Download to read offline
![x
Diabetes Medications for Adults With Type 2 Diabetes:
An Update
Structured Abstract
Objectives. To evaluate the comparative effectiveness and safety of monotherapy and
metformin-based combination therapy for type 2 diabetes.
Data sources. We searched MEDLINE®
, Embase®
, and the Cochrane Central Register of
Controlled Trials (CENTRAL) for English-language articles using the search developed for the
prior review (2011), with date restrictions of April 2009 through April 2015. We searched
Drugs@FDA and ClinicalTrials.gov for unpublished data.
Review methods. Two reviewers independently reviewed titles, abstracts, and full-text articles
to identify studies that assessed intermediate and clinical outcomes or safety for monotherapy
(metformin, sulfonylureas, thiazolidinediones, dipeptidyl peptidase-4 [DPP-4] inhibitors,
glucagon-like peptide-1 [GLP-1] agonists, and sodium glucose cotransporter-2 [SGLT-2]
inhibitors) or metformin-based combination therapy (metformin plus one of these monotherapy
drugs or insulin) comparisons. Two reviewers extracted data from included articles sequentially
using standardized protocols; risk of bias was assessed independently. Two reviewers graded the
strength of the evidence sequentially using a protocol adapted from the Grading of
Recommendations Assessment, Development, and Evaluation (GRADE) criteria.
Results. We included 216 studies and found moderate- or high-strength evidence for the
following. Hemoglobin A1c (HbA1c) reduction was similar across all monotherapy comparisons
and across metformin-based combination comparisons except DPP-4 inhibitors, which yielded
smaller reductions than metformin. Metformin, DPP-4 inhibitors, GLP-1 agonists, and SGLT-2
inhibitors reduced or maintained body weight, while sulfonylureas, thiazolidinediones, and
insulin increased weight; between-group differences ranged from 1 to 5 kilograms. SGLT-2
inhibitors and GLP-1 agonists plus metformin reduced systolic blood pressure by 3 to 5 mmHg
compared with metformin. Cardiovascular mortality in studies over 2 years in duration was 50 to
70 percent higher for sulfonylureas than metformin (risk difference, 0.1% to 2.9% in randomized
controlled trials). Sulfonylurea-based therapy increased the risk of total and severe hypoglycemia
versus most comparisons. Gastrointestinal adverse events were higher with metformin than other
drugs except GLP-1 agonists, which increased nausea/vomiting 1.5 times compared with
metformin. SGLT-2 inhibitors increased the risk of genital mycotic infections over other drugs.
The evidence did not support substantive conclusions for microvascular outcomes, congestive
heart failure, cancer, pancreatitis, or other safety outcomes.
Conclusions. Evidence from this updated systematic review supports metformin as firstline
therapy, given its beneficial effects on HbA1c, weight, and cardiovascular mortality (relative to
sulfonylureas) and its relative safety profile. In addition, evidence on comparative outcomes
associated with different medication classes can be used to facilitate personalized treatment
choices by patients and clinicians, guideline development, and decisionmaking by payers and
regulators.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-11-2048.jpg)
![ES-13
evidence was low for these outcomes because of less consistency in results across studies. It is of
note that losses to followup were greater than 20 percent in both RCTs. Losses to followup were
the same (20%) across arms in the study by Hong and colleagues (2013) and therefore not
anticipated to bias the comparison of arms.20
In A Diabetes Outcome Progression Trial
(ADOPT), losses to followup were higher in the sulfonylurea (44%) than the metformin (38%)
arm, with median followup of 3.3 years for the sulfonylurea arm versus 4.0 years for the
metformin arm.21
Therefore, study results were likely biased to the null, lending further support
to the inference that metformin was favored over sulfonylurea monotherapy.
Table D. Comparative effectiveness of sulfonylureas compared with metformin for long-term all-
cause mortality and cardiovascular mortality and morbidity—moderate strength of evidence or
consistent low-strength evidence
Outcome Range in RR From
RCTs
Range in RD From
RCTs
Adjusted HR From
Observational
Studies
SOE
All-cause mortality 1.0 to 2.1 (N = 2) 0.1% to 5.0% (N = 2) 1.2 to 1.9 (N = 7*) Low
CVD mortality 1.5 to 1.7 (N = 2) 0.1% to 2.9% (N = 2) 1.1 to 1.6 (N = 3) Moderate
CVD morbidity 0.7 to 1.4 (N = 2) -10.1% to 0.4% (N = 2) 1.1 to 3.3 (N = 5
†
) Low
CVD = cardiovascular disease; HR = hazard ratio; RCT = randomized controlled trial; RD = risk difference; RR = relative risk;
SOE = strength of evidence.
*One additional retrospective cohort study reported an odds ratio of 1.1.
†Additionally, 1 case-control study reported an odds ratio of 1.2.
Retinopathy, Nephropathy, and Neuropathy
While we found more evidence than in the prior report, there were still too few studies to
reach firm conclusions; all evidence for these outcomes was of low strength or insufficient.
Key Questions 3a and 3b: Comparative Safety
Of 145 studies identified for Key Question 3, 137 were RCTs and 8 were observational
(mainly retrospective cohort) studies. Most RCTs lasted a year or less, with only about 5 percent
lasting more than 2 years. Mean or median followup of the eight observational studies ranged
from 3 months to 5 years. The few longer studies were generally consistent with the shorter term
results; however, the losses to followup were often high (>20% in the majority of the longer
studies), making it difficult to draw firm long-term conclusions. Therefore, most safety
comparisons represent shorter term results unless specifically stated in the text or a figure.
Hypoglycemia
Sulfonylureas alone and in combination with metformin had a higher risk of mild, moderate,
or total hypoglycemia than any other monotherapies and metformin-based combinations for
which we identified evidence (Figure D). While studies were too heterogeneous for a meta-
analysis, sulfonylureas also had greater risk of hypoglycemia than GLP-1 receptor agonists
(range in odds ratio [OR], 3.1 to 5.3; range in risk difference [RD], 12% to 21%) and DPP-4
inhibitors (range in OR, 3.8 to 12.4; range in RD, 6% to 15%), with moderate strength of
evidence. In addition to the increased risk of hypoglycemia with metformin plus sulfonylurea
versus several comparators (Figure D), the combination of metformin plus sulfonylurea also had
greater risk of hypoglycemia compared with metformin monotherapy (range in OR, 2 to 17;
range in RD, 0% to 35%) and compared with the combination of metformin plus a GLP-1
receptor agonist (for studies lasting 104 to 234 weeks: range in OR, 3.4 to 7.1; range in RD, 15%
to 30%). When compared with metformin plus a basal or premixed insulin, metformin plus a](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-37-2048.jpg)
![ES-16
Figure E. Pooled odds ratios of gastrointestinal adverse events and strength of evidence for
monotherapy and metformin-based combination comparisons
*
CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; GI = gastrointestinal; GLP-1 = glucagon-like peptide-1
receptor agonists; H = high; M = moderate; Met = metformin; OR = odds ratio; RD = absolute risk difference; SGLT-2 =
sodium-glucose cotransporter-2 inhibitors; SOE = strength of evidence; SU = sulfonylurea; TZD = thiazolidinediones.
The width of the horizontal lines represents the 95% confidence intervals for each pooled analysis. Drug 1 is the reference group.
*
All results presented in this graph are based on short-term (less than 52 weeks) studies unless otherwise specified.
†
Based on studies with 104 weeks of followup.
Congestive Heart Failure
There was only one long-term trial, which lasted 4 years, and only a few observational
studies of medium quality with 6 to 8 years of followup that allow an assessment of the
comparative safety of diabetes medications regarding congestive heart failure. We found low
strength of evidence that the risk of congestive heart failure was 1.2 to 1.6 times as great with
thiazolidinediones as with sulfonylureas (pooled OR, 1.6; 95% CI, 0.96 to 2.8; range in RD, 0%
to 2%) or metformin (2 RCTs lasting less than a year with no events; 1 4-year RCT with an RD
of 3%; and range in hazard ratio of 1.2 to 1.5 in 2 observational studies). Despite recent concerns
about congestive heart failure with specific DPP-4 inhibitors, we found low or insufficient
strength of evidence on the comparative safety of this drug class for this outcome in studies
lasting less than 2 years (5 RCTs reporting no events in the DPP-4 inhibitor arms; 1 RCT with 1
event in the metformin plus DPP-4 inhibitor arm and none in the comparator arm; and 1 RCT of
metformin plus DPP-4 inhibitor vs. metformin plus sulfonylurea reporting fewer events in the
DPP-4 combination arm compared with the sulfonylurea combination arm [3 vs. 6 events]).](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-40-2048.jpg)
![ES-29
15. DerSimonian R, Laird N. Meta-analysis in
clinical trials. Control Clin Trials. 1986
Sep;7(3):177-88. PMID: 3802833.
16. Cornell JE, Mulrow CD, Localio R, et al.
Random-effects meta-analysis of
inconsistent effects: a time for change. Ann
Intern Med. 2014 Feb 18;160(4):267-70.
PMID: 24727843.
17. Owens DK, Lohr KN, Atkins D, et al.
AHRQ series paper 5: grading the strength
of a body of evidence when comparing
medical interventions--Agency for
Healthcare Research and Quality and the
Effective Health-Care Program. J Clin
Epidemiol. 2010 May;63(5):513-23. PMID:
19595577.
18. Vasilakou D, Karagiannis T, Athanasiadou
E, et al. Sodium-glucose cotransporter 2
inhibitors for type 2 diabetes: a systematic
review and meta-analysis. Ann Intern Med.
2013 Aug 20;159(4):262-74. PMID:
24026259.
19. Shyangdan DS, Royle P, Clar C, et al.
Glucagon-like peptide analogues for type 2
diabetes mellitus. Cochrane Database Syst
Rev [serial on the Internet]. 2011;(10).
http://onlinelibrary.wiley.com/doi/10.1002/1
4651858.CD006423.pub2/abstract.
20. Hong J, Zhang Y, Lai S, et al. Effects of
metformin versus glipizide on
cardiovascular outcomes in patients with
type 2 diabetes and coronary artery disease.
Diabetes Care. 2013 May;36(5):1304-11.
PMID: 23230096.
21. Kahn SE, Haffner SM, Heise MA, et al.
Glycemic durability of rosiglitazone,
metformin, or glyburide monotherapy. N
Engl J Med. 2006 Dec 7;355(23):2427-43.
PMID: 17145742.
22. Ahren B, Johnson SL, Stewart M, et al.
HARMONY 3: 104-week randomized,
double-blind, placebo- and active-controlled
trial assessing the efficacy and safety of
albiglutide compared with placebo,
sitagliptin, and glimepiride in patients with
type 2 diabetes taking metformin. Diabetes
Care. 2014 Aug;37(8):2141-8. PMID:
24898304.
23. Schernthaner G, Duran-Garcia S, Hanefeld
M, et al. Efficacy and tolerability of
saxagliptin compared with glimepiride in
elderly patients with type 2 diabetes: a
randomized, controlled study
(GENERATION). Diabetes Obes Metab.
2015 Jul;17(7):630-8. PMID: 25761977.
24. Schellhase KG, Koepsell TD, Weiss NS.
Glycemic control and the risk of multiple
microvascular diabetic complications. Fam
Med. 2005 Feb;37(2):125-30. PMID:
15690253.
25. Vijan S, Hofer TP, Hayward RA. Estimated
benefits of glycemic control in
microvascular complications in type 2
diabetes. Ann Intern Med. 1997 Nov
1;127(9):788-95. PMID: 9382399.
26. UK Prospective Diabetes Study (UKPDS)
Group. Intensive blood-glucose control with
sulphonylureas or insulin compared with
conventional treatment and risk of
complications in patients with type 2
diabetes (UKPDS 33). Lancet. 1998 Sep
12;352(9131):837-53. PMID: 9742976.
27. Liu SC, Tu YK, Chien MN, et al. Effect of
antidiabetic agents added to metformin on
glycaemic control, hypoglycaemia and
weight change in patients with type 2
diabetes: a network meta-analysis. Diabetes
Obes Metab. 2012 Sep;14(9):810-20. PMID:
22486990.
28. McIntosh B, Cameron C, Singh SR, et al.
Second-line therapy in patients with type 2
diabetes inadequately controlled with
metformin monotherapy: a systematic
review and mixed-treatment comparison
meta-analysis. Open Med. 2011;5(1):e35-48.
PMID: 22046219.
29. Kahn BB, Flier JS. Obesity and insulin
resistance. J Clin Invest. 2000
Aug;106(4):473-81. PMID: 10953022.
30. Purnell TS, Joy S, Little E, et al. Patient
preferences for noninsulin diabetes
medications: a systematic review. Diabetes
Care. 2014 Jul;37(7):2055-62. PMID:
24963113.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-53-2048.jpg)
![ES-30
31. Richter B, Bandeira-Echtler E, Bergerhoff
K, et al. Dipeptidyl peptidase-4 (DPP-4)
inhibitors for type 2 diabetes mellitus.
Cochrane Database Syst Rev [serial on the
Internet]. 2008;(2).
http://onlinelibrary.wiley.com/doi/10.1002/1
4651858.CD006739.pub2/abstract.
32. SHEP Cooperative Research Group.
Prevention of stroke by antihypertensive
drug treatment in older persons with isolated
systolic hypertension. Final results of the
Systolic Hypertension in the Elderly
Program (SHEP). JAMA. 1991 Jun
26;265(24):3255-64. PMID: 2046107.
33. UK Prospective Diabetes Study Group.
Tight blood pressure control and risk of
macrovascular and microvascular
complications in type 2 diabetes: UKPDS
38. BMJ. 1998 Sep 12;317(7160):703-13.
PMID: 9732337.
34. Adler AI, Stratton IM, Neil HA, et al.
Association of systolic blood pressure with
macrovascular and microvascular
complications of type 2 diabetes (UKPDS
36): prospective observational study. BMJ.
2000 Aug 12;321(7258):412-9. PMID:
10938049.
35. Wang JG, Staessen JA, Gong L, et al.
Chinese trial on isolated systolic
hypertension in the elderly. Systolic
Hypertension in China (Syst-China)
Collaborative Group. Arch Intern Med. 2000
Jan 24;160(2):211-20. PMID: 10647760.
36. Sacks FM, Svetkey LP, Vollmer WM, et al.
Effects on blood pressure of reduced dietary
sodium and the Dietary Approaches to Stop
Hypertension (DASH) diet. DASH-Sodium
Collaborative Research Group. N Engl J
Med. 2001 Jan 4;344(1):3-10. PMID:
11136953.
37. Wu J, Kraja AT, Oberman A, et al. A
summary of the effects of antihypertensive
medications on measured blood pressure.
Am J Hypertens. 2005 Jul;18(7):935-42.
PMID: 16053990.
38. Nauman J, Janszky I, Vatten LJ, et al.
Temporal changes in resting heart rate and
deaths from ischemic heart disease. JAMA.
2011 Dec 21;306(23):2579-87. PMID:
22187277.
39. Singh N. Diabetes, heart rate, and mortality.
J Cardiovasc Pharmacol Ther. 2002
Apr;7(2):117-29. PMID: 12075400.
40. Phung OJ, Schwartzman E, Allen RW, et al.
Sulphonylureas and risk of cardiovascular
disease: systematic review and meta-
analysis. Diabet Med. 2013
Oct;30(10):1160-71. PMID: 23663156.
41. Monami M, Genovese S, Mannucci E.
Cardiovascular safety of sulfonylureas: a
meta-analysis of randomized clinical trials.
Diabetes Obes Metab. 2013 Oct;15(10):938-
53. PMID: 23594109.
42. Nissen SE, Wolski K. Effect of rosiglitazone
on the risk of myocardial infarction and
death from cardiovascular causes. N Engl J
Med. 2007 Jun 14;356(24):2457-71. PMID:
17517853.
43. U.S. Food and Drug Administration. FDA
requires removal of some prescribing and
dispensing restrictions for rosiglitazone-
containing diabetes medicines. 2013.
www.fda.gov/downloads/Drugs/DrugSafety/
UCM381108.pdf. Accessed February 25,
2015.
44. Scirica BM, Bhatt DL, Braunwald E, et al.
Saxagliptin and cardiovascular outcomes in
patients with type 2 diabetes mellitus. N
Engl J Med. 2013 Oct 3;369(14):1317-26.
PMID: 23992601.
45. Green JB, Bethel MA, Armstrong PW, et al.
Effect of sitagliptin on cardiovascular
outcomes in type 2 diabetes. N Engl J Med.
2015 Jul 16;373(3):232-42. PMID:
26052984.
46. Zannad F, Cannon CP, Cushman WC, et al.
Heart failure and mortality outcomes in
patients with type 2 diabetes taking
alogliptin versus placebo in EXAMINE: a
multicentre, randomised, double-blind trial.
Lancet. 2015 May 23;385(9982):2067-76.
PMID: 25765696.
47. Duckworth W, Abraira C, Moritz T, et al.
Glucose control and vascular complications
in veterans with type 2 diabetes. N Engl J
Med. 2009 Jan 8;360(2):129-39. PMID:
19092145.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-54-2048.jpg)
![ES-31
48. Bonds DE, Miller ME, Bergenstal RM, et al.
The association between symptomatic,
severe hypoglycaemia and mortality in type
2 diabetes: retrospective epidemiological
analysis of the ACCORD study. BMJ.
2010;340:b4909. PMID: 20061358.
49. Holman RR, Farmer AJ, Davies MJ, et al.
Three-year efficacy of complex insulin
regimens in type 2 diabetes. N Engl J Med.
2009 Oct 29;361(18):1736-47. PMID:
19850703.
50. Budnitz DS, Shehab N, Kegler SR, et al.
Medication use leading to emergency
department visits for adverse drug events in
older adults. Ann Intern Med. 2007 Dec
4;147(11):755-65. PMID: 18056659.
51. Lago RM, Singh PP, Nesto RW. Congestive
heart failure and cardiovascular death in
patients with prediabetes and type 2 diabetes
given thiazolidinediones: a meta-analysis of
randomised clinical trials. Lancet. 2007 Sep
29;370(9593):1129-36. PMID: 17905165.
52. Singh S, Loke YK, Furberg CD. Long-term
risk of cardiovascular events with
rosiglitazone: a meta-analysis. JAMA. 2007
Sep 12;298(10):1189-95. PMID: 17848653.
53. Home PD, Pocock SJ, Beck-Nielsen H, et al.
Rosiglitazone evaluated for cardiovascular
outcomes in oral agent combination therapy
for type 2 diabetes (RECORD): a
multicentre, randomised, open-label trial.
Lancet. 2009 Jun 5;373(9681):2125-35.
PMID: 19501900.
54. GlaxoSmithKline. Highlights of Prescribing
Information: Avandia (rosiglitazone
maleate) tablets. 2014.
http://us.gsk.com/products/assets/us_avandia
.pdf. Accessed March 2, 2015.
55. Takeda Pharmaceuticals America.
Highlights of Prescribing Information: Actos
(pioglitazone) tables for oral use. 2013.
http://general.takedapharm.com/content/file/
pi.pdf?applicationcode=8a9c4571-a123-
4477-
91deb9cafe7d07e3&filetypecode=actospi.
Accessed March 2, 2015.
56. U.S. Food and Drug Administration. FDA
panel wants new DPP-4 inhibitor labels -
cardiovascular data warrant new risk
information for saxagliptin and alogliptin,
advisers say.
www.medpagetoday.com/PublicHealthPolic
y/ClinicalTrials/50990. Accessed July 25,
2015.
57. Clinicaltrials.gov. CAROLINA:
Cardiovascular Outcome Study of
Linagliptin Versus Glimepiride in Patients
With Type 2 Diabetes. 2010.
https://clinicaltrials.gov/ct2/show/NCT0124
3424. Accessed July 30, 2015.
58. Clinicaltrials.gov. Cardiovascular and Renal
Microvascular Outcome Study With
Linagliptin in Patients With Type 2 Diabetes
Mellitus (CARMELINA). 2013.
https://clinicaltrials.gov/ct2/show/NCT0189
7532. Accessed July 30, 2015.
59. U.S. Food and Drug Administration.
Highlights of Prescribing Information:
Victoza (liraglutide [rDNA origin] injection,
solution for subcutaneous use. 2011.
www.accessdata.fda.gov/drugsatfda_docs/la
bel/2011/022341s004lbl.pdf. Accessed
March 2, 2015.
60. U.S. Food and Drug Administration.
Highlights of Prescribing Information:
Tanzeum (albiglutide) for injection, for
subcutaneous use. 2014.
www.accessdata.fda.gov/drugsatfda_docs/la
bel/2014/125431s000lbl.pdf. Accessed
March 2, 2015.
61. U.S. Food and Drug Administration.
Highlights of Prescribing Information:
Bydureon (exenatide extended-release) for
injectable suspension. 2015.
www.fda.gov/safety/medwatch/safetyinform
ation/ucm400570.htm. Accessed August 7,
2015.
62. U.S. Food and Drug Administration.
Trulicity (dulaglutide) injection, for
subcutaneous use. 2015.
www.fda.gov/safety/medwatch/safetyinform
ation/ucm442202.htm. Accessed August 7,
2015.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-55-2048.jpg)
![1
Introduction
Background
Type 2 diabetes currently affects 9.3 percent of the US population, or 29.1 million people.1
The proportion of affected individuals in the US varies greatly by race and ethnicity: 16 percent
of American Indian/ Alaska Natives, 13 percent of non-Hispanic black Americans and Hispanic
Americans, 9 percent of Asian Americans, and 7 percent of non-Hispanic white Americans are
afflicted with diabetes. The vast majority of these cases are type 2 diabetes.2
Within these racial
categories, rates also vary substantially within sub-populations (e.g., South Asian-Americans and
East Asian-Americans).2
Estimates of diabetes incidence that include laboratory-diagnosed
diabetes, in addition to self-report, are higher than those reported by the US Centers for Disease
Control and Prevention.3
Encouragingly, most reports in the US and Europe suggest that the
incidence of disease has not been rising over the past decade.4
Similarly, the age at diagnosis has
been relatively stable at 55 years in non-Hispanic whites, and 49 years in non-Hispanic blacks
and Hispanics.5
Diabetes and its complications are a substantial public health burden, as they contribute
significantly to mortality, morbidity, and health care costs.1, 6
Costs related to diabetes were
approximately $245 billion in 2012.1
Complications of longstanding diabetes include the
microvascular complications of retinopathy and blindness, neuropathy, nephropathy, and end-
stage kidney disease. Diabetes is the most prevalent cause of new-onset blindness and new-onset
end-stage renal disease in adults in the US. Diabetes also contributes importantly to
macrovascular complications, including coronary artery disease, peripheral arterial disease, and
carotid artery disease, and increases the risk of cardiovascular-related death nearly two-fold.7
Lifestyle modification and pharmacologic therapy are the cornerstones of the management of
hyperglycemia for type 2 diabetes.8
Results from randomized controlled trials have established
that the risk of microvascular complications, particularly retinopathy, can be reduced with
glycemic control in patients with type 2 diabetes.9, 10
However, studies in the past decade have
suggested that using diabetes medications to achieve intensive glycemic control [hemoglobin
A1c (HbA1c) less than 7%] does not benefit cardiovascular morbidity and mortality11, 12
and may
harm patients, including those with important co-morbid conditions.13
Recent work also suggests
that the effects of intensive glucose lowering may vary across racial and ethnic groups.14
These
mixed results on the benefits and safety of glycemic control through pharmacologic therapy
suggest the need for further research, including investigation of the long-term impact of glucose
lowering therapies.
Even if questions about intensity of control are resolved, clinicians and other stakeholders
need to determine the optimal agent for glucose lowering. Given the ever-increasing literature
about type 2 diabetes medications and the recent approval of many new medications, an updated
systematic review evaluating the effects of these medications on intermediate and long-term
effectiveness and safety outcomes will be valuable to clinicians, patients, investigators, funders,
guideline developers, and payers. In this era of intensive, direct-to-consumer marketing of new
drugs, clinicians need a trustworthy source of comprehensive information about the comparative
effectiveness and safety of medications. This review seeks to provide information about
treatment options to a diverse set of clinicians, including family practitioners, general internists,
nurse practitioners, physician assistants, nurses, pharmacists, endocrinologists, cardiologists,
nephrologists, and others. Guideline developers may also find this review to be informative for
clinical practice guideline preparation. Patients and patient advocates will find the information](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-58-2048.jpg)
![2
valuable when making decisions about treatment options. Finally, investigators will be able to
use the results of this review to identify gaps in the literature and formulate original research
questions to fill these knowledge gaps.
Rationale for Update of Review on Comparative
Effectiveness of Diabetes Medications
The Effective Health Care (EHC) Program of the Agency for Healthcare Research and
Quality (AHRQ) has published two systematic reviews comparing monotherapies and
medication combinations for adults with type 2 diabetes.15, 16
In 2007, the AHRQ published its
first systematic review, including 216 studies, on this topic.15
This review concluded that most
diabetes medications approved by the U.S. Food and Drug Administration (FDA) had similar
effects on reducing HbA1c, and most drugs, except for metformin and acarbose, caused at least
modest increases in body weight. The sulfonylurea class was associated with an increased risk of
hypoglycemia, metformin with gastrointestinal problems, and the thiazolidinediones with heart
failure. Importantly, the literature was too sparse to support any conclusions about differential
effects of the oral diabetes medications on all-cause mortality, cardiovascular mortality and
morbidity, and microvascular complications. When asked by AHRQ to update that review in
2011, we identified an additional 140 randomized controlled trials and 26 observational studies.16
We found that most medications lowered HbA1c by 1 absolute percentage point, on average, but
metformin was more effective for HbA1c-lowering than the dipeptidyl-peptidase 4 (DPP-4)
inhibitors, a newer class of diabetes medications approved since the initial report. Mostly, the
two-drug combinations had similar effects on HbA1c reduction. Compared with metformin,
thiazolidinediones and sulfonylureas contributed to more weight gain. Sulfonylureas had a four-
fold higher risk of mild/moderate hypoglycemia compared with metformin alone, and, in
combination with metformin, had more than a five-fold increased risk of hypoglycemia when
compared with metformin plus thiazolidinediones. The risk of congestive heart failure was
higher with thiazolidinediones than with sulfonylureas, and the risk of bone fractures was higher
with thiazolidinediones than with metformin. Thus, the evidence continued to support use of
metformin as a first line agent, based on its effects on HbA1c and weight and side effect profile.
The risk of adverse effects was the main determinant of the risk-benefit balance for the two-drug
combinations.
Despite the addition of important evidence on the HbA1c-lowering and adverse effects of the
FDA-approved diabetes medications in 2011, data on the then recently-approved medication
classes (glucagon-like peptide-1 (GLP-1) agonists and DPP-4 inhibitors) were sparse, and data
on long-term outcomes for both older and newer medications were still lacking.17, 18
Based on
these prior systematic reviews, metformin has strong evidence to support its use as an initial
pharmacologic treatment for most patients with type 2 diabetes;7
It’s use as a first-line therapy
has been widely promoted by clinical practice guidelines.19-21
Not all patients, however, can
successfully use metformin due to contraindications to its use or intolerance of its side effects.
The evidence base regarding alternative monotherapies for these patients continues to evolve.
Since January 2010, one new medication class [the sodium-glucose cotransporter 2 (SGLT-2)
inhibitors, with three new medications] and several new DPP-4 inhibitors and GLP-1 receptor
agonists have been approved by the FDA. Also since 2010, additional data on previously-
approved medications have emerged that could change the balance of benefit and risk
attributable to these drugs or could alter the strength of evidence about some of the drug
comparisons previously reviewed.22-25
Including insulin, there are 10 medication classes with](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-59-2048.jpg)
![13
unable to draw conclusions about causality: long-term outcomes (KQ2), fractures, cancer, intermediate outcomes in studies
where there was a washout period of less than 3 months; and safety outcomes in studies where the washout period was less than a
month except for hypoglycemia, gastrointestinal side effects, and liver injury.
¶
We decided to include non-English language articles through the full text article review phase of the updated search and assess
the volume and content of these articles along with workload to determine if abstracting data from these articles would add value
to the review.
Data Extraction
We used a systematic approach to extract all data to minimize the risk of bias in this process.
We used standardized forms from the previous reviews as templates for data extraction and pilot
tested them for the new medications and outcomes (Appendix B). By creating standardized
forms for data extraction, we sought to maximize consistency in identifying all pertinent data
available for synthesis.
We double-reviewed all data abstracted from the studies. The second reviewer confirmed the
first reviewer’s abstracted data for completeness and accuracy. Reviewer pairs were formed to
include personnel with both clinical and methodological expertise. A third reviewer audited a
random sample of articles to ensure consistency in the data abstraction of the articles. Reviewers
were not masked to the authors of the articles, their respective institutions, nor the journals in
which their articles were published.
For all articles, the reviewers extracted information on the general study characteristics (e.g.,
study design, study period, and followup); study participants (e.g., age, sex, race, weight/body
mass index [BMI], hemoglobin A1c levels, and duration of diabetes); interventions (e.g., initial,
maximum, and mean doses, frequency of use, duration of use, and permissibility of treatment
intensification with additional therapies), comparisons; the method of ascertainment of safety
outcomes; and the outcome results, including measures of variability. We also collected data on
outcomes for the subgroups of interest: age, sex, race/ethnicity, and BMI.
For continuous outcomes, we extracted the mean difference between groups and a measure of
dispersion. If the between-group difference was not reported, we calculated the point estimate of
the difference using the mean difference from baseline for each group. If the mean difference
from baseline was not reported, we calculated this from the baseline and final values for each
group.32
If there were no measures of dispersion for the mean difference from baseline for each
group, we calculated the variance using the standard deviation of the baseline and final values,
assuming a correlation between baseline and final values of 0.5.
We entered all information from the article review process into a DistillerSR database
(Evidence Partners Inc., Ottawa, Canada). Reviewers entered comments into the system
whenever applicable. The DistillerSR database was used to maintain the data and to create
detailed evidence tables and summary tables. Data will later be uploaded into the Systematic
Review Data Repository.
Risk of Bias Assessment of Individual Studies
Two independent reviewers assessed study quality. We assessed the risk of bias in individual
randomized controlled trials (RCTs) using the Jadad criteria consistent with the prior report.33
Although newer quality assessment tools exist, we felt that continuing to use the Jadad criteria
would be adequate and consistent with our previous methods. We used the Downs and Black tool
for assessment of risk of bias for non-randomized trials and observational studies.34
Given that
observational studies with a high risk of bias add little value to a systematic review of
effectiveness,35
we included only medium- and high-quality observational studies as determined](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-70-2048.jpg)
![22
Key Points and Evidence Grades for Intermediate Outcomes
Hemoglobin A1c
Monotherapy Comparisons
Most oral diabetes medications had similar efficacy in achieving reductions in
hemoglobin A1c (HbA1c).
o In the prior report, the strength of evidence was graded as high that metformin was
similar to sulfonylurea (pooled between-group difference of 0.1%; 95% confidence
interval [CI], -0.1% to 0.3%). Therefore, we did not update this comparison for
HbA1c in this review.
o The strength of evidence (SOE) was graded as high that metformin was similar to
thiazolidinedione (pooled between-group difference of -0.04%; 95% CI,
-0.11% to 0.03%).
o Thiazolidinediones performed similarly to sulfonylureas (pooled between-group
difference of -0.04%; 95% CI, -0.13% to 0.06%). (SOE: High)
o The SOE was graded as low or insufficient for all the monotherapy comparisons of
the newer classes of sodium-glucose cotransporter (SGLT-2) inhibitors and glucagon-
like peptide-1 (GLP-1) agonists, and will warrant further study.
The one exception was that metformin had a greater reduction in HbA1c compared with
dipeptidyl peptidase-4 (DPP-4) inhibitors (pooled between-group difference of
-0.4%; 95% CI, -0.5% to -0.3%). (SOE: High)
Metformin-Based Combination Comparisons
The combination of metformin plus GLP-1 receptor agonists reduced HbA1c more than
metformin plus DPP-4 inhibitors, with a pooled between-group difference of -0.65%
(95% CI, -0.75% to -0.54%) in the short-term. (SOE: Moderate)
Most other combination therapy comparisons had either no significant or no clinically
meaningful (<0.3%) between-group differences in HbA1c between arms.
The evidence was graded as moderate for the following comparisons: metformin plus a
thiazolidinedione versus metformin plus a sulfonylurea, metformin plus a
thiazolidinedione versus metformin plus a DPP-4 inhibitor, metformin plus a
sulfonylurea versus metformin plus an SGLT-2 inhibitor, metformin plus a DPP-4
inhibitor versus metformin plus an SGLT-2 inhibitor, and metformin plus a DPP-4
inhibitor versus metformin plus a GLP-1 receptor agonist.
Despite the clinical interest in comparing metformin plus injectables, there was
insufficient or low strength of evidence on glycemic control for the following
comparisons: metformin plus the GLP-1 receptor agonists versus metformin plus basal or
premixed insulin, and metformin plus premixed insulin versus metformin plus basal
insulin.
Weight
Monotherapy Comparisons
In the 2011 report, metformin had greater weight reduction than thiazolidinediones
(pooled mean between-group difference of -2.6 kg; 95% CI, -4.1 kg to -1.2 kg) or](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-79-2048.jpg)
![99
Figure 39. Pooled odds ratio of short-term all-cause mortality comparing metformin with
pioglitazone
CI = confidence interval; Group 1 = metformin; Group 2 = thiazolidinediones; OR = odds ratio
Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing
more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The
diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
Two RCTs compared the effects of metformin with rosiglitazone on all-cause mortality.50, 59
The A Diabetes Outcome Progression Trial (ADOPT) randomized participants with recently-
diagnosed, untreated type 2 diabetes from 488 different centers in the United States, Canada, and
Europe to rosiglitazone, metformin, or glyburide and had long-term follow up.50
Mortality was
slightly lower in the metformin (31/1454; 2.1%) versus rosiglitazone (34/1456; 2.3%) arm with
median followup of 4.0 years.50
The actual number of participants for which ADOPT ascertained
mortality is unclear, and withdrawals were high across the arms: 37 percent (rosiglitazone) and
38 percent (metformin).50
The second trial was 32 weeks in duration and reported no deaths in
the metformin or rosiglitazone arms.59
Observational Studies
Two retrospective cohort studies compared the effects of initiating thiazolidinediones and
metformin (Table 19).233, 234
One study found no significant difference in all-cause mortality for
metformin and thiazolidinediones.233
The second study found a significantly increased risk of all-
cause mortality among women but not among men for rosiglitazone versus metformin.234
[SOE: Low for pioglitazone (short-term mortality); Neither metformin nor pioglitazone
favored] (SOE: Low for rosiglitazone; Metformin favored compared with rosiglitazone)](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-156-2048.jpg)
![113
Figure 46. Pooled odds ratio for long-term all-cause mortality comparing a combination of
metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor
CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = combination of metformin plus a sulfonylurea; Group 2 =
combination of metformin plus a dipeptidyl peptidase-4 inhibitor; Met = metformin; OR = odds ratio; SU = sulfonylurea
Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing
more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The
diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
Two additional RCTs evaluated this comparison but were not included in the meta-analysis
because of their shorter durations.190, 193
One trial with 52 weeks of followup conducted among
persons (predominantly men) greater than 65 years of age reported one death in each arm (1/360
(0.3%) in the metformin plus sulfonylurea arm; 1/360 (0.3%) in the metformin plus saxagliptin
arm).193
The other trial had 30 weeks of followup and reported one death in the metformin plus
sulfonylurea arm (1/519, 0.2%) and no deaths in the metformin plus sitagliptin arm (0/516,
0%).190
A single retrospective cohort study in the Danish National Registry reported a significantly
decreased risk of death among metformin plus DPP-4 inhibitor users (n=11,138) versus
metformin plus sulfonylurea users (n=25,092) with median follow up of 2.1 years (adjusted rate
ratio, 0.65; 95% CI, 0.54 to 0.8).227
(SOE: Low; Combination of metformin plus a DPP-4
inhibitor favored for long-term mortality; SOE: Insufficient for short-term mortality)
Combination of Metformin Plus a Sulfonylurea Versus a Combination of
Metformin Plus an SGLT-2 Inhibitor
Three long-term RCTs (reported in four publications), each with a duration of 104 to 208
weeks, reported on all-cause mortality for this comparison.54, 199-201
Mortality rates were low
across the studies.
An extension of Nauck 2011, with extremely high losses to followup, reported a higher rate
of mortality in the metformin plus sulfonylurea (5/408, 1.2%) versus metformin plus SGLT-2
inhibitor (2/406, 0.5%) arm at 208 weeks.54
Meta-analysis of the data from these trials at 104 weeks suggested that long-term all-cause
mortality [which was low (<1%) across studies] was similar for metformin plus SGLT-2](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-170-2048.jpg)
![122
Evidence for Cardiovascular Mortality
Monotherapy Comparisons
Metformin Versus Thiazolidinediones
Three RCTs compared metformin with thiazolidinediones and did not find differences in
cardiovascular mortality.50, 63, 70
Two of the RCTs were small, lasted less than one year, and did
not report any cardiovascular deaths (Table 29).63, 70
Studies were not combined because of
different lengths of followup and different thiazolidinediones under study. ADOPT was the
single long-term RCT (median followup of 4.0 years): the actual number of participants for
which ADOPT ascertained CVD mortality is unclear and withdrawals were high across the arms
[37% (rosiglitazone) and 38% (metformin)].50
(SOE: Low; Neither favored)
Table 29. Randomized controlled trials comparing metformin with thiazolidinediones on
cardiovascular mortality
Author, Year Followup Number of Events/N (%)
in the Metformin Arm
Number of Events/N (%) in the
Thiazolidinedione Arm
Lawrence, 2004
63
24 weeks 0/20 (0%) 0/20 (pioglitazone) (0%)
Erem, 2014
70
48 weeks 0/19 (0%) 0/19 (pioglitazone) (0%)
Kahn, 2006
50
4 years (median) 2/1454 (0.1%) 1/1456 (0.1%) (rosiglitazone)
Metformin Versus Sulfonylureas
Two high-quality RCTs compared metformin with sulfonylureas and reported on
cardiovascular mortality.50, 231
ADOPT, conducted among patients with newly diagnosed and
untreated diabetes (N=2895), reported a slightly higher incidence of fatal MI in the glyburide
(3/1441, 0.2%) versus the metformin (2/1454, 0.1%) arm (glyburide vs. metformin: calculated
RR, 1.5 (95% CI, 0.3 to 9.0); calculated between-group difference, 0.1%). Median followup was
4.0 years for the metformin (maximum dose 2,000 mg; mean dose not reported) arm and 3.3
years for the glyburide (maximum dose 15 mg; mean dose not reported) arm, and losses to
followup were also differential for the metformin (38%) arm vs. the glyburide (44%) arm.50
The
smaller RCT was conducted in China among patients with known coronary heart disease
(clinical evidence of acute MI or coronary stenosis >50% on angiogram) and also reported a
higher risk of cardiovascular mortality in the sulfonylurea (glipizide, mean dose 28.3 mg) arm
(11/148, 7.4%) compared with the metformin (mean dose 1,400 mg) arm (7/156, 4.5%) over 2.8
years.231
We calculated the RR of cardiovascular mortality comparing sulfonylurea with
metformin to be 1.66 (95% CI, 0.66 to 4.16) and the between-group difference to be 2.9 percent.
Losses to followup were 21 percent for each arm of this trial.231
Losses to follow up were the
same (20%) across arms in Hong et al.,231
decreasing the risk of non-conservative bias due to
losses of follow up across arms. In ADOPT, given differential losses to followup and followup
duration across arms, study results were likely biased to the null, lending further support to the
inference that metformin was favored over sulfonylurea monotherapy.
Three retrospective cohort studies, analyzing two cohorts compared metformin with a
sulfonylurea, and all found a higher risk of cardiovascular mortality for sulfonylurea users versus
metformin users (Table 30).225, 229, 230
To account for confounding, two of these studies used
propensity score matching,225, 229
and one used multivariate regression.230
(SOE: Moderate;
Metformin favored for long-term CVD mortality)](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-179-2048.jpg)
![159
Combination of Metformin Plus a Thiazolidinedione Versus a Combination of
Metformin Plus a GLP-1 Receptor Agonist
A single RCT (N=514) with 26 weeks of followup reported on percent change in urine
albumin-to-creatinine ratio for the combination of metformin plus pioglitazone and the
combination of metformin plus exenatide and found a reduction in urine albumin-to-creatinine
from baseline for the metformin plus exenatide arm [-16% (95% CI, -28% to -2%)] and no
significant reduction for the metformin plus pioglitazone arm [-4% (95% CI, -17% to 12%)].188
This study had approximately 20 percent losses to followup in both arms and did not use an
intention-to-treat analysis for this outcome.188
(SOE: Low; Combination of metformin plus a
GLP-1 receptor agonist favored)
Combination of Metformin Plus a Sulfonylurea Versus a Combination of
Metformin Plus a GLP-1 Receptor Agonist
A small RCT conducted in 42 participants with baseline microalbuminuria compared change
in 24-hour urine albumin over 16 weeks for metformin plus sub-maximally dosed glimepiride
and metformin plus exenatide. The authors reported a significant reduction in urine albumin in
the metformin plus exenatide arm (-42 mg/day; P for change for baseline <0.01) compared to the
metformin plus glimepiride arm (5 mg/day; P for change from baseline >0.05; calculated
between group difference, -37 mg/day; P for between-group difference <0.001).202
This small
study had more than 20 percent losses to followup across arms and did not use an intention-to-
treat analysis.202
(SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored)
Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of
Metformin Plus a GLP-1 Receptor Agonist
A single RCT (N=514) with 26 weeks of followup reported on percent change in urine
albumin-to-creatinine ratio for the combination of metformin plus sitagliptin and the
combination of metformin plus exenatide.188
The trial found a statistically significant reduction
in urine albumin-to-creatinine from baseline for the metformin plus exenatide arm (-16%; 95%
CI, -28% to -2%) but not for the metformin plus sitagliptin arm (-6.9%; 95% CI, -20% to unclear
but greater than 0%).188
This study had approximately 20 percent losses to followup in both arms
and did not use an intention-to-treat analysis for this outcome.188
(SOE: Low; Combination of
metformin plus a GLP-1 receptor agonist favored)
Strength of Evidence for Nephropathy
We found low or insufficient strength of evidence on nephropathy outcomes for all
comparisons of interest as described in the Key Points, Table 54, and Table 55.
The evidence on nephropathy was limited by the lack of studies. For RCTs, major study
limitations included small sample sizes and high rates of withdrawals (>20%), without use of an
intention-to-treat approach. We could usually not determine consistency because of a lack of
studies, and the evidence on all comparisons was imprecise because of insufficient sample size.
We did not detect reporting bias. However, the small number of studies limited our ability to
assess publication bias. Many of the studies did not provide measures of dispersion for
nephropathy outcomes, but we did not believe that this was actually a source of selective analysis
reporting bias as much as a reflection of a lack of a focus on reporting of these outcomes given
that they were not primary outcomes.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-216-2048.jpg)
![162
Evidence for Neuropathy
For the neuropathy analyses, we included studies where newly developed neuropathy was
reported for each treatment group. Four short trials reported on neuropathy as an adverse
outcome.121, 142, 154, 183
Monotherapy Comparisons
Metformin Versus DPP-4 Inhibitors
A single RCT compared the effects of metformin with alogliptin on undefined neuropathy.
At 26 weeks of followup, one participant developed unspecified neuropathy in each of the
alogliptin arms [12.5 mg (n=213) and 25 mg (n=210)]); neuropathy was not reported on in the
metformin arm.154
(SOE: Insufficient)
Metformin Versus a Metformin-Based Combination Comparisons
Metformin Versus a Combination of Metformin Plus a Thiazolidinedione
In a single RCT of 26 weeks duration, one withdrawal owing to undefined neuropathy
occurred in the metformin arm (n=34); no events were reported on in the two metformin plus
rosiglitazone arms (n=35 and n=36).121
(SOE: Insufficient)
Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor
One RCT reported the occurrence of undefined diabetic neuropathy as 2.1 percent among
participants in the metformin arm (n=94) and 4.2 percent in the metformin plus sitagliptin arm
(n=96) at 30 weeks.142
(SOE: Low; Metformin favored)
Metformin-Based Combination Comparisons
Combination of Metformin Plus a Thiazolidinedione Versus a Combination of
Metformin Plus a Sulfonylurea
In a 6-month trial, neuropathy was described but was not a pre-specified outcome.183
One of
103 participants in the metformin plus thiazolidinedione arm developed neuropathy, and none of
the 80 participants in the metformin plus sulfonylurea arm developed neuropathy.183
(SOE: Low;
Neither favored)
Strength of Evidence for Neuropathy
The evidence grading for neuropathy is summarized in Table 56.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-219-2048.jpg)
![170
Table 58. Studies comparing metformin with sulfonylureas for hypoglycemia
Author, Year
Study Design
Followup Metformin
(Dose in mg*)
SU (Dose in
mg*)
Definition of
Hypoglycemia
Results
†
(Metformin Vs
SU)
Hermann,
1994
134
RCT
24 weeks Metformin
(max 3000)
Glyburide
(max 14)
Severe (based on
clinical findings or
available BG)
8/38 (21.1%) vs 12/34
(35.3%); P = 0.18
Hong, 2013
231
RCT
36 months Metformin
(max 1500;
mean 1400)
Glipizide
(max 30;
mean 28.3)
Severe (required
assistance and/or
PG < 56 mg/dL [3.1
mmol/L])
3/156 (1.9%) vs 4/148
(2.7%)
Chien, 2007
138
RCT
16 weeks Metformin
(max 2000;
mean 1910)
Glyburide
(max 20;
mean 19)
Mild-moderate
(symptomatic or BG
< 60 mg/dL)
0/25 (0%) vs 0/23 (0%)
Blonde,
2002
131
RCT
16 weeks Metformin
(max 2000)
Glyburide
(fixed at 10)
Symptomatic and
FSG <= 60 mg/dL
1/153 (1%) vs. 3/164
(2%)
Garber,
2003
129
RCT
16 weeks Metformin
(max 2000)
Glyburide
(max 10)
Symptoms
suggesting
hypoglycemia
29/164 (18%) vs. 98/151
(65%)
Marre‡
,
2002
132
RCT
16 weeks Metformin
(max 2000)
Glibenclamid
e (max 20)
Symptoms or labs 0/104 (0%) vs. 7/103
(7%)
Garber,
2002
133
RCT
20 weeks Metformin
(max 2000)
Glyburide
(max 10)
Symptomatic and BG
< 60 mg/dL
0/159 (0%) vs. 10/160
(6%)
DeFronzo
‡
,
1995
137
RCT
29 weeks Metformin
(max 2500)
Glyburide
(max 20)
Not reported 4/210 (2%) vs. 6/209
(3%)
Kahn, 2006
50
RCT
4 years Metformin
(max 2000)
Glyburide
(max 15)
Total (self-reported) 168/1454 (11.6%) vs
557/1441 (38.7%)
Wright,
2006
258
RCT
6 years Metformin
(max 2550)
Glyburide
(max 20)
Mild to severe (not
just transient
symptoms)
Mean annual percentage
0.3% among 290 patients
vs 1.2% among 1418
patients
Yamanouchi‡
,
2005
60
RCT
12 weeks Metformin
(fixed at 750)
Glimepiride
(max 2)
Not reported 0/39 (0%) vs. 1/37 (3%)
Charpentier‡
,
2001
136
RCT
20 weeks Metformin
(fixed at 2550)
Glimepiride
(max 6)
Symptomatic 8/75 (11%) vs. 17/150
(11%)
Derosa,
2004
257
RCT
48 weeks Metformin
(max 3000)
Glimepiride
(max 4)
Mild-moderate (not
specified)
0/75 (0%) vs 0/73 (0%)
Yoon‡
, 2011
74
RCT
48 weeks Metformin
(mean 1234.2;
max 2000)
Glimepiride
(mean 4.5;
max: 8)
Symptomatic 4/114 (4%) vs. 23/118
(19%)
Weir, 2011
259
Retrospective
cohort
3 months Metformin
(NR)
Glyburide
(NR)
Total (presented to
an emergency room
or hospital with an
admission diagnosis
of hypoglycemia)
Among patients with
normal renal function
27/572 (4.7%) vs 53/193
(27.5%); aOR = 9.0 (95%
CI, 4.9 to 16.4)
Among patients with
impaired renal function
29/580 (5.0%) vs 109/444
(24.5%); aOR = 6.0 (95%
CI, 3.8 to 9.5)
aOR = adjusted odds ratio; BG = blood glucose; CI = confidence interval; max = maximum; mg = milligrams; mg/dL =
milligrams per deciliter; mmol/L = millimoles per liter; PG = plasma glucose; RCT = randomized controlled trial; SU =
sulfonylurea
* All doses were titrated, unless otherwise stated.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-227-2048.jpg)
![173
Table 59. Randomized controlled trials comparing metformin with GLP-1 receptor agonists on
hypoglycemia
Author, Year Followup
(Weeks)
Metformin
(Dose*)
GLP-1
Receptor
Agonist
(Dose*)
Definition of
Hypoglycemia
Results
†
(Metformin Vs
GLP-1 Receptor
Agonist)
Russell-Jones,
2012
73
26 Metformin
(max 2500
mg)
Exenatide
(fixed at 2.0
mg weekly)
Mild-moderate (signs or
symptoms associated
with BG < 3.0 mmol/L
(either self-treated or
resolved
independently))
Severe
‡
Total (signs or
symptoms)
0/246 (0%) vs 5/248
(2.0%)
0/246 (0%) vs 0/248 (0%)
10/246 (4.1%) vs 13/248
(5.2%)
Umpierrez,
2014
91
52 Metformin
(max 2000
mg or ≥ 1500
mg
depending on
tolerability)
Dulaglutide
(fixed at 0.75
mg weekly)
Total (signs or
symptoms and/or PG ≤
70 mg/dL [3.9 mmol/L])
Severe (required third
party assistance)
34/268 (12.7%) vs 30/270
(11.1%)
0/268 (0%) vs 0/270 (0%)
Umpierrez,
2014
91
52 Metformin
(max 2000
mg or ≥ 1500
mg
depending on
tolerability)
Dulaglutide
(fixed at 1.5
mg weekly)
Total (signs or
symptoms and/or PG ≤
70 mg/dL [3.9 mmol/L])
Severe (required third
party assistance)
34/268 (12.7%) vs 33/269
(12.3%)
0/268 (0%) vs 0/269 (0%)
Yuan, 2012
92
26 Metformin
(max 2000
mg)
Exenatide
(max 2.0 mg)
Mild (not specified)
Severe (required third
party assistance or
hospital treatment)
1/26 (3.8%) vs 4/33
(12.1%)
0/26 (0%) vs 0/33 (0%)
BG = blood glucose; GLP-1 = glucagon-like peptide-1; max = maximum; mg = milligrams; mg/dL = milligram per deciliter;
mmol/L = millimole per liter; PG = plasma glucose
* All doses were titrated, unless otherwise stated.
† Results are presented as n/N (%) unless otherwise stated.
‡ Severe hypoglycemia was defined as symptoms resulting in loss of consciousness or seizure that showed prompt recovery after
glucose administration, or documented blood glucose less than 3.0 mmol/L that required the assistance of another person because
of severe impairment in consciousness.
Thiazolidinediones Versus Sulfonylureas
Nine RCTs compared thiazolidinedione with sulfonylurea monotherapy and reported on
hypoglycemia (Table 60).50, 60, 74, 94, 95, 100, 103, 217, 253
Results from the meta-analysis of five sufficiently-homogeneous, short-term RCTs showed
that the risk of total hypoglycemia was higher for sulfonylurea compared with thiazolidinedione
monotherapy (pooled OR for sulfonylurea compared with thiazolidinedione monotherapy, 6.31;
95% CI, 4.08 to 9.76) (Figure 57). We did not find evidence of significant statistical
heterogeneity, and removal of any one study did not change the overall inference.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-230-2048.jpg)
![176
Table 61. Randomized controlled trials comparing thiazolidinediones with DPP-4 inhibitors on
hypoglycemia
Author, Year Followup
(Weeks)
TZD (Dose*) DPP-4
Inhibitor
(Dose*)
Definition of
Hypoglycemia
Results
†
(TZD Vs DPP-4
Inhibitor)
Alba, 2013
48
12 Pioglitazone
(fixed at 30
mg)
Sitagliptin
(fixed at 100
mg)
Total (all reports of
hypoglycemia; no
glucose measurement
required)
2/54 (3.7%) vs 0/52 (0%)
Rosenstock,
2010
104
26 Pioglitazone
(fixed at
30mg)
Alogliptin
(fixed at 25
mg)
Severe (required third
party assistance)
0/163 (0%) vs 0/164 (0%)
Russell-Jones,
2012
73
26 Pioglitazone
(max 45 mg)
Sitagliptin
(fixed at 100
mg)
Total (signs or
symptoms)
Severe
‡
6/163 (3.7%) vs 5/163
(3.1%)
0/163 (0%) vs 0/163 (0%)
DPP-4 = dipeptidyl peptidase-4; mg = milligrams; TZD = thiazolidinedione
* All doses were titrated, unless otherwise stated.
† Results are presented as n/N (%) unless otherwise stated.
‡ Severe hypoglycemia was defined as symptoms resulting in loss of consciousness or seizure that showed prompt recovery after
glucose administration, or documented blood glucose less than 3.0 mmol/L that required the assistance of another person because
of severe impairment in consciousness.
Thiazolidinediones Versus GLP-1 Receptor Agonists
Two RCTs (26 and 48 weeks in duration) found higher rates of non-severe hypoglycemia for
exenatide compared to pioglitazone but no difference in severe hypoglycemia (no events) across
arms.73, 105
In the 26-week trial 0/163 (0%) experienced mild hypoglycemia in the pioglitazone
arm versus 5/248 (2%) in the exenatide arm.73
In the 48-week trial, corresponding event
(symptoms with blood glucose <3.9 mmol/l [<70 mg/dl]) rates were 13/142 (9.2%) and 5/136
(3.7%) in the exenatide and pioglitazone arms, respectively.105
(SOE: Low; Thiazolidinediones
favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe
hypoglycemia)
Sulfonylureas Versus DPP-4 Inhibitors
Comparisons for both mild and severe hypoglycemia favored the DPP-4 inhibitor arms over
sulfonylureas (for mild, moderate, or total hypoglycemia for SU vs. DPP-4 inhibitors: range in
OR 3.8 to 12.4; range in RD 6% to 15%). Four RCTs examined hypoglycemia with this
comparison; differences in followup length and definitions of hypoglycemia precluded a meta-
analysis (Table 62). (SOE: Moderate; DPP-4 inhibitors favored for mild, moderate, or total
hypoglycemia) (SOE: Low; DPP-4 inhibitors favored for severe hypoglycemia)](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-233-2048.jpg)
![177
Table 62. Randomized controlled trials comparing sulfonylureas with DPP-4 inhibitors on
hypoglycemia
Author, Year Followup
(Weeks)
SU (Dose*) DPP-4
Inhibitor
(Dose*)
Definition of
Hypoglycemia
Results
†
(SU Vs DPP-4
Inhibitor)
Arjona Ferreira,
2013
107
58 Glipizide
(max 20 mg;
mean 7.7 mg)
Sitagliptin
(fixed at 50
mg for those
with
moderate
renal
insufficiency
and 25 mg for
those with
severe renal
insufficiency)
Severe (required third
party assistance or
medical intervention or
exhibited markedly
depressed level of
consciousness, loss of
consciousness, or
seizure)
Total (signs and/or
symptoms)
6/212 (2.8%) vs 3/210
(1.4%)
36/212 (17%) vs 13/210
(6.2%); P < 0.001
Gupta, 2013
260
24 Glimepiride
(max 4 mg)
Sitagliptin
(max 200 mg)
Total (not specified) 11 episodes among 71
patients vs 3 episodes
among 77 patients
Barnett,
2012
106
34 Glimepiride
(max 4 mg)
Linagliptin
(fixed at 5
mg)
Mild-moderate
(Symptoms and/or PG
≤ 70 mg/dL [3.9 mmol/])
Severe (required third
party assistance)
5/64 (7.8%) vs 3/137
(2.2%)
0/64 (0%) vs 0/137 (0%)
Scott, 2007
108
12 Glipizide
(max 20 mg)
Sitagliptin
(fixed at 25
mg)
Total (self-report and
glucose measurements)
21/123 (17.1%) vs 5/123
(4.1%)
Scott, 2007
108
12 Glipizide
(max 20 mg)
Sitagliptin
(fixed at 50
mg)
Total (self-report and
glucose measurements)
21/123 (17.1%) vs 5/123
(4.1%)
Scott, 2007
108
12 Glipizide
(max 20 mg)
Sitagliptin
(fixed at 100
mg)
Total (self-report and
glucose measurements)
21/123 (17.1%) vs 2/122
(1.6%)
DPP-4 = dipeptidyl peptidase-4; max = maximum; mg = milligrams; mg/dL = milligrams per deciliter; mmol/L = millimole per
liter; PG = plasma glucose; SU = sulfonylurea
* All doses were titrated, unless otherwise stated.
† Results are presented as n/N (%) unless otherwise stated.
Sulfonylureas Versus GLP-1 Receptor Agonists
Sulfonylureas had greater risk of mild to moderate hypoglycemia compared with GLP-1
receptor agonists (range in OR 3.1 to 5.3; range in RD 12% to 21%) (Table 63). Five studies
assessed this outcome and could not be pooled because of heterogeneity in outcome definitions
and followup length. No study reported any events of severe hypoglycemia. (SOE: Moderate;
GLP-1 receptor agonists favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither
favored for severe hypoglycemia)](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-234-2048.jpg)
![178
Table 63. Randomized controlled trials comparing sulfonylureas with GLP-1 receptor agonists on
hypoglycemia
Author, Year Followup
(Weeks)
SU (Dose*) GLP-1
Receptor
Agonist
(Dose*)
Definition of
Hypoglycemia
Results
†
(SU Vs GLP-1
Receptor Agonist)
Garber, 2011
113
104 Glimepiride
(max 8mg)
Liraglutide
(max 1.2 mg)
Mild-moderate (did not
require assistance, BG
< 56 mg/dL [3.1
mmol/L])
Severe (required third-
party assistance)
64/248 (25.8%) vs 30/251
(12%)
0/248 (0%) vs 0/251 (0%)
Garber, 2011
113
104 Glimepiride
(max 8 mg)
Liraglutide
(max 1.8 mg)
Mild-moderate (did not
require assistance, BG
< 56 mg/dL [3.1
mmol/L])
Severe (required third-
party assistance)
64/248 (25.8%) vs 25/247
(10.1%)
0/248 (0%) vs 1/247
(0.4%)
‡
Kaku, 2011
110
52 Glibenclamid
e (fixed at
1.25 -2.5 mg)
Liraglutide
(max 0.9 mg)
Mild-moderate (self-
treated)
Severe (required third
party assistance)
1.10 events per patient-
year vs 0.19 events per
patient-year
0/132 (0%) vs 0/268 (0%)
Madsbad,
2004
111
12 Glimepiride
(max 4 mg)
Liraglutide
(Fixed (0.60
mg))
Mild-moderate (BG <
2.8 mmol/L)
4/26 (15.4%) vs 1/30
(3.3%)
Madsbad,
2004
111
12 Glimepiride
(max 4 mg)
Liraglutide
(Fixed (0.75
mg))
Mild-moderate (BG <
2.8 mmol/L)
4/26 (15.4%) vs 0/28 (0%)
Seino, 2010
109
24 Glibenclamid
e (max 2.5
mg)
Liraglutide
(max 0.9 mg)
Mild-moderate
(symptoms)
Mild-moderate
(symptoms and BG <
3.1 mmol/L)
Severe (required third
party assistance)
45/132 (34.1%) vs 36/268
(13.4%)
29/132 (22%) vs 22/268
(8.2%)
0/132 (0%) vs 0/268 (0%)
BG = blood glucose; GLP-1 = glucagon-like peptide-1; max = maximum; mg = milligrams; mg/dL = milligrams per deciliter;
mmol/L = millimole per liter; SU = sulfonylurea
* All doses were titrated, unless otherwise stated.
† Results are presented as n/N (%) unless otherwise stated.
‡ Event in the context of insulin infusion as part of a “sub-study” procedure.
DPP-4 Inhibitors Versus SGLT-2 Inhibitors
One study assessed hypoglycemia for the comparison of DPP-4 inhibitors versus SGLT-2
inhibitors.114
The study compared sitagliptin with empagliflozin at 24 weeks, with one of 223
patients (<1%) in the sitagliptin arm with any hypoglycemia, one of 224 patients (<1%) in the 10
mg empagliflozin arm, and one of 223 patients (<1%) in the 25 mg empagliflozin arm. No](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-235-2048.jpg)
![179
patients experienced severe hypoglycemia. (SOE: Low; Neither favored for mild, moderate, or
total hypoglycemia)
DPP-4 Inhibitors Versus GLP-1 Receptor Agonists
One 26-week RCT assessed hypoglycemia for the comparison of DPP-4 inhibitors versus
GLP-1 receptor agonists.73
Investigators compared sitagliptin with exenatide at 26 weeks. Total
hypoglycemia was slightly higher for the GLP-1 receptor agonist arm: 13/248 (5.2%) patients in
the exenatide arm vs. 5/163 (3.1%) in the sitagliptin arm. Mild hypoglycemia was also higher in
the exenatide (5/248, 2.0%) than the sitagliptin arm (0/163, 0%). No patients had severe
hypoglycemia in either arm. (SOE: Low; DPP-4 inhibitors favored for mild, moderate, or total
hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia)
Metformin Versus Metformin-Based Combination Comparisons
Metformin Versus a Combination of Metformin Plus a Thiazolidinedione
More patients experienced hypoglycemia in the combination arm than in the metformin-
alone arm. We combined eight sufficiently-homogeneous, short-term RCTs and found an
increased odds of mild or moderate hypoglycemia for metformin plus thiazolidinedione versus
metformin alone (pooled OR, 1.56; 95% CI, 0.99 to 2.44) (Figure 58).59, 117-120, 122, 123, 247
We did
not find statistical heterogeneity.
We excluded one RCT from this meta-analysis because its longer followup. This study
compared metformin (titrated to a maximum of 2000 mg daily) with metformin plus
rosiglitazone (titrated to a maximum of 8 mg/2000 mg daily) at 80 weeks and reported 10 total
hypoglycemia events in the metformin-alone arm (3% of patients), and 20 hypoglycemia events
in the metformin-rosiglitazone arm (6% of patients).127
These results were consistent with our
findings reported above. Another study was excluded from the meta-analysis because of its low-
dose combination arm. The study compared metformin (fixed at 1700 mg daily) with a lower-
dose combination arm [metformin (fixed at 500 mg daily) and rosiglitazone (fixed at 4 mg
daily)] and reported no hypoglycemia in either arm at 24 weeks.124
(SOE: High, Metformin
favored for mild, moderate, or total hypoglycemia)](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-236-2048.jpg)
![187
Table 65. Randomized controlled trials comparing metformin with a combination of metformin
plus a GLP-1 receptor agonist on hypoglycemia
Author, Year Followup
(Weeks)
Metformin
(Dose*)
Metformin +
GLP-1 Receptor
Agonist (Dose*)
Definition of
Hypoglycemia
Results
†
(Metformin Vs
Metformin + GLP-1
Receptor Agonist)
Ahren, 2014
141
104 Metformin
(fixed at ≥
1500 mg)
Metformin (fixed
at ≥ 1500 mg) +
albiglutide (max
50 mg weekly)
Mild-moderate
(Asymptomatic, but PG
≤ 3.9 mmol/L)
Mild-moderate
(Symptomatic and PG ≤
3.9 mmol/L)
Severe (required third
party assistance)
1/101 (1%) vs 4/302
(1.3%)
4/101 (4.0%) vs 9/302
(3.0%)
0/101 (0%) vs 0/302 (0%)
Nauck, 2014
159
26 Metformin
(fixed at ≥
1500 mg)
Metformin (fixed
at ≥ 1500 mg) +
dulaglutide (fixed
at 0.75
mg/week)
Mild-moderate (Signs
and symptoms and/or
PG ≤ 70 mg/dL [3.9
mmol/L])
Severe (required third
party assistance)
2/177 (1.1%)
‡
vs 16/302
(5.3%)
0/177 (0%)
‡
vs 0/302
(0%)
Nauck, 2014
159
26 Metformin
(fixed at ≥
1500 mg)
Metformin (fixed
at ≥ 1500 mg) +
dulaglutide (fixed
at 1.5 mg/week)
Mild-moderate (Signs
and symptoms and/or
PG ≤ 70 mg/dL [3.9
mmol/L])
Severe (required third
party assistance)
2/177 (1.1%)
‡
vs 31/304
(10.2%)
0/177 (0%)
‡
vs 0/304
(0%)
Derosa,
2013
171
48 Metformin
(mean
2500 mg)
Metformin (mean
2500 mg) +
exenatide (max
20 mcg)
Total (FPG < 60 mg/dL) 0/85 (0%) vs 0/86 (0%)
DeFronzo,
2005
174
30 Metformin
(fixed at ≥
1500 mg)
Metformin (fixed
at ≥ 1500 mg) +
exenatide 10
mcg
Symptoms (with or
without PG <60 mg/dl
[3.3 mmol/l])
6/113 (5.3%) vs. 5/110
(4.5%)
DeFronzo,
2005
174
30 Metformin
(fixed at ≥
1500 mg)
Metformin (fixed
at ≥ 1500 mg) +
exenatide 20
mcg
Symptoms (with or
without PG <3.3 mmol/l)
6/113 (5.3%) vs. 6/110
(5.3%)
FPG = fasting plasma glucose; GLP-1 = glucagon-like peptide-1; mcg = micrograms; mg = milligrams; mg/dL = milligrams per
deciliter; mmol/L = millimole per liter; PG = plasma glucose; vs = versus
* All doses were titrated, unless otherwise stated.
† Results are presented as n/N (%) unless otherwise stated.
‡ Although study had 52 weeks of followup, incidence of hypoglycemia was only available at 26 weeks. In the trial, patients
were switched from a combination of metformin and placebo to a combination of metformin and sitagliptin.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-244-2048.jpg)
![196
Table 68. Randomized controlled trials comparing a combination of metformin plus a DPP-4
inhibitor with a combination of metformin plus a basal insulin on hypoglycemia
Author, Year Followup
(Weeks)
Metformin +
DPP-4
Inhibitor
(Dose*)
Metformin +
Basal Insulin
(Dose*)
Definition of
Hypoglycemia
Results
†
(Metformin +
DPP-4 Inhibitor Vs
Metformin + Basal
Insulin)
Aschner,
2012
211
24 Metformin
(baseline
dose 1835
mg) +
sitagliptin
(fixed at 100
mg)
Metformin
(baseline
dose 1852
mg) + insulin
glargine (max
0.5 U/kg)
Severe (severe
symptomatic)
Total (symptomatic and
BG ≤ 3.9 mmol/L)
1/264 (0.4%) vs 3/237
(1.3%)
28/264 (10.6%) vs 86/237
(36.3%)
BG = blood glucose; DPP-4 = dipeptidyl peptidase-4; max = maximum; mg = milligrams; mmol/L = millimole per liter; U/kg =
units per kilogram
* All doses were titrated, unless otherwise stated.
† Results are presented as n/N (%) unless otherwise stated.
Combination of Metformin Plus a GLP-1 Receptor Agonist Versus a
Combination of Metformin Plus a Basal Insulin
Two RCTs compared metformin plus basal insulin with the combination of metformin plus
exenatide, with lower risk of mild or moderate hypoglycemia in the metformin plus exenatide
arms in both studies (range in RD of -3% to -25%) (Table 69).212, 264
Both RCTs reported no
between-group differences in severe hypoglycemia.212, 264
(SOE: Low; Combination of
metformin plus a GLP-1 receptor agonist favored for mild, moderate, or total hypoglycemia)
(SOE: Low; Neither favored for severe hypoglycemia)
Table 69. Randomized controlled trials comparing a combination of metformin plus a GLP-1
receptor agonist with a combination of metformin plus a basal insulin on hypoglycemia
Author,
Year
Followup
(Weeks)
Metformin +
GLP-1
(Dose*)
Metformin + Basal
Insulin (Dose*)
Definition of
Hypoglycemia
Results
†
(Metformin +
GLP-1 Vs Metformin +
Basal Insulin)
Davies,
2013
264
26 Metformin
(fixed at ≥
1000 mg) +
exenatide
(fixed at 2 mg
weekly)
Metformin (fixed at ≥
1000 mg) + insulin
detemir (mean initial
dose 0.21 IU/kg;
mean end dose 20.8
IU, end dose 0.51
IU/kg)
Mild-moderate
(symptoms that were
self-treated or resolved
on their own, with
documented BG < 3.0
mmol/L)
Severe hypoglycemia
‡
0/33 (0%) vs 1/29
(3.4%)
0/33 (0%) vs 0/29 (0%)
Diamant,
2010
212
84 Metformin
(continued
stable dose)
+ exenatide
(fixed at 2 mg
weekly)
Metformin (continued
stable dose) + insulin
glargine (started at
10 IU then titrate to
glycemic goal of 4-
5.5 mmol/L)
Mild-moderate
(symptoms and BG < 3.0
mmol/L and was either
self-treated or resolved
independently)
Severe hypoglycemia‡
13/164 (7.9%) vs
51/157 (32.5%)
1/164 (0.6%) vs 1/157
(0.6%)
BG = blood glucose; GLP-1 = glucagon-like peptide-1 receptor agonist; IU = international units; IU/kg = international units per
kilogram; mg = milligrams; mmol/L = millimole per liter;
* All doses were titrated, unless otherwise stated.
† Results are presented as n/N (%) unless otherwise stated.
‡ Any hypoglycemic episode with symptoms consistent with hypoglycemia that led to loss of consciousness or seizure, with
prompt recovery in response to glucagon or glucose administration, or documented hypoglycemia [blood glucose <3.0mmol]
necessitating assistance of another person](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-253-2048.jpg)
![231
Evidence for Cancer
Monotherapy Comparisons
Metformin Versus Thiazolidinediones
A single retrospective cohort study, from the England Cancer Registry, reported no
difference in non-melanoma cancer risk for thiazolidinediones users (N=31,372) versus
metformin users (N=109,708) (adjusted RR, 0.96; 95% CI, 0.81 to 1.13, P not reported) over 4
years of followup.255
(Not graded)
Metformin Versus Sulfonylureas
Four retrospective, cohort studies compared cancer outcomes for metformin and sulfonylurea
users (Table 79).225, 254, 255, 266
Three studies reported no difference between metformin and
sulfonylurea users.225, 255, 266
The other study only provided results stratified by statin use,
indicating a possible interaction with statin use.254
(SOE: Low; Neither favored for long-term
cancer risk)
Table 79. Retrospective cohort studies comparing metformin with sulfonylureas on cancer
Author, Year Population Followup Outcome Results
van Staa, 2012
255
England Cancer
Registry
[n=68,209
(sulfonylurea);
n=109,708 (metformin)]
4-5 years Non-melanoma
cancer
HR 1.03; 95% CI, 0.91 to 1.17
Reference=metformin
Andersson,
2010
225
Danish Patient Registry
Patients with heart
failure
(N=5,852)
10 years Death from
cancer
HR 1.01; 95% CI, 0.72 to 1.43
Reference=sulfonylurea
Lehman, 2012
254
Veterans Health
Administration
[n=533 (metformin-
statin); n=2404
(sulfonylurea-statin);
n=175 (metformin-no
statin); n=1,930
(sulfonylurea-no statin)]
270.4 weeks Incident prostate
cancer
Statin users HR 0.69; 95% CI,
0.5 to 0.92, P = 0.01
Reference=sulfonylurea
Non users of statins HR 2.15;
95% CI, 1.83 to 2.52, P <
0.0001
Reference=sulfonylurea
Kowall, 2015
266
German Disease
Analyzer (IMS Health) –
primary care clinics
(N=22,556)
4.8 years Incident cancer
by ICD-10 code
Adjusted HR, 1.09; 95% CI,
0.87 to 1.36
Reference=metformin
CI = confidence interval; HR = hazard ratio; ICD-10 = International Classification of Diseases
Metformin Versus DPP-4 Inhibitors
Two RCTs of metformin plus placebo (N=510) versus a DPP-4 inhibitor plus placebo
(N=514) evaluated cancer outcomes.85, 87
Each study reported one cancer event: one death due to
pancreatic neoplasm/sepsis87
and one occurrence of esophageal cancer,85
in the metformin arms.
Neither study reported on cancer in the DPP-4 inhibitor arm.85, 87
Followup ranged from 7687
to
104 weeks.85
(SOE: Insufficient)](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-288-2048.jpg)
![272
Evidence for Urinary Tract Infections
Monotherapy Comparisons
Metformin Versus SGLT-2 Inhibitors
Three short RCTs (published in two articles), 12 to 24 weeks in duration, compared
dapagliflozin with metformin and reported on UTIs.88, 89
Since ORs did not appear to vary by
gender in these short-term studies, we present results for men and women combined. We found
significant statistical heterogeneity using a random effects meta-analysis (pooled OR, 1.54; 95%
CI, 0.56 to 4.22; I-squared, 61.1%). Exclusion of any one study did not change the inference of
the meta-analysis. We found similar non-significant increased odds for SGLT-2 inhibitors versus
metformin for UTIs (pooled OR, 1.5; 95% CI, 0.5 to 5.0 (Figure 89) using the profile likelihood
method. One of the RCTs used a lower dose of dapagliflozin of 5 mg88
relative to 10 mg in the
other two RCTs.
We did not include one RCT in the meta-analysis because it was longer (78 weeks) than the
other three studies.90
This study compared metformin to 10 mg of empagliflozin and 25 mg of
empagliflozin and reported similar overall incidences of UTIs with both doses of empagliflozin.
UTI rates among men receiving 25 mg of empagliflozin (4/57; 7.0%) were non-significantly
higher than among those receiving 10 mg of empagliflozin (0/49; 0%) and metformin (0/28; 0%)
and approached UTI rates among women [empagliflozin 10 mg: 4/57 (7.0%), empagliflozin 25
mg: 3/52 (5.8%), and metformin: 2/28 (7.1%)].90
(SOE: Low; Neither favored)
Figure 89. Pooled odds ratio of urinary tract infections comparing metformin with SGLT-2
inhibitors
CI = confidence interval; Group 1 = metformin; Group 2 = sodium-glucose co-transporter-2 inhibitors; OR = odds ratio; SGLT-2
= sodium-glucose co-transporter-2
Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing
more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The
diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-329-2048.jpg)
![284
Evidence for Genital Mycotic Infections
Monotherapy Comparisons
Metformin Versus SGLT-2 Inhibitors
Three medium- to high-quality, short RCTs (reported in two articles) compared metformin
with SGLT-2 inhibitors and found more genital infections in the SGLT-2 inhibitor vs. metformin
arms (pooled OR, 4.1; 95% CI, 2.0 to 8.3) (Figure 91).88, 89
ORs did appear to vary by gender.
No single study markedly influenced the results, and we did not find significant statistical
heterogeneity (I2
= 0.0%).
Figure 91. Pooled odds ratio of genital or mycotic infections comparing metformin with SGLT-2
inhibitors
CI = confidence interval; Group 1 = metformin; Group 2 = sodium-glucose co-transporter-2 inhibitors; OR = odds ratio; SGLT-2
= sodium-glucose co-transporter-2
Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing
more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The
diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
We did not include a low-quality, 78-week RCT in the meta-analysis because of its longer
duration.90
This study compared metformin with empagliflozin and reported slightly higher rates
of genital infections among females for SGLT-2 inhibitor therapy [1/28 (3.6%) with metformin
versus 3/57 (5.3%) with metformin plus low-dose empagliflozin and 3/52 (5.8%) with metformin
plus high-dose empagliflozin] and more genital infections among males with SGLT-2 inhibitors
[0/28 (0%) with metformin versus 2/49 (4.1%) with metformin plus low dose empagliflozin and
3/57 (5.3%) with metformin plus high- dose empagliflozin]. (SOE: Moderate; Metformin
favored)
DPP-4 Inhibitors Versus SGLT-2 Inhibitors
Two RCTs (24 to 26 weeks) compared outcomes from use of 100 mg of sitagliptin daily to
an SGLT-2 inhibitor, by gender.114, 240
Both trials reported higher rates of genital infections
among both women and men with use of SGLT-2 inhibitors compared with sitagliptin, with](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-341-2048.jpg)
![292
Evidence for Fracture
Metformin Versus Metformin-Based Combination Comparisons
Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor
Three RCTs compared metformin with metformin plus an SGLT-2 inhibitor and found no
differences in fractures.168, 170, 267
Two of these had followup for longer than one year: One 102-
week RCT compared metformin with metformin plus dapagliflozin and reported a slightly higher
incidence of fractures in the highest-dose dapagliflozin arm [2/137 (1.5%) for metformin versus
2/137 (1.5%) for dapagliflozin 2.5 mg, 2/137 (1.5%) for dapagliflozin 5 mg, and 3/135 (2.2%)
for dapagliflozin 10 mg]. There was a high loss to followup in this study, ranging from 30
percent to 47 percent across arms.170
Another 102-week RCT compared metformin with
metformin plus 10 mg of dapagliflozin and reported one fracture (1.1%) in each treatment arm
(n=91 for both arms).267
A single 16-week RCT compared metformin with metformin plus
dapagliflozin and reported no fractures in either arm.168
(SOE: Low; Neither favored for shorter
studies; SOE: Low; Neither favored for longer studies)
Metformin-Based Combination Comparisons
Combination of Metformin Plus a Sulfonylurea Versus a Combination of
Metformin Plus an SGLT-2 Inhibitor
Two studies compared metformin plus a GLP-1 receptor agonist with metformin plus a
SGLT-2 inhibitor, showing no differences in fracture risk.200, 219
One 104-week RCT compared
metformin plus glipizide with metformin plus dapagliflozin and showed a slightly higher
incidence of fractures in the metformin plus glipizide arm [9/408 (2.2%) for metformin plus
glipizide versus 6/406 (1.5%) for metformin plus dapagliflozin].219
Another 104-week RCT
compared metformin plus glimepiride to metformin plus empagliflozin showing similar
incidences of fractures in both arms (2%).200
(SOE: Insufficient)
Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of
Metformin Plus an SGLT-2 Inhibitor
One 24-week RCT compared metformin plus a DPP-4 inhibitor with metformin plus a
SGLT-2 inhibitor and reported a slightly higher risk of fracture in the metformin arm [2/176
(1.0%) for metformin plus saxagliptin versus 1/179 (0.6%) for metformin plus dapagliflozin].209
(SOE: Low; Metformin plus SGLT-2 inhibitor favored short-term fracture risk)
Strength of Evidence for Fracture
The strength of evidence for the comparative effects of monotherapy and metformin-based
combinations are presented in Table 122 and summarized in the key points. All studies were
RCTs. Study limitations for all the comparisons were low. We did not find any evidence of
publication bias in any of the comparisons for fractures. A single 52-week, unpublished trial
reported no fractures in either arm (NCT01368081), which is consistent with the other studies.
We also did not find any evidence of publication bias or reporting bias in the grey literature
review.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-349-2048.jpg)
![294
Evidence for Volume Depletion
Monotherapy Comparisons
Metformin Versus SGLT-2 Inhibitors
Two moderately-sized, short RCTs compared metformin with SGLT-2 inhibitors and
reported inconsistent results.88, 89
One 12-week RCT comparing metformin with dapagliflozin
reported a higher incidence of hypotensive events with metformin (4% for metformin versus 0%
for dapagliflozin 5 mg and 10 mg).89
One 24-week RCT comparing metformin with
dapagliflozin 5 mg reported significantly more events of hypotension or syncope with
dapagliflozin [0/201 (0%) for metformin versus 4/203 (2%) for dapagliflozin 5 mg].88
(SOE:
Low; Conflicting results)
Metformin Versus Metformin-Based Combination Comparisons
Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor
Six RCTs compared metformin with the combination of metformin plus an SGLT-2 inhibitor
for this outcome.88, 156, 165, 168, 170, 267
We did not combine these in a meta-analysis because of
differences in study duration and definition of volume depletion events (Table 123, Figure 94).
The two RCTs with long-term followup had large losses to followup and were conflicting with
one suggesting a higher risk of hypotension in the SGLT-2 inhibitor-based arm267
and the other
suggesting similar rates of volume depletion across arms.170
Volume depletion events were rare
in the four short-duration RCTs (Figure 94).88, 156, 165, 168
(SOE: Low; Neither favored for shorter
studies; SOE: Low; Neither Favored for longer studies)
Figure 94. Pooled odds ratio of volume depletion comparing metformin with a combination of
metformin plus an SGLT-2 inhibitor
CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2
inhibitor; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2
Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing
more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The
diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies
were excluded because they did not contribute any events.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-351-2048.jpg)
![297
metformin plus sitagliptin versus 4/64 (6%) for metformin plus canagliflozin 100 mg, 3/65 (5%)
for metformin plus canagliflozin 200 mg, and 1/64 (2%) for metformin plus canagliflozin 300
mg].156
(SOE: Low; Neither favored)
Strength of Evidence for Volume Depletion
The strength of evidence for the comparative effects of monotherapy and metformin-based
combinations are presented in Table 125 and summarized in the key points. All studies were
RCTs. Study limitations for all the comparisons were low. Where quality influences results, we
describe that under the appropriate comparisons. In general, we did not find strong differences in
outcomes in the lower- versus higher-quality studies. We did not find any evidence of
publication bias in any of the comparisons for volume depletion. We also did not find any
evidence of publication bias or reporting bias in the grey literature review.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-354-2048.jpg)
![307
or just increases cardiovascular mortality less than sulfonylureas, and likewise, we do not know
if sulfonylureas actually increase cardiovascular mortality or just decrease cardiovascular
mortality less than metformin.
We evaluated pioglitazone and rosiglitazone separately when they were compared with other
drug classes for all-cause mortality and cardiovascular outcomes, given concerns about
cardiovascular risk and mortality associated with rosiglitazone raised previously292
and acted
upon by the FDA through recommendations regarding the prescription of this medication in
2010.293, 294
While the strength of evidence was low, metformin was favored over the
combination of metformin plus rosiglitazone for short-term all-cause mortality, as well as
cardiovascular mortality and cardiovascular morbidity; metformin was also favored, albeit with
low strength of evidence, over rosiglitazone monotherapy for long-term mortality and long-term
cardiovascular morbidity.
We had little evidence on the combination of metformin plus a sulfonylurea or metformin
plus a thiazolidinedione. In The Rosiglitazone Evaluated for Cardiovascular Outcomes in oral
agent combination therapy for type 2 Diabetes (RECORD) trial, 4,447 participants with diabetes
were randomized to a rosiglitazone-based combination (with either metformin or sulfonylurea)
or to the combination of metformin plus a sulfonylurea.49
We excluded this study from this
outcome assessment, because it did not stratify results by a comparison of interest. This trial did
not find a difference between rosiglitazone-based two-drug therapy and metformin plus a
sulfonylurea for the primary endpoint, cardiovascular hospitalization or death. After 5.5 years,
all-cause mortality, cardiovascular death, and stroke were slightly higher in the metformin plus
sulfonylurea arm, and myocardial infarction rates were slightly (non-statistically significantly)
higher in the rosiglitazone combination therapy arm. Of note, the FDA commissioned an
independent re-adjudication and analysis of the data from RECORD and subsequently (in 2013)
lifted their restrictions on rosiglitazone usage.295
Although strength of evidence was low, compared with the combination of metformin plus a
sulfonylurea, the combination of metformin plus a DPP-4 inhibitor was associated with lower
rates of long-term all-cause mortality and cardiovascular mortality and morbidity; an
unpublished study with long-term followup was supportive of these conclusions. We also
identified many new studies evaluating short-term all-cause mortality and cardiovascular
outcomes for DPP-4 inhibitor comparisons, but most of this evidence was of low strength.
Several meta-analyses published since the 2011 report have evaluated DPP-4 inhibitors and
all-cause mortality and cardiovascular outcomes; they tended to combine comparators (active,
placebo, or both) against DPP-4 inhibitors so were not conclusive about specific direct
comparisons.296-301
The conclusions of these reviews have been based mainly on trials with less
than 2 years of followup and have reported mixed results on short-term mortality and
cardiovascular risk.
Outside of these meta-analyses, three large RCTs have evaluated DPP-4 inhibitors compared
with placebo, added to the standard diabetes treatment per routine clinical care, and
cardiovascular outcomes.302-304
None of these RCTs evaluated direct head-to-head comparisons
of interest, and none were included in this review. All of these trials designated a composite
cardiovascular outcome as the primary outcome and reported non-inferiority (but not superiority)
for the DPP-4 inhibitor addition compared with placebo addition for the composite
cardiovascular outcome.302-304
Rates of the primary endpoint were lower in the DPP-4 inhibitor
versus placebo arm in two of the three trials [risk difference (statistical comparison not
provided): -0.5%,303
-0.2%,302
0.1%302
]. Participants in all three trials were treated per clinical](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-364-2048.jpg)
![313
safety of this drug class for this outcome in mainly short duration studies [five short duration
RCTs reporting no events in the DPP-4 inhibitor arms, one short duration RCT with one event in
the metformin plus DPP-4 arm and none in the comparator arm, and one RCT which reported
few congestive heart failure events in the metformin plus DPP-4 inhibitor arm compared with the
metformin plus sulfonylurea arm (three versus six respectively)]. Several large, double-blind,
placebo-controlled RCTs evaluating DPP-4 inhibitors on cardiovascular outcomes in adults with
moderate to high cardiovascular risk were excluded from our systematic review of head-to-head
comparisons but deserve mention due to recent controversy.302, 303, 324
Two of these RCTs
(comparing either saxagliptin or alogliptin with placebo) reported a small increased risk of
hospitalization for congestive heart failure in adults at moderate to high cardiovascular risk
(between-group absolute risk differences of 0.7% and 0.9%).302, 324
The EXAMINE trial with the
alogliptin comparison reported these differences solely for the outcome of first hospitalization
for heart failure in adults without pre-existing congestive heart failure as part of a posthoc
subgroup analysis.324
The third placebo-controlled RCT303
compared sitagliptin with placebo on
cardiovascular outcomes and reported no between-group differences in hospitalization for
congestive heart failure. It is unclear if differences in these trials were due to differences in drug
type, chance, or other causes. Due to these findings, however, the FDA has requested additional
labeling for saxagliptin and alogliptin to reflect concerns of the potential increased risk of
hospitalization for congestive heart failure.325
Further research directly comparing DPP-4
inhibitors with other active comparators on heart failure outcomes will be useful in determining
the comparative safety of these medications on heart failure risk, including results of two RCTs
[the Cardiovascular Outcome Study of Linagliptin Versus Glimepiride in Patients with Type 2
Diabetes (CAROLINA) and the Cardiovascular and Renal Microvascular Outcome Study with
Linagliptin in Patients with Type 2 Diabetes Mellitus (CARMELINA) studies] of linagliptin,
which are in progress.326, 327
Liver Injury
Similar to the 2011 report, we found little evidence on liver injury. Compared to the prior
report showing no between-group differences in liver injury, we downgraded the available
evidence for metformin versus thiazolidinedione monotherapy (from moderate to low) and
downgraded the evidence for thiazolidinedione versus sulfonylurea monotherapy (from high to
low). Notably, there are FDA warnings of post-marketing cases of hepatic failure for both
alogliptin328
and pioglitazone,329
but we found low or insufficient strength of evidence for
thiazolidinedione- and DPP-4 inhibitor-based comparisons and liver injury.
Lactic Acidosis
Prior evidence on the elevated risk of lactic acidosis with phenformin, an earlier biguanide,
and case reports of lactic acidosis among metformin users have led to continued concern about
an increased risk of lactic acidosis with metformin; however, for most of the ~300 case reports
on metformin and lactic acidosis, other factors contributing to lactic acidosis could not be
excluded (e.g., acute myocardial infarction330, 331
or acute kidney failure). Consistent with the
prior report16
and a Cochrane review on this topic,332
we did not find an increased risk of lactic
acidosis with metformin based on the little evidence identified. A more recent systematic review
by Inzucchi et al.333
evaluated the risk of lactic acidosis associated with metformin use in adults
with mild to moderate chronic kidney disease (CKD; estimated glomerular filtration rates of 30-
60 mL/min per 1.73 m2
). Using data from 65 studies (mainly observational), they reported an](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-370-2048.jpg)
![329
52. Tolman KG, Freston JW, Kupfer S, Perez A.
Liver safety in patients with type 2 diabetes
treated with pioglitazone: results from a 3-
year, randomized, comparator-controlled
study in the US. Drug Saf. 2009;32(9):787-
800. PMID: 19670918.
53. Gallwitz B, Guzman J, Dotta F, et al.
Exenatide twice daily versus glimepiride for
prevention of glycaemic deterioration in
patients with type 2 diabetes with metformin
failure (EUREXA): an open-label,
randomised controlled trial. Lancet. 2012
Jun 16;379(9833):2270-8. PMID: 22683137.
54. Del Prato S, Nauck M, Duran-Garcia S, et
al. Long-term glycaemic response and
tolerability of dapagliflozin versus a
sulphonylurea as add-on therapy to
metformin in type 2 diabetes patients: 4-year
data. Diabetes Obes Metab. 2015 Mar 4.
PMID: 25735400.
55. Derosa G, Maffioli P, Salvadeo SA, et al.
Direct comparison among oral
hypoglycemic agents and their association
with insulin resistance evaluated by
euglycemic hyperinsulinemic clamp: the
60's study. Metabolism. 2009 2009 Apr 22.
PMID: 19394976
56. Gupta AK, Smith SR, Greenway FL, Bray
GA. Pioglitazone treatment in type 2
diabetes mellitus when combined with
portion control diet modifies the metabolic
syndrome. Diabetes Obes Metab. 2009 2009
Apr;11(4):330-7. PMID: 19267711
57. Erdem G, Dogru T, Tasci I, et al. The effects
of pioglitazone and metformin on plasma
visfatin levels in patients with treatment
naive type 2 diabetes mellitus. Diabetes Res
Clin Pract. 2008 2008 Nov;82(2):214-8.
PMID: 18778865
58. Iliadis F, Kadoglou NP, Hatzitolios A, et al.
Metabolic effects of rosiglitazone and
metformin in Greek patients with recently
diagnosed type 2 diabetes. In Vivo. 2007
2007 Nov-Dec;21(6):1107-14. PMID:
18210765
59. Rosenstock J, Rood J, Cobitz A, et al. Initial
treatment with rosiglitazone/metformin
fixed-dose combination therapy compared
with monotherapy with either rosiglitazone
or metformin in patients with uncontrolled
type 2 diabetes. Diabetes Obes Metab. 2006
2006 Nov;8(6):650-60. PMID: 17026489
60. Yamanouchi T, Sakai T, Igarashi K, et al.
Comparison of metabolic effects of
pioglitazone, metformin, and glimepiride
over 1 year in Japanese patients with newly
diagnosed Type 2 diabetes. Diabet Med.
2005 2005 Aug;22(8):980-5. PMID:
16026361
61. Ramachandran A, Snehalatha C, Salini J,
Vijay V. Use of glimepiride and insulin
sensitizers in the treatment of type 2
diabetes--a study in Indians. J Assoc
Physicians India. 2004 2004 Jun;52:459-63.
PMID: 15645955
62. Schernthaner G, Matthews DR, Charbonnel
B, et al. Efficacy and safety of pioglitazone
versus metformin in patients with type 2
diabetes mellitus: a double-blind,
randomized trial. J Clin Endocrinol Metab.
2004 2004 Dec;89(12):6068-76. PMID:
15579760
63. Lawrence JM, Reid J, Taylor GJ, et al.
Favorable effects of pioglitazone and
metformin compared with gliclazide on
lipoprotein subfractions in overweight
patients with early type 2 diabetes. Diabetes
Care. 2004 2004 Jan;27(1):41-6. PMID:
14693964
64. Pavo I, Jermendy G, Varkonyi TT, et al.
Effect of pioglitazone compared with
metformin on glycemic control and
indicators of insulin sensitivity in recently
diagnosed patients with type 2 diabetes. J
Clin Endocrinol Metab. 2003 2003
Apr;88(4):1637-45. PMID: 12679450
65. Hallsten K, Virtanen KA, Lonnqvist F, et al.
Rosiglitazone but not metformin enhances
insulin- and exercise-stimulated skeletal
muscle glucose uptake in patients with
newly diagnosed type 2 diabetes. Diabetes.
2002 2002 Dec;51(12):3479-85. PMID:
12453903
66. Kiyici S, Ersoy C, Kaderli A, et al. Effect of
rosiglitazone, metformin and medical
nutrition treatment on arterial stiffness,
serum MMP-9 and MCP-1 levels in drug
naive type 2 diabetic patients. Diabetes
Research and Clinical Practice. [doi: DOI:
10.1016/j.diabres.2009.07.004]. 2009
2009/10//;86(1):44-50. PMID: 19674806](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-386-2048.jpg)
![330
67. Perez A, Zhao Z, Jacks R, Spanheimer R.
Efficacy and safety of
pioglitazone/metformin fixed-dose
combination therapy compared with
pioglitazone and metformin monotherapy in
treating patients with T2DM. Curr Med Res
Opin. 2009 Dec;25(12):2915-23. PMID:
19827910.
68. Kato T, Sawai Y, Kanayama H, et al.
Comparative study of low-dose pioglitazone
or metformin treatment in Japanese diabetic
patients with metabolic syndrome. Exp Clin
Endocrinol Diabetes. 2009
Nov;117(10):593-9. PMID: 19924605.
69. Esteghamati A, Ghasemiesfe M,
Mousavizadeh M, et al. Pioglitazone and
metformin are equally effective in reduction
of chemerin in patients with type 2 diabetes.
J Diabetes Investig. 2014 May 4;5(3):327-
32. PMID: 24843782.
70. Erem C, Ozbas HM, Nuhoglu I, et al.
Comparison of effects of gliclazide,
metformin and pioglitazone monotherapies
on glycemic control and cardiovascular risk
factors in patients with newly diagnosed
uncontrolled type 2 diabetes mellitus. Exp
Clin Endocrinol Diabetes. 2014
May;122(5):295-302. PMID: 24710641.
71. Genovese S, De Berardis G, Nicolucci A, et
al. Effect of pioglitazone versus metformin
on cardiovascular risk markers in type 2
diabetes. Adv Ther. 2013 Feb;30(2):190-
202. PMID: 23359066.
72. Taslimi S, Esteghamati A, Rashidi A, et al.
Treatment with pioglitazone is associated
with decreased preprandial ghrelin levels: a
randomized clinical trial. Peptides. 2013
Feb;40:89-92. PMID: 23276779.
73. Russell-Jones D, Cuddihy RM, Hanefeld M,
et al. Efficacy and safety of exenatide once
weekly versus metformin, pioglitazone, and
sitagliptin used as monotherapy in drug-
naive patients with type 2 diabetes
(DURATION-4): a 26-week double-blind
study. Diabetes Care. 2012 Feb;35(2):252-8.
PMID: 22210563.
74. Yoon KH, Shin JA, Kwon HS, et al.
Comparison of the efficacy of glimepiride,
metformin, and rosiglitazone monotherapy
in korean drug-naive type 2 diabetic
patients: the practical evidence of
antidiabetic monotherapy study. Diabetes
Metab J. 2011 Feb;35(1):26-33. PMID:
21537410.
75. Fidan E, Onder Ersoz H, Yilmaz M, et al.
The effects of rosiglitazone and metformin
on inflammation and endothelial dysfunction
in patients with type 2 diabetes mellitus.
Acta Diabetol. 2011 Dec;48(4):297-302.
PMID: 21424914.
76. Esposito K, Maiorino MI, Di Palo C, et al.
Effects of pioglitazone versus metformin on
circulating endothelial microparticles and
progenitor cells in patients with newly
diagnosed type 2 diabetes--a randomized
controlled trial. Diabetes Obes Metab. 2011
May;13(5):439-45. PMID: 21255215.
77. Esteghamati A, Azizi R, Ebadi M, et al. The
Comparative Effect of Pioglitazone and
Metformin on Serum Osteoprotegerin,
Adiponectin and Intercellular Adhesion
Molecule Concentrations in Patients with
Newly Diagnosed Type 2 Diabetes: a
Randomized Clinical Trial. Exp Clin
Endocrinol Diabetes. 2015 Jan 21. PMID:
25607338.
78. Turkmen Kemal Y, Guvener Demirag N,
Yildirir A, et al. Effects of rosiglitazone on
plasma brain natriuretic peptide levels and
myocardial performance index in patients
with type 2 diabetes mellitus. Acta Diabetol.
2007 2007 Sep;44(3):149-56. PMID:
17721754
79. Gupta RK, Rehan HS, Rohatagi A, et al. The
effect of glipizide, metformin and
rosiglitazone on nontraditional
cardiovascular risk factors in newly
diagnosed patients with type 2 diabetes
mellitus. International journal of diabetes in
developing countries [serial on the Internet].
2010; (3): Available from:
http://onlinelibrary.wiley.com/o/cochrane/clcentr
al/articles/971/CN-00797971/frame.html.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-387-2048.jpg)
![333
110. Kaku K, Rasmussen MF, Nishida T, Seino
Y. Fifty-two-week, randomized, multicenter
trial to compare the safety and efficacy of
the novel glucagon-like peptide-1 analog
liraglutide vs glibenclamide in patients with
type 2 diabetes. J Diabetes Investig. 2011
Nov 30;2(6):441-7. PMID: 24843528.
111. Madsbad S, Schmitz O, Ranstam J, et al.
Improved glycemic control with no weight
increase in patients with type 2 diabetes
after once-daily treatment with the long-
acting glucagon-like peptide 1 analog
liraglutide (NN2211): a 12-week, double-
blind, randomized, controlled trial. Diabetes
Care. 2004 2004 Jun;27(6):1335-42. PMID:
15161785
112. Garber A, Henry R, Ratner R, et al.
Liraglutide versus glimepiride monotherapy
for type 2 diabetes (LEAD-3 Mono): a
randomised, 52-week, phase III, double-
blind, parallel-treatment trial. Lancet. 2009
2009 Feb 7;373(9662):473-81. PMID:
18819705
113. Garber A, Henry RR, Ratner R, et al.
Liraglutide, a once-daily human glucagon-
like peptide 1 analogue, provides sustained
improvements in glycaemic control and
weight for 2 years as monotherapy
compared with glimepiride in patients with
type 2 diabetes. Diabetes Obes Metab. 2011
Apr;13(4):348-56. PMID: 21205128.
114. Roden M, Weng J, Eilbracht J, et al.
Empagliflozin monotherapy with sitagliptin
as an active comparator in patients with type
2 diabetes: A randomised, double-blind,
placebo-controlled, phase 3 trial. The Lancet
Diabetes and Endocrinology [serial on the
Internet]. 2013; (3): Available from:
http://onlinelibrary.wiley.com/o/cochrane/clcentr
al/articles/599/CN-00915599/frame.html.
115. Suzuki K, Tanaka S, Aoki C, et al. Greater
efficacy and improved endothelial
dysfunction in untreated type 2 diabetes with
liraglutide versus sitagliptin. Dokkyo
Journal of Medical Sciences.
2014;41(3):211-20. PMID: COULD NOT
FIND THIS ONE
116. Leiter LA, Harris SB, Chiasson J-L, et al.
Efficacy and safety of Rosiglitazone as
monotherapy or in combination with
metformin in primary care settings. Can J
Diabetes. 2005 2005;29(4):384-92. PMID:
COULD NOT FIND THIS ONE
117. Kaku K. Efficacy and safety of therapy with
metformin plus pioglitazone in the treatment
of patients with type 2 diabetes: a double-
blind, placebo-controlled, clinical trial. Curr
Med Res Opin. 2009 2009 May;25(5):1111-
9. PMID: 19309251
118. Scott R, Loeys T, Davies MJ, Engel SS.
Efficacy and safety of sitagliptin when
added to ongoing metformin therapy in
patients with type 2 diabetes. Diabetes Obes
Metab. 2008 2008 Sep;10(10):959-69.
PMID: 18201203
119. Weissman P, Goldstein BJ, Rosenstock J, et
al. Effects of rosiglitazone added to
submaximal doses of metformin compared
with dose escalation of metformin in type 2
diabetes: the EMPIRE Study. Curr Med Res
Opin. 2005 2005 Dec;21(12):2029-35.
PMID: 16368054
120. Bailey CJ, Bagdonas A, Rubes J, et al.
Rosiglitazone/metformin fixed-dose
combination compared with uptitrated
metformin alone in type 2 diabetes mellitus:
a 24-week, multicenter, randomized, double-
blind, parallel-group study. Clin Ther. 2005
2005 Oct;27(10):1548-61. PMID: 16330291
121. Gomez-Perez FJ, Fanghanel-Salmon G,
Antonio Barbosa J, et al. Efficacy and safety
of rosiglitazone plus metformin in Mexicans
with type 2 diabetes. Diabetes Metab Res
Rev. 2002 2002 Mar-Apr;18(2):127-34.
PMID: 11994904
122. Einhorn D, Rendell M, Rosenzweig J, et al.
Pioglitazone hydrochloride in combination
with metformin in the treatment of type 2
diabetes mellitus: a randomized, placebo-
controlled study. The Pioglitazone 027
Study Group. Clin Ther. 2000 2000
Dec;22(12):1395-409. PMID: 11192132
123. Fonseca V, Rosenstock J, Patwardhan R,
Salzman A. Effect of metformin and
rosiglitazone combination therapy in
patients with type 2 diabetes mellitus: a
randomized controlled trial. JAMA. 2000
2000 Apr 5;283(13):1695-702. PMID:
10755495](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-390-2048.jpg)
![335
139. Forst T, Uhlig-Laske B, Ring A, et al.
Linagliptin (BI 1356), a potent and selective
DPP-4 inhibitor, is safe and efficacious in
combination with metformin in patients with
inadequately controlled Type 2 diabetes.
Diabet Med. 2010 Dec;27(12):1409-19.
PMID: 21059094.
140. Kim HS, Kim DM, Cha BS, et al. Efficacy
of glimepiride/metformin fixed-dose
combination vs metformin uptitration in
type 2 diabetic patients inadequately
controlled on low-dose metformin
monotherapy: A randomized, open label,
parallel group, multicenter study in Korea.
Journal of Diabetes Investigation.
2014((Kim H.-S.) Department of Internal
Medicine Keimyung University School of
Medicine Daegu Korea). PMID: 2542271
141. Ahren B, Johnson SL, Stewart M, et al.
HARMONY 3: 104-Week Randomized,
Double-Blind, Placebo- and Active-
Controlled Trial Assessing the Efficacy and
Safety of Albiglutide Compared With
Placebo, Sitagliptin, and Glimepiride in
Patients With Type 2 Diabetes Taking
Metformin. Diabetes Care. 2014 Jun 4.
PMID: 24898304.
142. Raz I, Chen Y, Wu M, et al. Efficacy and
safety of sitagliptin added to ongoing
metformin therapy in patients with type 2
diabetes. Curr Med Res Opin. 2008 2008
Feb;24(2):537-50. PMID: 18194595
143. Charbonnel B, Karasik A, Liu J, et al.
Efficacy and safety of the dipeptidyl
peptidase-4 inhibitor sitagliptin added to
ongoing metformin therapy in patients with
type 2 diabetes inadequately controlled with
metformin alone. Diabetes Care. 2006 2006
Dec;29(12):2638-43. PMID: 17130197
144. DeFronzo RA, Hissa MN, Garber AJ, et al.
The efficacy and safety of saxagliptin when
added to metformin therapy in patients with
inadequately controlled type 2 diabetes with
metformin alone. Diabetes Care. 2009
Sep;32(9):1649-55. PMID: 19478198.
145. Reasner C, Olansky L, Seck TL, et al. The
effect of initial therapy with the fixed-dose
combination of sitagliptin and metformin
compared with metformin monotherapy in
patients with type 2 diabetes mellitus.
Diabetes Obes Metab. 2011 Jul;13(7):644-
52. PMID: 21410627.
146. Yang W, Pan CY, Tou C, et al. Efficacy and
safety of saxagliptin added to metformin in
Asian people with type 2 diabetes mellitus: a
randomized controlled trial. Diabetes Res
Clin Pract. 2011 Nov;94(2):217-24. PMID:
21871686.
147. Fonseca V, Zhu T, Karyekar C, Hirshberg
B. Adding saxagliptin to extended-release
metformin vs. uptitrating metformin dosage.
Diabetes Obes Metab. 2012 Apr;14(4):365-
71. PMID: 22192246.
148. Yang W, Guan Y, Shentu Y, et al. The
addition of sitagliptin to ongoing metformin
therapy significantly improves glycemic
control in Chinese patients with type 2
diabetes. J Diabetes. 2012 Sep;4(3):227-37.
PMID: 22672586.
149. Kadowaki T, Tajima N, Odawara M, et al.
Addition of sitagliptin to ongoing metformin
monotherapy improves glycemic control in
Japanese patients with type 2 diabetes over
52 weeks. J Diabetes Investig. 2013 Mar
18;4(2):174-81. PMID: 24843649.
150. Derosa G, Carbone A, D'Angelo A, et al.
Variations in inflammatory biomarkers
following the addition of sitagliptin in
patients with type 2 diabetes not controlled
with metformin. Intern Med.
2013;52(19):2179-87. PMID: 24088749.
151. White JL, Buchanan P, Li J, Frederich R. A
randomized controlled trial of the efficacy
and safety of twice-daily saxagliptin plus
metformin combination therapy in patients
with type 2 diabetes and inadequate
glycemic control on metformin
monotherapy. BMC Endocr Disord.
2014;14(1):17. PMID: 24565221.
152. Ross SA, Rafeiro E, Meinicke T, et al.
Efficacy and safety of linagliptin 2.5?mg
twice daily versus 5?mg once daily in
patients with type 2 diabetes inadequately
controlled on metformin: a randomised,
double-blind, placebo-controlled trial. Curr
Med Res Opin [serial on the Internet]. 2012;
(9): Available from:
http://onlinelibrary.wiley.com/o/cochrane/clcentr
al/articles/674/CN-00859674/frame.html.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-392-2048.jpg)
![336
153. Rosenstock J, Seman LJ, Jelaska A, et al.
Efficacy and safety of empagliflozin, a
sodium glucose cotransporter 2 (SGLT2)
inhibitor, as add-on to metformin in type 2
diabetes with mild hyperglycaemia.
Diabetes Obes Metab [serial on the
Internet]. 2013; (12): Available from:
http://onlinelibrary.wiley.com/o/cochrane/clcentr
al/articles/648/CN-00915648/frame.html.
154. Nauck MA, Ellis GC, Fleck PR, et al.
Efficacy and safety of adding the dipeptidyl
peptidase-4 inhibitor alogliptin to metformin
therapy in patients with type 2 diabetes
inadequately controlled with metformin
monotherapy: a multicentre, randomised,
double-blind, placebo-controlled study. Int J
Clin Pract. 2009 Jan;63(1):46-55. PMID:
19125992.
155. Taskinen MR, Rosenstock J, Tamminen I, et
al. Safety and efficacy of linagliptin as add-
on therapy to metformin in patients with
type 2 diabetes: a randomized, double-blind,
placebo-controlled study. Diabetes Obes
Metab. 2011 Jan;13(1):65-74. PMID:
21114605.
156. Rosenstock J, Aggarwal N, Polidori D, et al.
Dose-ranging effects of canagliflozin, a
sodium-glucose cotransporter 2 inhibitor, as
add-on to metformin in subjects with type 2
diabetes. Diabetes Care. 2012
Jun;35(6):1232-8. PMID: 22492586.
157. Seino Y, Miyata Y, Hiroi S, et al. Efficacy
and safety of alogliptin added to metformin
in Japanese patients with type 2 diabetes: a
randomized, double-blind, placebo-
controlled trial with an open-label, long-
term extension study. Diabetes Obes Metab.
2012 Oct;14(10):927-36. PMID: 22583697.
158. Lavalle-Gonzalez FJ, Januszewicz A,
Davidson J, et al. Efficacy and safety of
canagliflozin compared with placebo and
sitagliptin in patients with type 2 diabetes on
background metformin monotherapy: a
randomised trial. Diabetologia. 2013
Dec;56(12):2582-92. PMID: 24026211.
159. Nauck M, Weinstock RS, Umpierrez GE, et
al. Efficacy and Safety of Dulaglutide
Versus Sitagliptin After 52 Weeks in Type 2
Diabetes in a Randomized Controlled Trial
(AWARD-5). Diabetes Care. 2014 Apr 17.
PMID: 24742660.
160. Wang W, Yang J, Yang G, et al. Efficacy
and safety of linagliptin in Asian patients
with type 2 diabetes mellitus inadequately
controlled by metformin: A multinational
24-week, randomized clinical trial. J
Diabetes. 2015 Mar 6. PMID: 25753488.
161. Hermans MP, Delibasi T, Farmer I, et al.
Effects of saxagliptin added to sub-maximal
doses of metformin compared with
uptitration of metformin in type 2 diabetes:
the PROMPT study. Curr Med Res Opin.
2012 Oct;28(10):1635-45. PMID:
23020253.
162. Ji L, Zinman B, Patel S, et al. Efficacy and
safety of linagliptin co-administered with
low-dose metformin once daily versus high-
dose metformin twice daily in treatment-
naive patients with type 2 diabetes: a
double-blind randomized trial. Adv Ther.
2015 Mar;32(3):201-15. PMID: 25805187.
163. Aaboe K, Knop FK, Vilsboll T, et al.
Twelve weeks treatment with the DPP-4
inhibitor, sitagliptin, prevents degradation of
peptide YY and improves glucose and non-
glucose induced insulin secretion in patients
with type 2 diabetes mellitus. Diabetes Obes
Metab. 2010 Apr;12(4):323-33. PMID:
20380653.
164. Haak T, Meinicke T, Jones R, et al. Initial
combination of linagliptin and metformin in
patients with type 2 diabetes: efficacy and
safety in a randomised, double-blind 1-year
extension study. Int J Clin Pract. 2013
Dec;67(12):1283-93. PMID: 24118640.
165. Qiu R, Capuano G, Meininger G. Efficacy
and safety of twice-daily treatment with
canagliflozin, a sodium glucose co-
transporter 2 inhibitor, added on to
metformin monotherapy in patients with
type 2 diabetes mellitus. Journal of Clinical
and Translational Endocrinology.
2014;1(2):54-60. PMID: COULD NOT
FIND THIS ONE
166. Haring HU, Merker L, Seewaldt-Becker E,
et al. Empaglif lozin as add-on to metformin
in patients with type 2 diabetes: A 24-week,
randomized, double-blind, placebo-
controlled trial. Diabetes Care.
2014;37(6):1650-9. PMID: 24722494](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-393-2048.jpg)
![344
270. Kawai T, Funae O, Shimada A, et al. Effects
of pretreatment with low-dose metformin on
metabolic parameters and weight gain by
pioglitazone in Japanese patients with type 2
diabetes. Intern Med. 2008
2008;47(13):1181-8. PMID: 18591838
271. United Kingdom Prospective Diabetes Study
24: a 6-year, randomized, controlled trial
comparing sulfonylurea, insulin, and
metformin therapy in patients with newly
diagnosed type 2 diabetes that could not be
controlled with diet therapy. United
Kingdom Prospective Diabetes Study
Group. Ann Intern Med. 1998 1998 Feb
1;128(3):165-75. PMID: 9454524
272. Gallwitz B, Rosenstock J, Patel S, et al.
Regardless of the degree of glycaemic
control, linagliptin has lower hypoglycaemia
risk than all doses of glimepiride, at all time
points, over the course of a 2-year trial.
Diabetes Obes Metab. 2015 Mar;17(3):276-
84. PMID: 25425502.
273. Kahn SE, Zinman B, Lachin JM, et al.
Rosiglitazone-associated fractures in type 2
diabetes: an Analysis from A Diabetes
Outcome Progression Trial (ADOPT).
Diabetes Care. 2008 2008 May;31(5):845-
51. PMID: 18223031
274. Dormuth CR, Carney G, Carleton B, et al.
Thiazolidinediones and fractures in men and
women. Arch Intern Med. 2009 2009 Aug
10;169(15):1395-402. PMID: 19667303
275. Schellhase KG, Koepsell TD, Weiss NS.
Glycemic control and the risk of multiple
microvascular diabetic complications. Fam
Med. 2005 Feb;37(2):125-30. PMID:
15690253.
276. Vijan S, Hofer TP, Hayward RA. Estimated
benefits of glycemic control in
microvascular complications in type 2
diabetes. Ann Intern Med. 1997 Nov
1;127(9):788-95. PMID: 9382399.
277. Shyangdan DS, Royle P, Clar C, et al.
Glucagon-like peptide analogues for type 2
diabetes mellitus. Cochrane Database of
Systematic Reviews [serial on the Internet].
2011; (10): Available from:
http://onlinelibrary.wiley.com/doi/10.1002/14651
858.CD006423.pub2/abstract.
278. Liu SC, Tu YK, Chien MN, Chien KL.
Effect of antidiabetic agents added to
metformin on glycaemic control,
hypoglycaemia and weight change in
patients with type 2 diabetes: A network
meta-analysis. Diabetes, obesity &
metabolism. 2012 Sep;14(9):810-20. PMID:
22486990.
279. McIntosh B, Cameron C, Singh SR, et al.
Second-line therapy in patients with type 2
diabetes inadequately controlled with
metformin monotherapy: a systematic
review and mixed-treatment comparison
meta-analysis. Open Med. 2011;5(1):e35-48.
PMID: 22046219.
280. Kahn BB, Flier JS. Obesity and insulin
resistance. J Clin Invest. 2000
Aug;106(4):473-81. PMID: 10953022.
281. Purnell TS, Joy S, Little E, et al. Patient
preferences for noninsulin diabetes
medications: a systematic review. Diabetes
Care. 2014 Jul;37(7):2055-62. PMID:
24963113.
282. Vasilakou D, Karagiannis T, Athanasiadou
E, et al. Sodium-glucose cotransporter 2
inhibitors for type 2 diabetes: a systematic
review and meta-analysis. Ann Intern Med.
2013 Aug 20;159(4):262-74. PMID:
24026259.
283. Richter B, Bandeira-Echtler E, Bergerhoff
K, Lerch C. Dipeptidyl peptidase-4 (DPP-4)
inhibitors for type 2 diabetes mellitus.
Cochrane Database of Systematic Reviews
[serial on the Internet]. 2008; (2): Available
from:
http://onlinelibrary.wiley.com/doi/10.1002/14651
858.CD006739.pub2/abstract.
284. Prevention of stroke by antihypertensive
drug treatment in older persons with isolated
systolic hypertension. Final results of the
Systolic Hypertension in the Elderly
Program (SHEP). SHEP Cooperative
Research Group. JAMA. 1991 Jun
26;265(24):3255-64. PMID: 2046107.
285. Tight blood pressure control and risk of
macrovascular and microvascular
complications in type 2 diabetes: UKPDS
38. UK Prospective Diabetes Study Group.
BMJ. 1998 Sep 12;317(7160):703-13.
PMID: 9732337.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-401-2048.jpg)
![346
302. Scirica BM, Bhatt DL, Braunwald E, et al.
Saxagliptin and cardiovascular outcomes in
patients with type 2 diabetes mellitus. The
New England journal of medicine. 2013 Oct
3;369(14):1317-26. PMID: 23992601.
303. Green JB, Bethel MA, Armstrong PW, et al.
Effect of Sitagliptin on Cardiovascular
Outcomes in Type 2 Diabetes. The New
England journal of medicine. 2015 Jul
16;373(3):232-42. PMID: 26052984.
304. White WB, Cannon CP, Heller SR, et al.
Alogliptin after acute coronary syndrome in
patients with type 2 diabetes. The New
England journal of medicine. 2013 Oct
3;369(14):1327-35. PMID: 23992602.
305. Monami M, Dicembrini I, Nardini C, et al.
Effects of glucagon-like peptide-1 receptor
agonists on cardiovascular risk: A meta-
analysis of randomized clinical trials.
Diabetes, obesity & metabolism.
2014;16(1):38-47. PMID: 23829656
306. Bonds DE, Miller ME, Bergenstal RM, et al.
The association between symptomatic,
severe hypoglycaemia and mortality in type
2 diabetes: retrospective epidemiological
analysis of the ACCORD study. BMJ.
2010;340:b4909. PMID: 20061358.
307. Holman RR, Farmer AJ, Davies MJ, et al.
Three-year efficacy of complex insulin
regimens in type 2 diabetes. The New
England journal of medicine. 2009 Oct
29;361(18):1736-47. PMID: 19850703.
308. Budnitz DS, Shehab N, Kegler SR, Richards
CL. Medication use leading to emergency
department visits for adverse drug events in
older adults. Ann Intern Med. 2007 Dec
4;147(11):755-65. PMID: 18056659.
309. Franciosi M, Lucisano G, Lapice E, et al.
Metformin therapy and risk of cancer in
patients with type 2 diabetes: systematic
review. PLoS One. 2013;8(8):e71583.
PMID: 23936520.
310. Zhang ZJ, Bi Y, Li S, et al. Reduced risk of
lung cancer with metformin therapy in
diabetic patients: a systematic review and
meta-analysis. Am J Epidemiol. 2014 Jul
1;180(1):11-4. PMID: 24920786.
311. Ferwana M, Firwana B, Hasan R, et al.
Pioglitazone and risk of bladder cancer: a
meta-analysis of controlled studies. Diabet
Med. 2013 Sep;30(9):1026-32. PMID:
23350856.
312. Dormandy JA, Charbonnel B, Eckland DJ,
et al. Secondary prevention of
macrovascular events in patients with type 2
diabetes in the PROactive Study
(PROspective pioglitAzone Clinical Trial In
macroVascular Events): a randomised
controlled trial. Lancet. 2005 Oct
8;366(9493):1279-89. PMID: 16214598.
313. Erdmann E, Song E, Spanheimer R, et al.
Observational follow-up of the PROactive
study: a 6-year update. Diabetes Obes
Metab. 2014 Jan;16(1):63-74. PMID:
23859428.
314. U.S. Food and Drug Administration. Incretin
Mimetic Drugs for Type 2 Diabetes: Early
Communication - Reports of Possible
Increased Risk of Pancreatitis and Pre-
cancerous Findings of the Pancreas. 2013;
http://www.fda.gov/Safety/MedWatch/SafetyInfo
rmation/SafetyAlertsforHumanMedicalProducts/
ucm343805.htm. Accessed 2015 August 1.
315. U.S. Food and Drug Administration.
Highlights of Prescribing Information:
Victoza (liraglutide [rDNA origin] injection,
solution for subcutaneous use. 2011;
http://www.accessdata.fda.gov/drugsatfda_docs/l
abel/2011/022341s004lbl.pdf. Accessed 2015
March 2.
316. U.S. Food and Drug Administration.
Highlights of Prescribing Information:
Tanzeum (albiglutide) for injection, for
subcutaneous use. 2014;
http://www.accessdata.fda.gov/drugsatfda_docs/l
abel/2014/125431s000lbl.pdf. Accessed 2015
March 2.
317. U.S. Food and Drug Administration.
Highlights of Prescribing Information:
Bydureon (exenatide extended-release) for
injectable suspension. 2015;
http://www.fda.gov/safety/medwatch/safetyinfor
mation/ucm400570.htm. Accessed 2015
August 7.
318. U.S. Food and Drug Administration.
Trulicity (dulaglutide) Injection, for
Subcutaneous Use. 2015;
http://www.fda.gov/safety/medwatch/safetyinfor
mation/ucm442202.htm. Accessed 2015
August 7.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-403-2048.jpg)
![347
319. U.S. Food and Drug Administration. Victoza
(liraglutide [rDNA origin]) Injection: REMS
- Risk of Thyroid C-cell Tumors, Acute
Pancreatitis. 2011;
http://www.fda.gov/Safety/MedWatch/SafetyInfo
rmation/SafetyAlertsforHumanMedicalProducts/
ucm258826.htm; . Accessed 2015 February
25.
320. Lago RM, Singh PP, Nesto RW. Congestive
heart failure and cardiovascular death in
patients with prediabetes and type 2 diabetes
given thiazolidinediones: a meta-analysis of
randomised clinical trials. Lancet. 2007 Sep
29;370(9593):1129-36. PMID: 17905165.
321. Singh S, Loke YK, Furberg CD. Long-term
risk of cardiovascular events with
rosiglitazone: a meta-analysis. JAMA. 2007
Sep 12;298(10):1189-95. PMID: 17848653.
322. GlaxoSmithKline. Highlights of Prescribing
Information: Avandia (rosiglitazone
maleate) tablets. 2014;
http://us.gsk.com/products/assets/us_avandia.pdf.
Accessed 2015 March 2.
323. Takeda Pharmaceuticals America.
Highlights of Prescribing Information: Actos
(pioglitazone) tables for oral use. 2013;
http://general.takedapharm.com/content/file/pi.pd
f?applicationcode=8a9c4571-a123-4477-
91deb9cafe7d07e3&filetypecode=actospi.
Accessed 2015 March 2.
324. Zannad F, Cannon CP, Cushman WC, et al.
Heart failure and mortality outcomes in
patients with type 2 diabetes taking
alogliptin versus placebo in EXAMINE: a
multicentre, randomised, double-blind trial.
Lancet. 2015 May 23;385(9982):2067-76.
PMID: 25765696.
325. U.S. Food and Drug Administration. FDA
Panel Wants New DPP-4 Inhibitor Labels -
Cardiovascular data warrant new risk
information for saxagliptin and alogliptin,
advisers say.;
http://www.medpagetoday.com/PublicHealthPoli
cy/ClinicalTrials/50990. Accessed 2015 July
25, .
326. Clinicaltrials.gov. CAROLINA:
Cardiovascular Outcome Study of
Linagliptin Versus Glimepiride in Patients
With Type 2 Diabetes. 2010;
https://clinicaltrials.gov/ct2/show/NCT0124
3424. Accessed 2015 July 30, .
327. Clinicaltrials.gov. Cardiovascular and Renal
Microvascular Outcome Study With
Linagliptin in Patients With Type 2 Diabetes
Mellitus (CARMELINA). 2013;
https://clinicaltrials.gov/ct2/show/NCT0189
7532. Accessed 2015 July 30, .
328. NESINA (alogliptin) tablets. 2015;
http://general.takedapharm.com/content/file.aspx
?FileTypeCode=NESINAPI&cacheRandomizer=
f3b08184-5c02-4cd5-a603-466fafb32324.
Accessed 2015 August 7.
329. ACTOS (pioglitazone) tablets for oral use.
2013;
http://general.takedapharm.com/content/file.aspx
?filetypecode=actospi&cacheRandomizer=bd24d
b50-cfd1-4e08-a5f0-1414e6c0d33a. Accessed
2015 August 7.
330. Brown JB, Pedula K, Barzilay J, et al. Lactic
acidosis rates in type 2 diabetes. Diabetes
Care. 1998 Oct;21(10):1659-63. PMID:
9773726.
331. Misbin RI, Green L, Stadel BV, et al. Lactic
acidosis in patients with diabetes treated
with metformin. The New England journal
of medicine. 1998 Jan 22;338(4):265-6.
PMID: 9441244.
332. Salpeter SR, Greyber E, Pasternak GA,
Salpeter EE. Risk of fatal and nonfatal lactic
acidosis with metformin use in type 2
diabetes mellitus. Cochrane Database Syst
Rev. 2010(4):CD002967. PMID: 20393934.
333. Inzucchi SE, Lipska KJ, Mayo H, et al.
Metformin in patients with type 2 diabetes
and kidney disease: a systematic review.
JAMA. 2014 Dec 24-31;312(24):2668-75.
PMID: 25536258.
334. Sethi BK, Viswanathan V, Kumar A, et al.
Liraglutide in Clinical Practice: Insights
from LEAD Programme. Supplement to
JAPI. 2010 June 2010;58:18-22. PMID:
COULD NOT FIND THIS ONE
335. Franks AS, Lee PH, George CM.
Pancreatitis: a potential complication of
liraglutide? Ann Pharmacother. 2012
Nov;46(11):1547-53. PMID: 23136352.
336. Li L, Shen J, Bala MM, et al. Incretin
treatment and risk of pancreatitis in patients
with type 2 diabetes mellitus: systematic
review and meta-analysis of randomised and
non-randomised studies. BMJ.
2014;348:g2366. PMID: 24736555.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-404-2048.jpg)
![A-1
Appendix A. Detailed Electronic Database Search
Strategies
PubMed Strategy
Search String
#1 (“diabetes mellitus, type 2”[mh] or (diabet*[tiab] and (“non-insulin
dependent”[tiab] or type-2[tiab] or “type II”[tiab] or “type 2”[tiab]))) AND
(“metformin”[mh] or “thiazolidinediones”[mh] or “glipizide”[mh] or
“glyburide”[mh] or “Dipeptidyl-Peptidase IV Inhibitors”[mh] or “Glucagon-Like
Peptide 1”[mh] or biguanide*[tiab] or metformin[tiab] or
thiazolidinedione*[tiab] or pioglitazone[tiab] or rosiglitazone[tiab] or
sulfonylurea*[tiab] or sulphonylurea*[tiab] or glipizide[tiab] or glyburide[tiab]
or glimepiride[tiab] or glibenclamide[tiab] or “insulin secretagogues”[tiab] or
sitagliptin*[tiab] or saxagliptin*[tiab] or dpp-4[tiab] or dpp-iv[tiab] or
liraglutide[tiab] or exenatide[tiab]) NOT (animal[mh] NOT human[mh]) NOT
(letter[pt] or comment[pt] or editorial[pt]) AND (("2009/04/01"[edat] :
"2014/07/11"[edat]))
#2 (“diabetes mellitus, type 2”[mh] or (diabet*[tiab] and (“non-insulin
dependent”[tiab] or type-2[tiab] or “type II”[tiab] or “type 2”[tiab]))) AND
(linagliptin*[tiab] or alogliptin*[tiab] or albiglutide*[tiab] or dulaglutide*[tiab]
or "sodium-glucose co-transporter 2 inhibitors”[tiab] or “sodium-glucose co-
transporter 2 inhibitor” [tiab] or “SGLT-2” [tiab] or “canagliflozin”[tiab] or
“dapagliflozin”[tiab]) NOT (animal[mh] NOT human[mh]) NOT (letter[pt] or
comment[pt] or editorial[pt])
#3 (“diabetes mellitus, type 2”[mh] or (diabet*[tiab] and (“non-insulin
dependent”[tiab] or type-2[tiab] or “type II”[tiab] or “type 2”[tiab]))) AND
(empagliflozin*[tiab]) NOT (animal[mh] NOT human[mh]) NOT (letter[pt] or
comment[pt] or editorial[pt])](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-410-2048.jpg)
![A-2
EMBASE Strategy
Search String
#1 ('non insulin dependent diabetes mellitus'/exp OR 'non insulin dependent
diabetes mellitus' or (diabet*:ti,ab and (‘non-insulin dependent’:ti,ab or type-
2:ti,ab or ‘type II’:ti,ab or ‘type 2’:ti,ab))) AND ('thiazolidinedione'/exp or
'rosiglitazone'/exp or 'pioglitazone'/exp or 'glipizide'/exp or 'glyburide'/exp or
‘glimepiride’/exp or 'metformin'/exp or ‘sitagliptin’/exp or
thiazolidinedione*:ti,ab or pioglitazone:ti,ab or rosiglitazone:ti,ab or
sulfonylurea*:ti,ab or sulphonylurea*:ti,ab or glipizide:ti,ab or glyburide:ti,ab or
glimepiride:ti,ab or glibenclamide:ti,ab or biguanide*:ti,ab or metformin:ti,ab or
‘insulin secretagogues’:ti,ab or ‘Dipeptidyl-Peptidase IV Inhibitor’/de or
saxagliptin/exp or saxagliptin*:ti,ab or sitagliptin/exp or sitagliptin*:ti,ab or
dpp-4:ti,ab or dpp-iv:ti,ab or exenatide/exp or exenatide:ti,ab or liraglutide/exp
or liraglutide:ti,ab) NOT ([animals]/lim NOT [humans]/lim) NOT (letter:it or
comment:it or editorial:it) AND [2009-2014]/py
#2 ('non insulin dependent diabetes mellitus'/exp OR 'non insulin dependent
diabetes mellitus' or (diabet*:ti,ab and (‘non-insulin dependent’:ti,ab or type-
2:ti,ab or ‘type II’:ti,ab or ‘type 2’:ti,ab))) AND (linagliptin/exp or
linagliptin*:ti,ab or alogliptin/exp or alogliptin*:ti,ab or albiglutide/exp or
albiglutide*:ti,ab or dulaglutide/exp or dulaglutide*:ti,ab or ‘sodium glucose
cotransporter 2 inhibitor’/de or ‘sodium-glucose co-transporter 2 inhibitors’:ti,ab
or ‘sodium-glucose co-transporter 2 inhibitor’:ti,ab or ‘sodium glucose
cotransporter 2 inhibitors’:ti,ab or ‘sodium glucose cotransporter 2
inhibitor’:ti,ab or ‘SGLT-2”:ti,ab or canagliflozin/exp or canagliflozin:ti,ab or
dapagliflozin/exp or dapagliflozin:ti,ab) NOT ([animals]/lim NOT
[humans]/lim) NOT (letter:it or comment:it or editorial:it)
#3 ('non insulin dependent diabetes mellitus'/exp OR 'non insulin dependent
diabetes mellitus' or (diabet*:ti,ab and (‘non-insulin dependent’:ti,ab or type-
2:ti,ab or ‘type II’:ti,ab or ‘type 2’:ti,ab))) AND (empagliflozin/exp or
empagliflozin*:ti,ab) NOT ([animals]/lim NOT [humans]/lim) NOT (letter:it or
comment:it or editorial:it)](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-411-2048.jpg)
![C-2
hemodialysis. Expert Opin Pharmacother.
2010 Jul;11(10):1611-20. PMID: 20540652.
Background medications
Adetunji O, Skrivanek Z, Tahbaz A, et al. A
post-hoc pooled analysis of two placebo
controlled phase 3 trials, Assessment of
Weekly AdministRation of LY2189265 in
Diabetes-1 and-5 (AWARD-1 and
AWARD-5): Dulaglutide compared with
exenatide, sitagliptin, and placebo.
Diabetologie und Stoffwechsel.
2014;9((Adetunji O.; Tahbaz A.) Eli Lilly
and Company, Medical Affairs,
Basingstoke, United Kingdom).
Meeting abstract
Adetunji O, Skrivanek Z, Tahbaz A, et al. A
posthoc pooled analysis of two placebo-
controlled phase 3 trials, Assessment of
Weekly Administration of LY2189265 in
Diabetes 1 and 5 (AWARD-1 and AWARD-
5): Dulaglutide compared with exenatide,
sitagliptin and placebo. Diabetic Medicine.
2014;31((Adetunji O.; Tahbaz A.) Medical
Department, Eli Lilly and Company,
Basingstoke, United Kingdom):50-1.
Meeting abstract
Agarwala A, Givens E, McGuire DK, et al.
Rosiglitazone increases cholesterol efflux
capacity in patients with type 2 diabetes.
Journal of Investigative Medicine.
2014;62(2):510-1.
Meeting abstract
Agrawal A, Pradeep. To study the pattern of
use and efficacy of anti-diabetic drugs in
controlling adequate glycemic levels in
diabetic patients in Navi Mumbai.
Australasian Medical Journal. 2012;5(1):88-
9.
Meeting abstract
Ajdi F, Khabbal Y, Safi S. ADR of oral
antidiabetic. Drug Safety. 2009;32(10):949-
50.
Meeting abstract
Al Sifri S, Basiounny A, Echtay A, et al.
The incidence of hypoglycaemia in Muslim
patients with type 2 diabetes treated with
sitagliptin or a sulphonylurea during
Ramadan: a randomised trial. Int J Clin
Pract. 2011 Nov;65(11):1132-40. PMID:
21951832.
No drug comparison of interest
Alba M, Ahren B, Inzucchi SE, et al. Initial
combination therapy with sitagliptin and
pioglitazone: Complementary effects on
postprandial glucose and islet cell function.
Canadian Journal of Diabetes.
2009;33(3):319-20.
Meeting abstract
Alexanderson-Rosas E, de Jesus Martinez
A, Ochoa-Lopez JM, et al. [Effects of the
combined treatment with
Metformin/Glimepiride on endothelial
function of patients with type 2 diabetes
mellitus. A positron emission tomography
(PET) evaluation study]. Arch Cardiol Mex.
2009 Oct-Dec;79(4):249-56. PMID:
20191984.
Follow-up less than three months
Alkharfy KM, Al-Daghri NM, Sabico SB, et
al. Vitamin D supplementation in patients
with diabetes mellitus type 2 on different
therapeutic regimens: a one-year prospective
study. Cardiovasc Diabetol. 2013 Aug
7;12(1):113. PMID: 23924389.
No drug comparison of interest
Allen E, Berglind N. Saxagliptin vs glipizide
as add-on therapy to metformin in patients
with type 2 diabetes: A 2-year assessment of
HbA1c, hypoglycaemia, and weight gain in
a randomised, double-blind study.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-434-2048.jpg)
![C-3
Diabetologia. 2011;54((Allen E.; Berglind
N.) Bristol-Myers Squibb, Princeton, United
States):S337.
Meeting abstract
Allen E, Donovan M, Berglind N, et al.
Efficacy of saxagliptin according to patient
baseline characteristics: A pooled analysis
of three add-on pivotal randomised phase 3
clinical trials. Diabetologia. 2010;53((Allen
E.; Donovan M.; Berglind N.) Bristol-Myers
Squibb, Princeton, United States):S328.
Meeting abstract
Allen E, Karyekar C, Ohman P. Safety
profile of saxagliptin (SAXA) in
combination with 2 other agents: Data from
dual-therapy trials in patients receiving
rescue treatment. Diabetes. 2011;60((Allen
E.; Karyekar C.; Ohman P.) Princeton,
United States):A619-A20.
Meeting abstract
Allen E, Slater J, Bryzinski B, et al. Efficacy
and safety of saxagliptin (SAXA) in patients
with type 2 diabetes stratified by
cardiovascular risk factors. Diabetologia.
2012;55((Allen E.; Slater J.) Medical
Affairs, Bristol-Myers Squibb, Princeton,
United States):S346-S7.
Meeting abstract
Alvarez-Guisasola F, Yin DD, Nocea G, et
al. Association of hypoglycemic symptoms
with patients' rating of their health-related
quality of life state: a cross sectional study.
Health Qual Life Outcomes. 2010;8:86.
PMID: 20723229.
Does not apply; No drug comparison of
interest
Ambrosius WT, Danis RP, Goff DC, Jr., et
al. Lack of association between
thiazolidinediones and macular edema in
type 2 diabetes: the ACCORD eye substudy.
Arch Ophthalmol. 2010 Mar;128(3):312-8.
PMID: 20212201.
Background medications
Ametov AS, Gusenbekova DG. [Effect of
dipeptidyl peptidase-4 inhibitors on lipid
metabolism in patients with type 2 diabetes
mellitus]. Ter Arkh. 2014;86(8):85-9.
PMID: 25306750.
Non-English Language
Araki A, Iimuro S, Sakurai T, et al. Long-
term multiple risk factor interventions in
Japanese elderly diabetic patients: The
Japanese Elderly Diabetes Intervention Trial
- study design, baseline characteristics and
effects of intervention. Geriatrics and
Gerontology International.
2012;12(SUPPL.1):7-17.
Does not apply
Aravind SR, Ismail SB, Balamurugan R, et
al. Hypoglycemia in patients with type 2
diabetes from India and Malaysia treated
with sitagliptin or a sulfonylurea during
Ramadan: a randomized, pragmatic study.
Curr Med Res Opin. 2012 Aug;28(8):1289-
96. PMID: 22738801.
Follow-up less than three months;
Background medications
Arcidiacono B, Capula C, Chiefari E, et al.
Glycemic efficacy of liraglutide is linked to
gender in italian type 2 diabetic patients.
Diabetes. 2014;63((Arcidiacono B.; Capula
C.; Chiefari E.; Vero A.; Oliverio R.; Puccio
L.; Liguori R.; Pullano V.; Tirinato D.; Foti
D.; Vero R.; Brunetti A.) Catanzaro, Italy,
Soverato, Italy):A288.
Meeting abstract
Ardawi MS, Akbar D, Al-Shaik A, et al.
Circulating sclerostin, bone turnover
markers and BMD in type-2 diabetic women
treated with metformin or pioglitazone.
Journal of Bone and Mineral Research.](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-435-2048.jpg)
![C-5
Edgbaston, Birmingham, United
Kingdom):S20.
Meeting abstract
Arnolds S, Sawicki PT. Liraglutide and the
preservation of pancreatic (beta)-cell
function in early type 2 diabetes: The libra
trial. Diabetes care 2014;37:3270-3278.
Diabetes Care. 2015;38(2):e25.
Meeting abstract
Arulanandham A, Raju A, Pradeep
Rajkumar LA, et al. Prevalence of clinically
significant macular edema [CSME] among
glitazone users and non- users of type-2 DM
patients with diabetic retinopathy.
International Journal of Drug Development
and Research. 2012;4(2):132-7.
No drug comparison of interest
Asanuma H, Kitakaze M. [Prospective
pioglitazone clinical trial in macrovascular
events]. Nihon Rinsho. 2012 May;70 Suppl
3:301-8. PMID: 22768537.
No original data; No drug comparison of
interest
Aschner P, Sethi B, Gomez-Peralta F, et al.
Glargine vs. premixed insulin for
management of type 2 diabetes patients
failing oral antidiabetic drugs: The
GALAPAGOS study. Diabetes.
2013;62((Aschner P.; Sethi B.; Gomez-
Peralta F.; Landgraf W.; Dain M.-P.;
Pilorget V.; Comlekci A.) Bogota,
Colombia, Hyderabad, India, Segovia,
Spain, Frankfurt, Germany, Paris, France,
Izmir, Turkey):A241-A2.
Meeting abstract
Aschner P, Sethi B, Gomez-Peralta F, et al.
Insulin glargine compared with premixed
insulin for management of insulin-naive
type 2 diabetes patients uncontrolled on oral
antidiabetic drugs: the open-label,
randomized GALAPAGOS study. J
Diabetes Complications. 2015
Aug;29(6):838-45. PMID: 25981123.
Background medications
Aso Y, Takebayashi K, Inukai T, et al.
Pioglitazone and cardiovascular events in
type 2 diabetes: Effects of pioglitazone on
cardiovascular outcomes in Japanese
patients with type 2 diabetes in higashi-
saitama (EPOCH Trial). Diabetes.
2011;60((Aso Y.; Takebayashi K.; Inukai
T.; Katsumori K.; Owada K.; Nakamura T.;
Naito T.; Itabashi H.; Morita K.; Sekine M.;
Takahashi K.; Miyano H.; Takai T.)
Koshigaya, Japan):A557.
Meeting abstract
Atchison L, Steinke EL. Relationship
between social and economic factors in
diabetes medication-prescribing patterns.
Journal of the American Pharmacists
Association. 2011;51(2):228.
Meeting abstract
Aubert RE, Herrera V, Chen W, et al.
Rosiglitazone and pioglitazone increase
fracture risk in women and men with type 2
diabetes. Diabetes Obes Metab. 2010
Aug;12(8):716-21. PMID: 20590749.
Aydin Y, Erden M, Ermis F, et al. Oral
antidiabetics and insulins do not increase
cancer risk. Acta Medica Mediterranea.
2013;29(4):859-67.
Background medications
Azar S, El-Mollayess GM, Al Shaar L, et al.
Impact of thiazolidinediones on macular
thickness and volume in diabetic eyes. Can J
Ophthalmol. 2013 Aug;48(4):312-6. PMID:
23931472.
No drug comparison of interest; Does not
account for confounding
Azar ST, Malha LP, Zantout MS, et al.
Management and control of patients with
type 2 diabetes mellitus in Lebanon: results](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-437-2048.jpg)
![C-6
from the International Diabetes Management
Practices Study (IDMPS). J Med Liban.
2013 Jul-Sep;61(3):127-31. PMID:
24422361.
Background medications
Azoulay L, Dell'Aniello S, Gagnon B, et al.
Metformin and the incidence of prostate
cancer in patients with type 2 diabetes.
Cancer Epidemiol Biomarkers Prev. 2011
Feb;20(2):337-44. PMID: 21148757.
No drug comparison of interest
Azoulay L, Schneider-Lindner V,
Dell'aniello S, et al. Combination therapy
with sulfonylureas and metformin and the
prevention of death in type 2 diabetes: a
nested case-control study.
Pharmacoepidemiol Drug Saf. 2010
Apr;19(4):335-42. PMID: 20052677.
Background medications
Azoulay L, Schneider-Lindner V,
Dell'aniello S, et al. Thiazolidinediones and
the risk of incident strokes in patients with
type 2 diabetes: a nested case-control study.
Pharmacoepidemiol Drug Saf. 2010
Apr;19(4):343-50. PMID: 19998318.
No drug comparison of interest
Azoulay L, Yin H, Filion KB, et al. The use
of pioglitazone and the risk of bladder
cancer in patients with type 2 diabetes.
Pharmacoepidemiology and Drug Safety.
2012;21((Azoulay L.; Yin H.; Filion K.B.;
Assayag J.; Suissa S.) Centre for Clinical
Epidemiolog, Jewish General Hospital,
Montreal, Canada):271.
Meeting abstract
Azoulay L, Yin H, Filion KB, et al. The use
of pioglitazone and the risk of bladder
cancer in people with type 2 diabetes: nested
case-control study. BMJ. 2012;344:e3645.
PMID: 22653981
Does not account for confounding
Bach RG, Brooks MM, Lombardero M, et
al. Rosiglitazone and outcomes for patients
with diabetes mellitus and coronary artery
disease in the Bypass Angioplasty
Revascularization Investigation 2 Diabetes
(BARI 2D) trial. Circulation. 2013 Aug
20;128(8):785-94. PMID: 23857320.
No drug comparison of interest
Bailey CJ, Day C, Campbell IW, et al.
Glycaemic control and cardiovascular
outcome trials in type 2 diabetes. British
Journal of Diabetes and Vascular Disease.
2012;12(4):161-4.
No original data
Bailey CJ, Gross JL, Bastone L, et al.
Dapagliflozin as an add-on to metformin
lowers hyperglycaemia in type 2 diabetes
patients inadequately controlled with
metformin alone. Diabetologia.
2009;52(S1):S76.
Meeting abstract
Bailey CJ, Gross JL, Hennicken D, et al.
Correction to Dapagliflozin add-on to
metformin in type 2 diabetes inadequately
controlled with metformin: A randomized,
double-blind, placebo-controlled 102-week
trial [BMC Medicine, 11, 193, (2013)].
BMC Med. 2013;11(1).
Meeting abstract
Bailey CJ, Gross JL, Yadav M, et al.
Sustained efficacy of dapagliflozin when
added to metformin in type 2 diabetes
inadequately controlled by metformin
monotherapy. Diabetologia. 2011;54((Bailey
C.J.) Aston University, Birmingham, United
Kingdom):S67.
Meeting abstract
Bailey CJ, Iqbal N, T'Joen C, et al.
Dapagliflozin monotherapy in drug-naive
patients with diabetes: a randomized-](https://image.slidesharecdn.com/diabetesmedicationsforadultswithtype2diabetesanupdate-231206032316-386ea178/75/Diabetes-Medications-for-Adults-With-Type-2-Diabetes_-An-Update-pdf-438-2048.jpg)
More Related Content
Similar to Diabetes Medications for Adults With Type 2 Diabetes_ An Update.pdf
Diabetes Medications for Adults With Type 2 Diabetes_ An Update.pdf
- 1. Comparative Effectiveness Review Number173 Diabetes Medications for Adults With Type 2 Diabetes: An Update
- 2. Comparative Effectiveness Review Number173 Diabetes Medications for Adults With Type 2 Diabetes: An Update Prepared for: Agency for Healthcare Research and Quality U.S. Department of Health and Human Services 5600 Fishers Lane Rockville, MD 20857 www.ahrq.gov Contract No. 290-2012-00007-I Prepared by: Johns Hopkins University Evidence-based Practice Center Baltimore, MD Investigators: Shari Bolen, M.D., M.P.H. Eva Tseng, M.D., M.P.H. Susan Hutfless, Ph.D. Jodi B. Segal, M.D., M.P.H. Catalina Suarez-Cuervo, M.D. Zackary Berger, M.D., Ph.D. Lisa M. Wilson, Sc.M. Yue Chu, M.S.P.H. Emmanuel Iyoha, M.B.Ch.B., M.P.H. Nisa M. Maruthur, M.D., M.H.S. AHRQ Publication No. 16-EHC013-EF April 2016
- 3. Addendum and Errata Introduction Duringreport dissemination in a peer-reviewed journal, Annals of Internal Medicine requested an update of our search and evidence. Methods Search Strategy We updated the MEDLINE search to identify randomized controlled trials indexed through December 31, 2015. Evidence Grading We updated the evidence from the final Agency for Healthcare Research and Quality report with results from newly identified randomized trials if these results increased the strength of evidence from low or moderate to moderate or high. Results From the updated MEDLINE search, we identified eight new studies (published in nine articles)1-9 which met our inclusion criteria, plus six additional publications that were either extensions or additional analyses of included studies.10-15 Of these, four studies contributed results which increased the strength of evidence to moderate or high strength; the results of these studies were incorporated in the publication (Appendix Figure 1).16 The updated strength of evidence is shown in Table 1. We report four new findings: (1) sulfonylureas had greater reductions in hemoglobin A1c than dipeptidyl peptidase-4 (DPP-4) inhibitors; (2) sulfonylureas had less beneficial effects on weight than DPP-4 inhibitors; (3) metformin plus glucagon-like peptide-1 receptor agonists had greater weight reductions than metformin plus premixed insulins; and (4) metformin plus thiazolidinediones had less diarrhea than metformin alone (Figures 1 and 4; Appendix Table 6 in the Annals of Internal Medicine manuscript16 ). Conclusions These four new findings strengthen the overall conclusions from the main report.17 ii
- 4. Table 1. Strengthof evidence domains for the comparisons and outcomes that changed with the updated search Comparison Outcome Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary SU vs. DPP-4 inhibitors HbA1c 4 (1659) Medium Consistent Direct Precise Undetected Moderate SU favored; pooled mean between-group difference, -0.2% (95% CI, -0.3 to -0.1%) SU vs. DPP-4 inhibitors Weight 4 (1659) Low Consistent Direct Precise Undetected Moderate DPP-4 inhibitors favored; range in between-group differences of 0.7 to 1.8 kg Metformin + GLP-1 receptor agonists vs. metformin + premixed insulin Weight 2 (426) Medium Consistent Direct Precise Undetected Moderate Metformin + GLP-1 receptor agonists favored; range in between-group differences 1.9 to 5.1 kg Metformin vs. metformin + TZD GI side effects 11 (4,271) 6 studies on diarrhea; 1-2 studies for other GI-related outcomes Medium Consistent Direct Imprecise Undetected Moderate for diarrhea; Low for other GI- related outcomes Metformin + TZD favored for diarrhea; Neither favored for other GI-related outcomes DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GI = gastrointestinal; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; GI = gastrointestinal; kg = kilograms; SU = sulfonylurea; TZD = thiazolidinedione iii
- 5. References 1. Yang W,Han P, Min KW, et al. Efficacy and safety of dapagliflozin in Asian patients with type 2 diabetes after metformin failure: A randomized controlled trial. J Diabetes. 2015 Nov 20. 2. Lu Y, Rodriguez LA, Malgerud L, et al. New-onset type 2 diabetes, elevated HbA1c, anti-diabetic medications, and risk of pancreatic cancer. Br J Cancer. 2015 Dec 1;113(11):1607-14. 3. Tanaka K, Saisho Y, Manesso E, et al. Effects of Liraglutide Monotherapy on Beta Cell Function and Pancreatic Enzymes Compared with Metformin in Japanese Overweight/Obese Patients with Type 2 Diabetes Mellitus: A Subpopulation Analysis of the KIND-LM Randomized Trial. Clin Drug Investig. 2015 Oct;35(10):675-84. 4. Ma Z, Chen R, Liu Y, et al. Effect of liraglutide vs. NPH in combination with metformin on blood glucose fluctuations assessed using continuous glucose monitoring in patients with newly diagnosed type 2 diabetes. Int J Clin Pharmacol Ther. 2015 Nov;53(11):933- 9. 5. Hartley P, Shentu Y, Betz-Schiff P, et al. Efficacy and Tolerability of Sitagliptin Compared with Glimepiride in Elderly Patients with Type 2 Diabetes Mellitus and Inadequate Glycemic Control: A Randomized, Double-Blind, Non-Inferiority Trial. Drugs Aging. 2015 Jun;32(6):469-76. 6. Cai XL, Chen YL, Zhao JJ, et al. Efficacy and safety of avandamet or uptitrated metformin treatment in patients with type 2 diabetes inadequately controlled with metformin alone: a multicenter, randomized, controlled trial. Chin Med J (Engl). 2015 May 20;128(10):1279-87. 7. Xiao CC, Ren A, Yang J, et al. Effects of pioglitazone and glipizide on platelet function in patients with type 2 diabetes. Eur Rev Med Pharmacol Sci. 2015;19(6):963-70. 8. Ross S, Thamer C, Cescutti J, et al. Efficacy and safety of empagliflozin twice daily versus once daily in patients with type 2 diabetes inadequately controlled on metformin: a 16-week, randomized, placebo-controlled trial. Diabetes Obes Metab. 2015 Jul;17(7):699-702. 9. Tanaka K, Saisho Y, Kawai T, et al. Efficacy and safety of liraglutide monotherapy compared with metformin in Japanese overweight/obese patients with type 2 diabetes. Endocr J. 2015;62(5):399-409. 10. Roden M, Merker L, Christiansen AV, et al. Safety, tolerability and effects on cardiometabolic risk factors of empagliflozin monotherapy in drug-naive patients with type 2 diabetes: a double-blind extension of a Phase III randomized controlled trial. Cardiovasc Diabetol. 2015;14(1):154. 11. Chirila C, Zheng Q, Davenport E, et al. Treatment satisfaction in type 2 diabetes patients taking empagliflozin compared with patients taking glimepiride. Qual Life Res. 2015 Sep 30. 12. Li D, Xu X, Zhang Y, et al. Liraglutide treatment causes upregulation of adiponectin and downregulation of resistin in Chinese type 2 diabetes. Diabetes Res Clin Pract. 2015 Nov;110(2):224-8. 13. Simo R, Guerci B, Schernthaner G, et al. Long-term changes in cardiovascular risk markers during administration of exenatide twice daily or glimepiride: results from the European exenatide study. Cardiovasc Diabetol. 2015;14:116. iv
- 6. 14. Li R,Xu W, Luo S, et al. Effect of exenatide, insulin and pioglitazone on bone metabolism in patients with newly diagnosed type 2 diabetes. Acta Diabetol. 2015 Dec;52(6):1083-91. 15. Weinstock RS, Guerci B, Umpierrez et al. Safety and efficacy of once-weekly dulaglutide versus sitagliptin after 2 years in metformin-treated patients with type 2 diabetes (AWARD-5): a randomized, phase III study. Diabetes Obes Metab. 2015 Sep;17(9):849-58. 16. Maruthur NM, Tseng E, Hutfless S, et al. Diabetes Medications as Monotherapy or Metformin-Based Combination Therapy for Type 2 Diabetes: Systematic Review and Meta-Analysis. Ann Intern Med. 2016;In press. 17. Bolen S, Tseng E, Hutfless S, et al. Diabetes Medications for Adults with Type 2 Diabetes: An Update. Comparative Effectiveness Review No. 173. (Prepared by the Johns Hopkins University Evidence-based Practice Center under Contract No. 290-2012- 00007-I.) Rockville, MD: Agency for Healthcare Research and Quality; April 2016. www.effectivehealthcare.ahrq.gov/reports/final.cfm Errata Upon closer review, we found that the primary outcome for Hong 2013 was indeed a composite cardiovascular outcome, which was not what we had stated and that the followup time was 5.0 years rather than 3.0 years. We also note that in the Executive Summary we stated that metformin and GLP-1 receptor agonists were similar for diarrhea, but this was of low and not moderate or high strength and therefore should have not appeared in that section. v
- 7. vi This report isbased on research conducted by the Johns Hopkins University Evidence-based Practice Center (EPC) under contract to the Agency for Healthcare Research and Quality (AHRQ), Rockville, MD (Contract No. 290-2012-00007-I). The findings and conclusions in this document are those of the authors, who are responsible for its contents; the findings and conclusions do not necessarily represent the views of AHRQ. Therefore, no statement in this report should be construed as an official position of AHRQ or of the U.S. Department of Health and Human Services. None of the investigators have any affiliations or financial involvement that conflicts with the material presented in this report. The information in this report is intended to help health care decisionmakers—patients and clinicians, health system leaders, and policymakers, among others—make well-informed decisions and thereby improve the quality of health care services. This report is not intended to be a substitute for the application of clinical judgment. Anyone who makes decisions concerning the provision of clinical care should consider this report in the same way as any medical reference and in conjunction with all other pertinent information, i.e., in the context of available resources and circumstances presented by individual patients. This report is made available to the public under the terms of a licensing agreement between the author and the Agency for Healthcare Research and Quality. This report may be used and reprinted without permission except those copyrighted materials that are clearly noted in the report. Further reproduction of those copyrighted materials is prohibited without the express permission of copyright holders. AHRQ or U.S. Department of Health and Human Services endorsement of any derivative products that may be developed from this report, such as clinical practice guidelines, other quality enhancement tools, or reimbursement or coverage policies, may not be stated or implied. This report may periodically be assessed for the currency of conclusions. If an assessment is done, the resulting surveillance report describing the methodology and findings will be found on the Effective Health Care Program Web site at www.effectivehealthcare.ahrq.gov. Search on the title of the report. Persons using assistive technology may not be able to fully access information in this report. For assistance contact [email protected]. Suggested citation: Bolen S, Tseng E, Hutfless S, Segal JB, Suarez-Cuervo C, Berger Z, Wilson LM, Chu Y, Iyoha E, Maruthur NM. Diabetes Medications for Adults With Type 2 Diabetes: An Update. Comparative Effectiveness Review No. 173. (Prepared by the Johns Hopkins University Evidence-based Practice Center under Contract No. 290-2012-00007-I.) AHRQ Publication No. 16-EHC013-EF. Rockville, MD: Agency for Healthcare Research and Quality; April 2016. www.effectivehealthcare.ahrq.gov/reports/final.cfm.
- 8. vii Preface The Agency forHealthcare Research and Quality (AHRQ), through its Evidence-based Practice Centers (EPCs), sponsors the development of systematic reviews to assist public- and private-sector organizations in their efforts to improve the quality of health care in the United States. These reviews provide comprehensive, science-based information on common, costly medical conditions, and new health care technologies and strategies. Systematic reviews are the building blocks underlying evidence-based practice; they focus attention on the strength and limits of evidence from research studies about the effectiveness and safety of a clinical intervention. In the context of developing recommendations for practice, systematic reviews can help clarify whether assertions about the value of the intervention are based on strong evidence from clinical studies. For more information about AHRQ EPC systematic reviews, see www.effectivehealthcare.ahrq.gov/reference/purpose.cfm. AHRQ expects that these systematic reviews will be helpful to health plans, providers, purchasers, government programs, and the health care system as a whole. Transparency and stakeholder input are essential to the Effective Health Care Program. Please visit the Web site (www.effectivehealthcare.ahrq.gov) to see draft research questions and reports or to join an email list to learn about new program products and opportunities for input. If you have comments on this systematic review, they may be sent by mail to the Task Order Officer named below at: Agency for Healthcare Research and Quality, 5600 Fishers Lane, Rockville, MD 20857, or by email to [email protected]. Richard G. Kronick, Ph.D. Arlene S. Bierman, M.D., M.S. Director Director Agency for Healthcare Research and Quality Center for Evidence and Practice Improvement Agency for Healthcare Research and Quality Stephanie Chang, M.D., M.P.H. Elisabeth U. Kato, M.D. Director Task Order Officer Evidence-based Practice Center Program Center for Evidence and Practice Center for Evidence and Practice Improvement Improvement Agency for Healthcare Research and Quality Agency for Healthcare Research and Quality Karen C. Lee, M.D., M.P.H. Task Order Officer Center for Evidence and Practice Improvement Agency for Healthcare Research and Quality
- 9. viii Acknowledgments We would liketo thank Jessica Gayleard for her help in reviewing articles, Emily Little for her assistance with the evidence tables, Jeanette Edelstein for copy editing, and Supriya Janakiraman. We would also like to thank the Technical Expert Panel, Peer Reviewers, Task Order Officers, and our Associate Editor. Technical Expert Panel In designing the study questions and methodology at the outset of this report, the EPC consulted several technical and content experts. Broad expertise and perspectives were sought. Divergent and conflicted opinions are common and perceived as healthy scientific discourse that results in a thoughtful, relevant systematic review. Therefore, in the end, study questions, design, methodologic approaches, and/or conclusions do not necessarily represent the views of individual technical and content experts. Technical Experts must disclose any financial conflicts of interest greater than $10,000 and any other relevant business or professional conflicts of interest. Because of their unique clinical or content expertise, individuals with potential conflicts may be retained. The TOO and the EPC work to balance, manage, or mitigate any potential conflicts of interest identified. The list of Technical Experts who provided input into this report follows: John Anderson, M.D. The Frist Clinic Nashville, TN Vanita Aroda, M.D.* MedStar Health Research Institute Hyattsville, MD Michael Barry, M.D.* Massachusetts General Hospital Harvard Medical School Boston, MA Judith Fradkin, M.D.* National Institute of Diabetes and Digestive and Kidney Diseases Bethesda, MD Linda Humphrey, M.D., M.P.H.* Portland Veterans Affairs Medical Center Oregon Health & Science University Portland, OR Leonard Pogach, M.D., M.B.A.* Veterans Affairs Central Office Washington, D.C. Robert Ratner, M.D. American Diabetes Association Alexandria, VA Christopher Schmid, Ph.D. Brown University School of Public Health Providence, RI *Provided input on Draft Report. Peer Reviewers Prior to publication of the final evidence report, EPCs sought input from independent Peer Reviewers without financial conflicts of interest. However, the conclusions and synthesis of the scientific literature presented in this report do not necessarily represent the views of individual reviewers.
- 10. ix Peer Reviewers mustdisclose any financial conflicts of interest greater than $10,000 and any other relevant business or professional conflicts of interest. Because of their unique clinical or content expertise, individuals with potential nonfinancial conflicts may be retained. The TOO and the EPC work to balance, manage, or mitigate any potential nonfinancial conflicts of interest identified. The list of Peer Reviewers follows: Jay Desai, Ph.D., M.P.H. HealthPartners Minneapolis, MN David Graham, M.D., M.P.H. U.S. Food and Drug Administration Silver Spring, MD Patrick O’Connor, M.D., M.A., M.P.H. HealthPartners Center for Chronic Care Innovation and HealthPartners Institute for Education and Research Bloomington, MN John Steiner, M.D., M.P.H. Kaiser Permanente Denver, CO
- 11. x Diabetes Medications forAdults With Type 2 Diabetes: An Update Structured Abstract Objectives. To evaluate the comparative effectiveness and safety of monotherapy and metformin-based combination therapy for type 2 diabetes. Data sources. We searched MEDLINE® , Embase® , and the Cochrane Central Register of Controlled Trials (CENTRAL) for English-language articles using the search developed for the prior review (2011), with date restrictions of April 2009 through April 2015. We searched Drugs@FDA and ClinicalTrials.gov for unpublished data. Review methods. Two reviewers independently reviewed titles, abstracts, and full-text articles to identify studies that assessed intermediate and clinical outcomes or safety for monotherapy (metformin, sulfonylureas, thiazolidinediones, dipeptidyl peptidase-4 [DPP-4] inhibitors, glucagon-like peptide-1 [GLP-1] agonists, and sodium glucose cotransporter-2 [SGLT-2] inhibitors) or metformin-based combination therapy (metformin plus one of these monotherapy drugs or insulin) comparisons. Two reviewers extracted data from included articles sequentially using standardized protocols; risk of bias was assessed independently. Two reviewers graded the strength of the evidence sequentially using a protocol adapted from the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) criteria. Results. We included 216 studies and found moderate- or high-strength evidence for the following. Hemoglobin A1c (HbA1c) reduction was similar across all monotherapy comparisons and across metformin-based combination comparisons except DPP-4 inhibitors, which yielded smaller reductions than metformin. Metformin, DPP-4 inhibitors, GLP-1 agonists, and SGLT-2 inhibitors reduced or maintained body weight, while sulfonylureas, thiazolidinediones, and insulin increased weight; between-group differences ranged from 1 to 5 kilograms. SGLT-2 inhibitors and GLP-1 agonists plus metformin reduced systolic blood pressure by 3 to 5 mmHg compared with metformin. Cardiovascular mortality in studies over 2 years in duration was 50 to 70 percent higher for sulfonylureas than metformin (risk difference, 0.1% to 2.9% in randomized controlled trials). Sulfonylurea-based therapy increased the risk of total and severe hypoglycemia versus most comparisons. Gastrointestinal adverse events were higher with metformin than other drugs except GLP-1 agonists, which increased nausea/vomiting 1.5 times compared with metformin. SGLT-2 inhibitors increased the risk of genital mycotic infections over other drugs. The evidence did not support substantive conclusions for microvascular outcomes, congestive heart failure, cancer, pancreatitis, or other safety outcomes. Conclusions. Evidence from this updated systematic review supports metformin as firstline therapy, given its beneficial effects on HbA1c, weight, and cardiovascular mortality (relative to sulfonylureas) and its relative safety profile. In addition, evidence on comparative outcomes associated with different medication classes can be used to facilitate personalized treatment choices by patients and clinicians, guideline development, and decisionmaking by payers and regulators.
- 12. xi Contents Executive Summary................................................................................................................ ES-1 Introduction...................................................................................................................................1 Background................................................................................................................................. 1 Rationale for Update of Review on Comparative Effectiveness of Diabetes Medications ........ 2 Analytic Framework……………………………………………………………………………4 Scope........................................................................................................................................... 7 Key Questions............................................................................................................................. 7 Methods........................................................................................................................................ 10 Topic Refinement and Review Protocol................................................................................... 10 Literature Search Strategy......................................................................................................... 10 Search Strategy ..................................................................................................................... 10 Study Selection ..................................................................................................................... 10 Data Extraction ......................................................................................................................... 13 Risk of Bias Assessment of Individual Studies ........................................................................ 13 Data Synthesis........................................................................................................................... 14 Reporting Bias Assessment................................................................................................... 15 Strength of the Body of Evidence............................................................................................. 15 Applicability ............................................................................................................................. 17 Peer Review and Public Commentary ...................................................................................... 17 Results........................................................................................................................................... 18 Results of Literature Searches .................................................................................................. 18 Study Duration of RCTs for All Key Questions (KQ1–KQ4).................................................. 19 Key Questions 1a and 1b: Intermediate Outcomes................................................................... 20 Study Design and Population Characteristics....................................................................... 20 Risk of Bias........................................................................................................................... 21 Key Points and Evidence Grades for Intermediate Outcomes.............................................. 22 Evidence for Hemoglobin A1c ............................................................................................. 25 Evidence for Weight ............................................................................................................. 52 Evidence for Systolic Blood Pressure................................................................................... 77 Evidence for Heart Rate........................................................................................................ 89 Key Questions 2a and 2b: All-Cause Mortality and Macrovascular and Microvascular Outcomes .................................................................................................................................. 97 Study Design and Population Characteristics....................................................................... 97 Risk of Bias........................................................................................................................... 97 Key Points and Evidence Grades.......................................................................................... 98 Evidence for All-Cause Mortality......................................................................................... 98 Evidence for Cardiovascular Mortality............................................................................... 122 Evidence for Cardiovascular and Cerebrovascular Morbidity............................................ 135 Evidence for Retinopathy ................................................................................................... 154 Evidence for Nephropathy.................................................................................................. 156 Evidence for Neuropathy.................................................................................................... 162 Key Questions 3a and 3b: Safety............................................................................................ 164 Study Design and Population Characteristics..................................................................... 164 Risk of Bias......................................................................................................................... 164 Key Points and Evidence Grades........................................................................................ 164
- 13. xii Evidence for Hypoglycemia............................................................................................... 167 Evidence for Gastrointestinal Side Effects ......................................................................... 207 Evidence for Cancer............................................................................................................ 231 Evidence for Congestive Heart Failure............................................................................... 242 Evidence for Liver Injury.................................................................................................... 251 Evidence for Lactic Acidosis.............................................................................................. 256 Evidence for Pancreatitis .................................................................................................... 258 Evidence for Severe Allergic Reactions ............................................................................. 267 Evidence for Macular Edema or Decreased Vision............................................................ 270 Evidence for Urinary Tract Infections ................................................................................ 272 Evidence for Impaired Renal Function............................................................................... 278 Evidence for Genital Mycotic Infections............................................................................ 284 Evidence for Fracture.......................................................................................................... 292 Evidence for Volume Depletion ......................................................................................... 294 Key Question 4: Subgroups .................................................................................................... 299 Subgroups Defined by Age................................................................................................. 299 Subgroups Defined by Sex.................................................................................................. 300 Subgroups Defined by Race/Ethnicity................................................................................ 301 Subgroups Defined by Body Mass Index ........................................................................... 301 Discussion................................................................................................................................... 302 Key Findings in Context......................................................................................................... 302 Intermediate Outcomes ....................................................................................................... 302 All-Cause Mortality and Macrovascular and Microvascular Outcomes............................. 306 Adverse Events ................................................................................................................... 309 Subgroups ........................................................................................................................... 315 Applicability ........................................................................................................................... 316 Implications for Clinical and Policy Decisionmaking............................................................ 317 Limitations of the Comparative Effectiveness Review Process ............................................. 318 Limitations of the Evidence Base ........................................................................................... 320 Research Gaps and Future Research Needs............................................................................ 321 Conclusions............................................................................................................................. 325 References.................................................................................................................................. 326 Abbreviations ............................................................................................................................ 352 Tables Table A. Priority medication comparisons included for each Key Question ........................... ES-3 Table B. Study inclusion criteria .............................................................................................. ES-5 Table C. Summary of the moderate- to high-strength evidence on the comparative effectiveness and safety of diabetes medications as monotherapy and metformin-based combination therapy for systolic blood pressure and heart rate ............................................. ES-12 Table D. Comparative effectiveness of sulfonylureas compared with metformin for long- term all-cause mortality and cardiovascular mortality and morbidity—moderate strength of evidence or consistent low-strength evidence .................................................................... ES-13 Table E. Summary of the moderate- to high-strength evidence on the comparative safety of diabetes medications as monotherapy and metformin-based combination therapy for genital mycotic infections....................................................................................................... ES-17
- 14. xiii Table F. Evidencegaps and future research needs for the comparative effectiveness and safety of diabetes medications for adults with type 2 diabetes............................................... ES-25 Table 1. Characteristics of medications included in this report...................................................... 3 Table 2. Priority medication comparisons included for each Key Question .................................. 9 Table 3. Inclusion and exclusion criteria ...................................................................................... 11 Table 4. Optimal information size for one arm and classification of dichotomous outcomes for optimal information size.......................................................................................................... 16 Table 5. Pooled mean between-group difference in HbA1c comparing metformin with a combination of metformin plus a thiazolidinedione stratified by baseline HbA1c and dosing differences..................................................................................................................................... 31 Table 6. Strength of evidence domains for monotherapy comparisons in terms of hemoglobin A1c among adults with type 2 diabetes ........................................................................................ 47 Table 7. Strength of evidence domains for metformin versus metformin-based combination comparisons in terms of hemoglobin A1c among adults with type 2 diabetes............................. 49 Table 8. Strength of evidence domains for metformin-based combination comparisons in terms of hemoglobin A1c among adults with type 2 diabetes...................................................... 50 Table 9. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus a sulfonylurea, stratified by baseline weight.............................. 59 Table 10. Strength of evidence domains for monotherapy comparisons in terms of weight among adults with type 2 diabetes................................................................................................ 72 Table 11. Strength of evidence domains for metformin versus metformin-based combination comparisons in terms of weight among adults with type 2 diabetes............................................. 74 Table 12. Strength of evidence domains for metformin-based combination comparisons in terms of weight among adults with type 2 diabetes...................................................................... 75 Table 13. Strength of evidence domains for monotherapy comparisons in terms of systolic blood pressure among adults with type 2 diabetes........................................................................ 86 Table 14. Strength of evidence domains for metformin versus metformin-based combination comparisons in terms of systolic blood pressure among adults with type 2 diabetes................... 87 Table 15. Strength of evidence domains for metformin-based combination comparisons in terms of systolic blood pressure among adults with type 2 diabetes ............................................ 88 Table 16. Strength of evidence domains for monotherapy comparisons in terms of heart rate among adults with type 2 diabetes................................................................................................ 94 Table 17. Strength of evidence domains for metformin versus metformin-based combination comparisons in terms of heart rate among adults with type 2 diabetes ........................................ 95 Table 18. Strength of evidence domains for metformin-based combination comparisons in terms of heart rate among adults with type 2 diabetes.................................................................. 96 Table 19. Observational studies comparing metformin with thiazolidinediones on all-cause mortality...................................................................................................................................... 100 Table 20. Randomized controlled trials comparing metformin with sulfonylureas on all-cause mortality...................................................................................................................................... 101 Table 21. Observational studies comparing metformin with sulfonylureas on all-cause mortality...................................................................................................................................... 102 Table 22. Randomized controlled trials comparing sulfonylureas with GLP-1 receptor agonists on all-cause mortality.................................................................................................... 106
- 15. xiv Table 23. Randomizedcontrolled trials or arms of randomized controlled trials excluded from the meta-analysis comparing metformin with a combination of metformin plus a DPP-4 inhibitor on all-cause mortality................................................................................................... 110 Table 24. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a GLP-1 receptor agonist on all-cause mortality...................................................................................................................................... 114 Table 25. Randomized controlled trials comparing a combination of metformin plus sitagliptin with a combination of metformin plus a GLP-1 receptor agonist on all-cause mortality...................................................................................................................................... 116 Table 26. Strength of evidence domains for monotherapy comparisons in terms of all-cause mortality among adults with type 2 diabetes .............................................................................. 117 Table 27. Strength of evidence domains for metformin versus metformin-based combination comparisons in terms of all-cause mortality among adults with type 2 diabetes........................ 119 Table 28. Strength of evidence domains for metformin-based combination comparisons in terms of all-cause mortality among adults with type 2 diabetes................................................. 120 Table 29. Randomized controlled trials comparing metformin with thiazolidinediones on cardiovascular mortality.............................................................................................................. 122 Table 30. Observational studies comparing metformin with sulfonylureas on cardiovascular mortality...................................................................................................................................... 123 Table 31. Randomized controlled trials comparing metformin with DPP-4 inhibitors on cardiovascular mortality.............................................................................................................. 123 Table 32. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor on cardiovascular mortality...................................................................................................................................... 129 Table 33. Strength of evidence domains for monotherapy comparisons in terms of cardiovascular mortality among adults with type 2 diabetes...................................................... 132 Table 34. Strength of evidence domains for metformin versus metformin-based combination comparisons in terms of cardiovascular mortality among adults with type 2 diabetes .............. 133 Table 35. Strength of evidence domains for metformin-based combination comparisons in terms of cardiovascular mortality among adults with type 2 diabetes........................................ 134 Table 36. Randomized controlled trials comparing metformin with rosiglitazone on cardiovascular morbidity ............................................................................................................ 135 Table 37. Randomized controlled trials comparing metformin with pioglitazone on cardiovascular morbidity ............................................................................................................ 135 Table 38. Retrospective cohort studies comparing metformin with rosiglitazone on cardiovascular morbidity ............................................................................................................ 136 Table 39. Retrospective cohort studies comparing metformin with pioglitazone on cardiovascular morbidity ............................................................................................................ 136 Table 40. Randomized controlled trials comparing metformin with sulfonylureas on cardiovascular morbidity ............................................................................................................ 137 Table 41. Retrospective cohort studies comparing metformin with sulfonylureas on cardiovascular morbidity ............................................................................................................ 138 Table 42. Nested case-control study comparing metformin with sulfonylureas on hospitalization for incidence of myocardial infarction............................................................... 138 Table 43. Studies comparing rosiglitazone with sulfonylureas on cardiovascular morbidity.... 140 Table 44. Studies comparing pioglitazone with sulfonylureas on cardiovascular morbidity..... 141
- 16. xv Table 45. Randomizedcontrolled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on cardiovascular morbidity................................................. 144 Table 46. Randomized controlled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on nonfatal stroke................................................................. 144 Table 47. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor on cardiovascular morbidity..................................................................................................................................... 147 Table 48. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor on cerebrovascular morbidity..................................................................................................................................... 147 Table 49. Strength of evidence domains for monotherapy comparisons in terms of cardiovascular and cerebrovascular morbidity among adults with type 2 diabetes.................... 151 Table 50. Strength of evidence domains for metformin versus metformin-based combination comparisons in terms of cardiovascular and cerebrovascular morbidity among adults with type 2 diabetes............................................................................................................................. 152 Table 51. Strength of evidence domains for combination therapy comparisons in terms of cardiovascular and cerebrovascular morbidity among adults with type 2 diabetes.................... 153 Table 52. Strength of evidence domains for comparisons in terms of retinopathy among adults with type 2 diabetes.......................................................................................................... 155 Table 53. Retrospective cohort studies comparing metformin with sulfonylureas on nephropathy................................................................................................................................. 157 Table 54. Strength of evidence domains for monotherapy comparisons in terms of nephropathy among adults with type 2 diabetes......................................................................... 160 Table 55. Strength of evidence domains for metformin-based combination comparisons in terms of nephropathy among adults with type 2 diabetes........................................................... 161 Table 56. Strength of evidence domains for comparisons in terms of neuropathy among adults with type 2 diabetes.......................................................................................................... 163 Table 57. Randomized controlled trials comparing metformin with thiazolidinediones on hypoglycemia.............................................................................................................................. 168 Table 58. Studies comparing metformin with sulfonylureas for hypoglycemia......................... 170 Table 59. Randomized controlled trials comparing metformin with GLP-1 receptor agonists on hypoglycemia......................................................................................................................... 173 Table 60. Randomized controlled trials comparing thiazolidinediones with sulfonylureas on mild to moderate hypoglycemia.................................................................................................. 175 Table 61. Randomized controlled trials comparing thiazolidinediones with DPP-4 inhibitors on hypoglycemia......................................................................................................................... 176 Table 62. Randomized controlled trials comparing sulfonylureas with DPP-4 inhibitors on hypoglycemia.............................................................................................................................. 177 Table 63. Randomized controlled trials comparing sulfonylureas with GLP-1 receptor agonists on hypoglycemia........................................................................................................... 178 Table 64. Additional randomized controlled trials comparing metformin with a combination of metformin plus a sulfonylurea on hypoglycemia ................................................................... 182 Table 65. Randomized controlled trials comparing metformin with a combination of metformin plus a GLP-1 receptor agonist on hypoglycemia...................................................... 187 Table 66. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a premixed insulin on hypoglycemia ....... 193
- 17. xvi Table 67. Randomizedcontrolled trials comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist on hypoglycemia... 195 Table 68. Randomized controlled trials comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a basal insulin on hypoglycemia .................... 196 Table 69. Randomized controlled trials comparing a combination of metformin plus a GLP-1 receptor agonist with a combination of metformin plus a basal insulin on hypoglycemia......... 196 Table 70. Randomized controlled trials comparing a combination of metformin plus a basal insulin with a combination of metformin plus a premixed insulin on hypoglycemia ................ 197 Table 71. Strength of evidence domains for monotherapy comparisons in terms of hypoglycemia among adults with type 2 diabetes ...................................................................... 199 Table 72. Strength of evidence domains for metformin versus metformin-based combination comparisons in terms of hypoglycemia among adults with type 2 diabetes............................... 202 Table 73. Strength of evidence domains for metformin-based combination comparisons in terms of hypoglycemia among adults with type 2 diabetes ........................................................ 203 Table 74. Studies comparing metformin with sulfonylureas on gastrointestinal adverse events208 Table 75. Randomized controlled trials comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a sulfonylurea on gastrointestinal adverse events ............................................................................................................................. 222 Table 76. Strength of evidence domains for monotherapy comparisons in terms of gastrointestinal side effects among adults with type 2 diabetes ................................................. 226 Table 77. Strength of evidence domains for metformin versus metformin combination comparisons in terms of gastrointestinal side effects among adults with type 2 diabetes.......... 228 Table 78. Strength of evidence domains for metformin-based combination comparisons in terms of gastrointestinal side effects among adults with type 2 diabetes ................................... 229 Table 79. Retrospective cohort studies comparing metformin with sulfonylureas on cancer.... 231 Table 80. Randomized controlled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on cancer .............................................................................. 233 Table 81. Randomized controlled trials comparing metformin with a combination of metformin plus an SGLT-2 inhibitor on cancer.......................................................................... 234 Table 82. Randomized controlled trials comparing a combination of metformin with a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor on cancer ..................... 235 Table 83. Strength of evidence domains for monotherapy comparisons and cancer outcomes among adults with type 2 diabetes.............................................................................................. 239 Table 84. Strength of evidence domains for metformin versus metformin-based combination comparisons and cancer outcomes among adults with type 2 diabetes ...................................... 240 Table 85. Strength of evidence domains for combination therapy comparisons and cancer among adults with type 2 diabetes.............................................................................................. 241 Table 86. Studies comparing metformin with thiazolidinediones on congestive heart failure... 242 Table 87. Studies comparing metformin with sulfonylureas on congestive heart failure .......... 243 Table 88. Observational studies comparing thiazolidinediones with sulfonylureas on congestive heart failure............................................................................................................... 243 Table 89. Randomized controlled trials comparing metformin with a combination of metformin plus a thiazolidinedione on congestive heart failure................................................. 245 Table 90. Strength of evidence domains for monotherapy comparisons in terms of congestive heart failure among adults with type 2 diabetes.......................................................................... 248
- 18. xvii Table 91. Strengthof evidence domains for monotherapy versus metformin-based combination comparisons in terms of congestive heart failure among adults with type 2 diabetes ....................................................................................................................................... 249 Table 92. Strength of evidence domains for metformin-based combination comparisons in terms of congestive heart failure among adults with type 2 diabetes ......................................... 250 Table 93. Randomized controlled trials comparing metformin with thiazolidinediones on liver injury................................................................................................................................... 251 Table 94. Randomized controlled trials comparing metformin with sulfonylureas on liver injury........................................................................................................................................... 252 Table 95. Randomized controlled trials comparing thiazolidinediones with sulfonylureas on liver injury................................................................................................................................... 252 Table 96. Randomized controlled trials comparing metformin with metformin plus DPP-4 inhibitors on liver injury ............................................................................................................. 253 Table 97. Strength of evidence domains for comparisons in terms of liver injury among adults with type 2 diabetes.......................................................................................................... 255 Table 98. Strength of evidence domains for comparisons in terms of lactic acidosis among adults with type 2 diabetes.......................................................................................................... 257 Table 99. Randomized controlled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on pancreatitis ...................................................................... 259 Table 100. Randomized controlled trials comparing metformin with a combination of metformin plus a GLP-1 receptor agonist on pancreatitis .......................................................... 260 Table 101. Randomized controlled trials comparing the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor for pancreatitis ......... 261 Table 102. Strength of evidence domains for monotherapy comparisons in terms of pancreatitis among adults with type 2 diabetes........................................................................... 264 Table 103. Strength of evidence domains for metformin monotherapy versus metformin- based combination comparisons in terms of pancreatitis among adults with type 2 diabetes.... 265 Table 104. Strength of evidence domains for metformin-based combination comparisons in terms of pancreatitis among adults with type 2 diabetes ............................................................ 266 Table 105. Randomized controlled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on severe allergic reactions.................................................. 268 Table 106. Strength of evidence domains for comparisons in terms of severe allergic reactions among adults with type 2 diabetes............................................................................... 269 Table 107. Strength of evidence domains for monotherapy comparisons in terms of macular edema or decreased vision among adults with type 2 diabetes................................................... 271 Table 108. Definitions of urinary tract infections used in randomized controlled trials comparing metformin with a combination of metformin and SGLT-2 inhibitor........................ 273 Table 109. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor on urinary tract infections..................................................................................................................................... 275 Table 110. Randomized controlled trials comparing a combination of metformin plus a DPP- 4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor on urinary tract infections..................................................................................................................................... 276 Table 111. Strength of evidence domains for monotherapy and metformin-based combination comparisons in terms of urinary tract infections among adults with type 2 diabetes................. 277
- 19. xviii Table 112. Randomizedcontrolled trials comparing metformin with SGLT-2 inhibitors on impaired renal function............................................................................................................... 278 Table 113. Randomized controlled trials comparing metformin with a combination of metformin plus an SGLT-2 inhibitor on impaired renal function............................................... 279 Table 114. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor on impaired renal function ....................................................................................................................................... 281 Table 115. Randomized controlled trials comparing a combination of metformin plus a DPP- 4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor on impaired renal function ....................................................................................................................................... 282 Table 116. Strength of evidence domains for monotherapy and metformin-based combination comparisons in terms of impaired renal function among adults with type 2 diabetes................ 283 Table 117. Randomized controlled trials comparing DPP-4 inhibitors with SGLT-2 inhibitors on genital infections.................................................................................................................... 285 Table 118. Randomized controlled trials comparing metformin with a combination of metformin plus an SGLT-2 inhibitor on genital infections ........................................................ 287 Table 119. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor on genital infections288 Table 120. Randomized controlled trials comparing a combination of metformin plus a DPP- 4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor on genital infections .. 290 Table 121. Strength of evidence domains for monotherapy and metformin-based combination comparisons in terms of genital mycotic infections among adults with type 2 diabetes............ 291 Table 122. Strength of evidence domains for monotherapy and metformin-based combination comparisons in terms of fracture among adults with type 2 diabetes......................................... 293 Table 123. Randomized controlled trials comparing metformin with a combination of metformin plus an SGLT-2 inhibitor on volume depletion ........................................................ 295 Table 124. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor on volume depletion296 Table 125. Strength of evidence domains for monotherapy and metformin-based combination comparisons in terms of volume depletion among adults with type 2 diabetes.......................... 298 Table 126. Placebo-controlled RCTs evaluating DPP-4 inhibitors added to standard treatment with composite cardiovascular primary outcome ....................................................................... 308 Table 127. Evidence gaps and future research needs for the comparative effectiveness and safety of diabetes medications for adults with type 2 diabetes................................................... 323 Figures Figure A. Duration of followup for randomized controlled trials comparing the effects of diabetes medications among adults with type 2 diabetes (N = 174)......................................... ES-8 Figure B. Pooled between-group differences in hemoglobin A1c and strength of evidence for monotherapy and metformin-based combination comparisons .......................................... ES-9 Figure C. Pooled between-group differences in weight and strength of evidence for monotherapy and metformin-based combination comparisons.............................................. ES-11 Figure D. Pooled odds ratios of mild/moderate hypoglycemia and strength of evidence for monotherapy and metformin-based combination comparisons ........................................ ES-14 Figure E. Pooled odds ratios of gastrointestinal adverse events and strength of evidence for monotherapy and metformin-based combination comparisons ........................................ ES-16 Figure 1. Analytic framework......................................................................................................... 6
- 20. xix Figure 2. Summaryof the search (number of articles) ................................................................. 19 Figure 3. Duration of followup for randomized controlled trials comparing the effects of diabetes medications among adults with type 2 diabetes (N = 177)............................................. 20 Figure 4. Summary of the risk of bias of randomized controlled trials evaluating intermediate outcomes ....................................................................................................................................... 21 Figure 5. Pooled mean between-group difference in hemoglobin A1c comparing metformin with thiazolidinediones................................................................................................................. 26 Figure 6. Pooled mean between-group difference in hemoglobin A1c comparing metformin with DPP-4 inhibitors ................................................................................................................... 27 Figure 7. Pooled mean between-group difference in hemoglobin A1c comparing thiazolidinediones with sulfonylureas........................................................................................... 29 Figure 8. Pooled mean between-group difference in hemoglobin A1c comparing metformin with a combination of metformin plus a sulfonylurea.................................................................. 32 Figure 9. Pooled mean between-group difference in hemoglobin A1c comparing metformin with a combination of metformin plus a DPP-4 inhibitor............................................................. 34 Figure 10. Pooled mean between-group difference in hemoglobin A1c comparing metformin with a combination of metformin plus an SGLT-2 inhibitor........................................................ 35 Figure 11. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a sulfonylurea................................................................................................................................... 37 Figure 12. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a DPP-4 inhibitor............................................................................................................................. 38 Figure 13. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor, stratified by study duration............................................................................................ 39 Figure 14. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor .......................................................................................................................... 40 Figure 15. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus daily exenatide ....................................................................................................................................... 42 Figure 16. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor .......................................................................................................................... 43 Figure 17. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist................................................................................................................. 44 Figure 18. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a basal insulin with a combination of metformin plus a premixed insulin............................................................................................................................ 45 Figure 19. Pooled mean between-group difference in weight comparing metformin with DPP-4 inhibitors............................................................................................................................ 53 Figure 20. Pooled mean between-group difference in weight comparing thiazolidinediones with sulfonylureas......................................................................................................................... 55
- 21. xx Figure 21. Pooledmean between-group difference in weight comparing sulfonylureas with GLP-1 receptor agonists ............................................................................................................... 56 Figure 22. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus a thiazolidinedione ..................................................................... 58 Figure 23. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus a sulfonylurea............................................................................. 59 Figure 24. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus a DPP-4 inhibitor, stratified by study duration .......................... 60 Figure 25. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus an SGLT-2 inhibitor................................................................... 61 Figure 26. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus a GLP-1 receptor agonist ........................................................... 62 Figure 27. Pooled mean between-group difference in weight comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a sulfonylurea......... 63 Figure 28. Pooled mean between-group difference in weight comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a DPP-4 inhibitor ... 64 Figure 29. Pooled mean between-group difference in weight comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor, stratified by study duration ........................................................................................................... 66 Figure 30. Pooled mean between-group difference in weight comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor....... 67 Figure 31. Pooled mean between-group difference in weight comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist........................................................................................................................................... 69 Figure 32. Pooled mean between-group difference in weight comparing a combination of metformin plus a basal insulin with a combination of metformin plus a premixed insulin.......... 71 Figure 33. Pooled mean between-group difference in systolic blood pressure comparing metformin with SGLT-2 inhibitors............................................................................................... 77 Figure 34. Pooled mean between-group difference in systolic blood pressure comparing metformin with a combination of metformin plus an SGLT-2 inhibitor...................................... 80 Figure 35. Pooled mean between-group difference in systolic blood pressure comparing metformin with a combination of metformin plus a GLP-1 receptor agonist............................... 81 Figure 36. Pooled mean between-group difference in systolic blood pressure comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor .......................................................................................................................... 82 Figure 37. Pooled mean between-group difference in systolic blood pressure comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor .......................................................................................................................... 84 Figure 38. Pooled mean between-group difference in heart rate comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor....... 91 Figure 39. Pooled odds ratio of short-term all-cause mortality comparing metformin with pioglitazone................................................................................................................................... 99 Figure 40. Pooled odds ratio of short-term all-cause mortality comparing metformin with DPP-4 inhibitors.......................................................................................................................... 103 Figure 41. Pooled odds ratio for short-term all-cause mortality comparing metformin with SGLT-2 inhibitors....................................................................................................................... 104
- 22. xxi Figure 42. Pooledodds ratio of short-term all-cause mortality comparing metformin with a combination of metformin plus rosiglitazone............................................................................. 107 Figure 43. Pooled odds ratio of short-term all-cause mortality comparing metformin with a combination of metformin plus a sulfonylurea........................................................................... 108 Figure 44. Pooled odds ratio for short-term all-cause mortality comparing metformin with a combination of metformin plus a DPP-4 inhibitor ..................................................................... 109 Figure 45. Pooled odds ratio for short-term all-cause mortality comparing metformin with a combination of metformin plus an SGLT-2 inhibitor, stratified by study duration.................... 111 Figure 46. Pooled odds ratio for long-term all-cause mortality comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor ......... 113 Figure 47. Pooled odds ratio for long-term all-cause mortality comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor..... 114 Figure 48. Pooled odds ratio for short-term cardiovascular mortality comparing metformin with a combination of metformin plus rosiglitazone .................................................................. 125 Figure 49. Pooled odds ratio for short-term cardiovascular mortality comparing metformin with a combination of metformin plus a DPP-4 inhibitor........................................................... 126 Figure 50. Pooled odds ratio for long-term cardiovascular mortality comparing combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor..... 128 Figure 51. Pooled odds ratio of cardiovascular morbidity comparing metformin with a combination of metformin plus rosiglitazone............................................................................. 142 Figure 52. Pooled odds ratio of short-term cardiovascular morbidity comparing metformin with a combination of metformin plus a DPP-4 inhibitor........................................................... 143 Figure 53. Pooled odds ratio of cardiovascular morbidity comparing combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor ......... 148 Figure 54. Pooled odds ratio of mild or moderate hypoglycemia comparing metformin with sulfonylureas............................................................................................................................... 169 Figure 55. Pooled odds ratio of symptomatic hypoglycemia comparing metformin with DPP- 4 inhibitors .................................................................................................................................. 171 Figure 56. Pooled odds ratio of any hypoglycemia comparing metformin with SGLT-2 inhibitors ..................................................................................................................................... 172 Figure 57. Pooled odds ratio of any hypoglycemia comparing thiazolidinediones with sulfonylureas............................................................................................................................... 174 Figure 58. Pooled odds ratio of any hypoglycemia comparing metformin with combination of metformin plus a thiazolidinedione ............................................................................................ 180 Figure 59. Odds ratios for studies evaluating mild or moderate hypoglycemia comparing metformin with combination of metformin plus a sulfonylurea................................................. 181 Figure 60. Pooled odds ratio of mild or moderate hypoglycemia comparing metformin with combination of metformin plus a DPP-4 inhibitor ..................................................................... 183 Figure 61. Pooled odds ratio of severe hypoglycemia comparing metformin with combination of metformin plus a DPP-4 inhibitor .......................................................................................... 184 Figure 62. Pooled odds ratio of any hypoglycemia comparing metformin with combination of metformin plus a DPP-4 inhibitor............................................................................................... 185 Figure 63. Pooled odds ratio of any hypoglycemia comparing metformin with combination of metformin plus an SGLT-2 inhibitor.......................................................................................... 186 Figure 64. Pooled odds ratio of any hypoglycemia comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a sulfonylurea......................... 188
- 23. xxii Figure 65. Pooledodds ratio of hypoglycemia comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor, stratified by study duration and severity of hypoglycemia....................................................................................... 190 Figure 66. Pooled odds ratio of mild or moderate hypoglycemia comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor..... 191 Figure 67. Odds ratio of hypoglycemia comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a GLP-1 receptor agonist, stratified by study duration and severity of hypoglycemia ............................................................................. 192 Figure 68. Odds ratio of hypoglycemia comparing a combination of metformin plus an SGLT-2 inhibitor with a combination of metformin plus a DPP-4 inhibitor, stratified by severity of hypoglycemia............................................................................................................ 194 Figure 69. Pooled odds ratio of gastrointestinal adverse events comparing metformin with thiazolidinediones ....................................................................................................................... 207 Figure 70. Odds ratio of gastrointestinal adverse events comparing metformin with DPP-4 inhibitors ..................................................................................................................................... 209 Figure 71. Odds ratio of gastrointestinal adverse events comparing metformin with SGLT-2 inhibitors ..................................................................................................................................... 210 Figure 72. Odds ratio of gastrointestinal adverse events comparing metformin with GLP-1 receptor agonists ......................................................................................................................... 211 Figure 73. Odds ratio of gastrointestinal adverse events comparing thiazolidinediones with sulfonylureas............................................................................................................................... 212 Figure 74. Pooled odds ratio of gastrointestinal adverse events comparing sulfonylureas with GLP-1 receptor agonists ............................................................................................................. 213 Figure 75. Odds ratio of gastrointestinal adverse events comparing metformin with a combination of metformin plus a thiazolidinedione ................................................................... 214 Figure 76. Odds ratio of gastrointestinal adverse events comparing metformin with a combination of metformin plus a sulfonylurea........................................................................... 215 Figure 77. Pooled odds ratio of abdominal pain or nausea comparing metformin with a combination of metformin plus a DPP-4 inhibitor ..................................................................... 216 Figure 78. Pooled odds ratio of any gastrointestinal adverse event comparing metformin with a combination of metformin plus a DPP-4 inhibitor................................................................... 217 Figure 79. Pooled odds ratio of diarrhea comparing metformin with a combination of metformin plus a DPP-4 inhibitor............................................................................................... 218 Figure 80. Pooled odds ratio of vomiting comparing metformin with a combination of metformin plus a DPP-4 inhibitor............................................................................................... 219 Figure 81. Pooled odds ratio of gastrointestinal adverse events comparing metformin with a combination of metformin plus an SGLT-2 inhibitor................................................................. 220 Figure 82. Odds ratio of gastrointestinal adverse events comparing metformin with a combination of metformin plus a GLP-1 receptor agonist ......................................................... 221 Figure 83. Odds ratio of gastrointestinal adverse events comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor ......... 223 Figure 84. Odds ratio of gastrointestinal adverse events comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist......................................................................................................................................... 224 Figure 85. Pooled odds ratio of cancer events comparing the combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist ................. 237
- 24. xxiii Figure 86. Pooledodds ratio of congestive heart failure events comparing thiazolidinediones with sulfonylureas....................................................................................................................... 244 Figure 87. Pooled odds ratio of congestive heart failure events comparing metformin with a combination of metformin plus a DPP-4 inhibitor ..................................................................... 245 Figure 88. Odds ratio of pancreatitis comparing the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor .................................. 262 Figure 89. Pooled odds ratio of urinary tract infections comparing metformin with SGLT-2 inhibitors ..................................................................................................................................... 272 Figure 90. Pooled odds ratio of short-term risk of urinary tract infections comparing metformin with a combination of metformin plus an SGLT-2 inhibitor.................................... 274 Figure 91. Pooled odds ratio of genital or mycotic infections comparing metformin with SGLT-2 inhibitors....................................................................................................................... 284 Figure 92. Pooled odds ratio of genital or mycotic infections comparing metformin with a combination of metformin plus an SGLT-2 inhibitor................................................................. 286 Figure 93. Pooled odds ratio of genital or mycotic infections comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor..... 289 Figure 94. Pooled odds ratio of volume depletion comparing metformin with a combination of metformin plus an SGLT-2 inhibitor...................................................................................... 294 Figure 95. Pooled odds ratio of volume depletion comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor ...................... 296 Figure 96. Pooled between-group differences in hemoglobin A1c and strength of evidence for monotherapy and metformin-based combination comparisons ............................................ 303 Figure 97. Pooled between-group differences in weight and strength of evidence for monotherapy and metformin-based combination comparisons.................................................. 305 Figure 98. Pooled odds ratios of mild/moderate hypoglycemia and strength of evidence for monotherapy and metformin-based combination comparisons.................................................. 310 Figure 99. Pooled odds ratios of gastrointestinal adverse events and strength of evidence for monotherapy and metformin-based combination comparisons.................................................. 311 Appendixes Appendix A. Detailed Electronic Database Search Strategies Appendix B. Forms Appendix C. List of Excluded Articles Appendix D. Evidence Tables Appendix E. Gray Literature Appendix F. Key Points and Evidence Grades Appendix G. References
- 25. ES-1 Executive Summary Condition andTherapeutic Strategies Type 2 diabetes affects more than 9.3 percent of the U.S. population, or 29.1 million people.1 Diabetes and its complications are a substantial public health burden, as they contribute significantly to mortality, morbidity, and health care costs.1 Complications of longstanding diabetes include the microvascular complications of retinopathy and blindness, neuropathy, nephropathy, and end-stage kidney disease. Diabetes also contributes importantly to macrovascular complications, including coronary artery disease, peripheral arterial disease, and carotid artery disease, and increases the risk of cardiovascular-related death nearly twofold.2 Lifestyle modification and pharmacologic therapy are the cornerstones of the management of hyperglycemia for type 2 diabetes to reduce diabetes complications.3-5 When beginning medical treatment, patients usually begin with a medication from one of six drug classes that have been approved by the Food and Drug Administration (FDA) for use as monotherapy, although several guidelines recommend use of metformin when not contraindicated as the first therapy after lifestyle modifications.3, 4 The approved drug classes are metformin (alone in the biguanide class), sulfonylureas, thiazolidinediones, dipeptidyl peptidase- 4 (DPP-4) inhibitors, glucagon-like peptide-1 (GLP-1) agonists, and sodium-glucose cotransporter-2 (SGLT-2) inhibitors. Clinical guidelines, including those of the American Diabetes Association, recommend monitoring hemoglobin A1c (HbA1c) to determine the need for changing the medication dose or adding another agent to improve glycemic control.4 Clinicians also monitor other intermediate outcomes, including body weight, and short-term and long-term safety and adverse effects of the drugs, which vary by drug class, with the goal of improving long-term clinical outcomes. The Effective Health Care Program of the Agency for Healthcare Research and Quality (AHRQ) has published two prior systematic reviews comparing monotherapies and medication combinations for adults with type 2 diabetes.6, 7 Since January 2010, the month of the last publications included in the past review, the FDA has approved one new medication class (SGLT-2 inhibitors, with 3 new medications) and several new DPP-4 inhibitors and GLP-1 receptor agonists. Additional data on previously approved medications have also emerged, which could change the balance of benefit and risk attributable to these drugs or could alter the strength of evidence about some of the drug comparisons previously reviewed.8-11 Given the ever- increasing literature about type 2 diabetes medications and the recent approval of many new medications, an updated systematic review evaluating the effects of these medications on intermediate and long-term effectiveness and safety outcomes will be valuable to clinicians, patients, investigators, guideline developers, and payers. Scope and Key Questions This review updates the 2011 review on oral diabetes medications for adults with type 2 diabetes.7 We are focusing on priority head-to-head drug class comparisons identified, a priori, as clinically relevant comparisons for which there are evidence gaps (Table A). Given the unique and emerging potential benefits and harms of some of these medications, we have included additional intermediate and safety outcomes in the review: for studies including either SGLT-2 inhibitors or GLP-1 receptor agonists, systolic blood pressure and heart rate, and for studies that
- 26. ES-2 include a comparisonwith SGLT-2 inhibitors, impaired renal function, urinary tract infections, genital infections, volume depletion, and bone fractures. The Key Questions that we address in this review are as follows: Key Question 1a: In adults ages 18 or older with type 2 diabetes mellitus, what is the comparative effectiveness of the specified monotherapy FDA-approved diabetes medications for the intermediate outcomes of HbA1c, weight, systolic blood pressure (for comparisons including SGLT-2 inhibitors or GLP-1 receptor agonists), and heart rate (for comparisons including SGLT- 2 inhibitors or GLP-1 receptor agonists)? Key Question 1b: In adults ages 18 or older with type 2 diabetes mellitus, what is the comparative effectiveness of metformin-based combinations of FDA-approved diabetes medications for the intermediate outcomes of HbA1c, weight, systolic blood pressure (for comparisons including SGLT-2 inhibitors or GLP-1 receptor agonists), and heart rate (for comparisons including SGLT-2 inhibitors or GLP-1 receptor agonists)? Key Question 2a: In adults ages 18 or older with type 2 diabetes mellitus, what is the comparative effectiveness of the monotherapy FDA-approved diabetes medications for the long- term clinical outcomes of all-cause mortality, cardiovascular and cerebrovascular morbidity and mortality, retinopathy, nephropathy, and neuropathy? Key Question 2b: In adults ages 18 or older with type 2 diabetes mellitus, what is the comparative effectiveness of the metformin-based combinations of FDA-approved diabetes medications for the long-term clinical outcomes of all-cause mortality, cardiovascular and cerebrovascular morbidity and mortality, retinopathy, nephropathy, and neuropathy? Key Question 3a: In adults ages 18 or older with type 2 diabetes mellitus, what is the comparative safety of the monotherapy FDA-approved diabetes medications regarding liver injury, lactic acidosis, pancreatitis, hypoglycemia, congestive heart failure, cancer, severe allergic reactions, macular edema or decreased vision, and gastrointestinal side effects; and for comparisons including SGLT-2 inhibitors, what is the comparative safety regarding urinary tract infections, impaired renal function, genital mycotic infections, fracture, and volume depletion? Key Question 3b: In adults ages 18 or older with type 2 diabetes mellitus, what is the comparative safety of metformin-based combinations of FDA-approved diabetes medications regarding liver injury, lactic acidosis, pancreatitis, hypoglycemia, congestive heart failure, cancer, severe allergic reactions, macular edema or decreased vision, and gastrointestinal side effects; and for comparisons including SGLT-2 inhibitors, what is the comparative safety regarding urinary tract infections, impaired renal function, genital mycotic infections, fracture, and volume depletion? Key Question 4: Do the comparative safety and effectiveness of these treatments differ across subgroups defined by the age, sex, race/ethnicity, and body mass index of adults with type 2 diabetes?
- 27. ES-3 Table A. Prioritymedication comparisons included for each Key Question Intervention Main Intervention Class (Generic Individual Drug Names) Comparisons Monotherapy as main intervention Biguanides (metformin) Thiazolidinediones* Sulfonylureas † DPP-4 inhibitors SGLT-2 inhibitors GLP-1 receptor agonists ‡ Combination of metformin plus thiazolidinedione Combination of metformin plus sulfonylurea Combination of metformin plus DPP-4 inhibitor Combination of metformin plus SGLT-2 inhibitor Combination of metformin plus GLP-1 receptor agonist Thiazolidinediones (rosiglitazone or pioglitazone) Sulfonylureas DPP-4 inhibitors SGLT-2 inhibitors GLP-1 receptor agonists Sulfonylureas (glimepiride, glyburide, ¶ glibenclamide, ¶ or glipizide) DPP-4 inhibitors SGLT-2 inhibitors GLP-1 receptor agonists DPP-4 inhibitors (alogliptin, linagliptin, saxagliptin, or sitagliptin) SGLT-2 inhibitors GLP-1 receptor agonists SGLT-2 inhibitors (canagliflozin, dapagliflozin, or empagliflozin) GLP-1 receptor agonists Combination therapy as main intervention Combination of metformin plus thiazolidinedione or sulfonylurea or DPP-4 inhibitor or SGLT-2 inhibitor or GLP-1 receptor agonist or basal insulin Combination of metformin plus sulfonylurea or DPP-4 inhibitor or SGLT-2 inhibitor or GLP-1 receptor agonist or basal insulin ‡ or premixed insulin ‡ DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; HbA1c = hemoglobin A1c; SGLT-2 = sodium-glucose cotransporter 2. *For studies comparing thiazolidinediones with metformin, we reviewed only HbA1c, long-term outcomes, and selected safety outcomes, given the high strength of evidence from our prior Comparative Effectiveness Review for other outcomes (specifically fracture and weight).7 † For studies comparing sulfonylureas with metformin, we reviewed only long-term outcomes and cancer, given the high strength of evidence on the other outcomes from our prior Comparative Effectiveness Review.7 ‡ The generic individual drug names for the GLP-1 receptor agonists are exenatide, liraglutide, dulaglutide, and albiglutide. The generic individual drug names for basal insulin are insulin glargine, insulin detemir, and neutral protamine Hagedorn (NPH) insulin. The generic individual drug names for premixed insulin are NPH/regular 50/50, NPH/regular 70/30, insulin lispro 50/50, insulin lispro 75/25, and insulin aspart 70/30. ¶ Glyburide and glibenclamide are the same drug. Methods Topic Refinement and Review Protocol This review updates the 2011 Comparative Effectiveness Review on diabetes medications for adults with type 2 diabetes.7 We recruited a Technical Expert Panel (TEP) to review a draft of the protocol and a summary of the revisions from the 2011 review. The TEP included endocrinologists, general internists, biostatisticians, and representatives from government agencies. The TEP reviewed our protocol and provided feedback on the proposed methods for addressing the Key Questions. With the feedback from the TEP and the AHRQ representatives, we finalized and posted the protocol (www.effectivehealthcare.ahrq.gov).
- 28. ES-4 Literature Search Strategy SearchStrategy We searched MEDLINE® , Embase® , and the Cochrane Central Register of Controlled Trials (CENTRAL). We ran the search developed for the 2011 review with the date restrictions of April 2009 through April 2015. (See Appendix A.) The expanded search included medical subject headings (MeSH) and text words for all of the new medications included in this updated report, without date restrictions. Additionally, we searched ClinicalTrials.gov to identify relevant registered trials. We reviewed the FDA Web site for any unpublished additional studies relevant to the topic as part of our gray literature search. We also provided an opportunity for manufacturers of interventions to submit unpublished data. Study Selection Two independent reviewers conducted title scans and advanced articles if either one thought them relevant. The abstract review phase was designed to identify studies reporting the effectiveness or safety of the medications and medication combinations of interest. Two investigators independently reviewed abstracts. Differences between investigators regarding the inclusion or exclusion of abstracts were resolved through consensus adjudication. Full articles underwent another independent parallel review regarding their appropriateness for inclusion. Selection criteria for studies are provided in Table B.
- 29. ES-5 Table B. Studyinclusion criteria PICOTS Inclusion Criteria Population We included studies of adult humans with type 2 diabetes, non–insulin-dependent diabetes mellitus, or adult-onset diabetes. Interventions We included studies that evaluated a diabetes medication of interest or drug combination of interest. (See Table A.) Comparisons We included studies that evaluated a comparison of interest. (See Table A.) Outcomes* We included studies addressing the following intermediate outcomes for KQ1: Hemoglobin A1c Weight Systolic blood pressure Heart rate We included studies addressing the following microvascular, macrovascular, and mortality outcomes for KQ2: All-cause mortality Cardiovascular and cerebrovascular morbidity and mortality Retinopathy Nephropathy Neuropathy We included studies addressing the following safety outcomes for KQ3: Liver injury Impaired renal function Lactic acidosis Pancreatitis Hypoglycemia Gastrointestinal side effects Congestive heart failure Cancer Macular edema or decreased vision Fractures Urinary tract infections Genital mycotic infections Volume depletion KQ4 included studies considering any of the above outcomes. Type of study For KQ1, we included only RCTs. For KQ2 and KQ3, we included RCTs, nonrandomized experimental studies with a comparison group, and high-quality observational studies with a comparison group. We included randomized trials that used a crossover design, with some exceptions. Only studies published in English were included. Timing and setting We included studies in which the observed intervention or exposure period was more than 3 months. KQ = Key Question; PICOTS = populations, interventions, comparisons, outcomes, timing, and settings; RCT = randomized controlled trial. *Not every outcome was assessed for each comparison. Data Extraction Reviewers extracted information on the general study characteristics, study participant characteristics, interventions, comparisons, method of ascertainment of safety outcomes, and outcome results, including measures of variability. We also collected data on outcomes for the subgroups of interest, which were defined by age, sex, race/ethnicity, and body mass index. Risk-of-Bias Assessment of Individual Studies Two independent reviewers assessed risk of bias. We assessed the risk of bias in individual randomized controlled trials (RCTs) using the Jadad criteria, consistent with the prior report.12 We used the Downs and Black tool for assessment of internal validity for nonrandomized trials and observational studies.13 We included only medium- or high-quality observational studies, as
- 30. ES-6 determined by assessmentof each study’s risk of bias. The Downs and Black tool was also applied to the observational studies that had been included in the prior report;7 some of the previously included observational studies were excluded owing to methodological deficiencies. Data Synthesis For each Key Question, we created a set of detailed evidence tables containing all information extracted from eligible studies, including those from the prior Comparative Effectiveness Reviews. We conducted meta-analyses when there were sufficient data (at least 3 trials) and studies were sufficiently homogeneous with respect to key variables (population characteristics, study duration, and drug dose). We included in the quantitative pooling those study arms with drug doses and study durations most commonly reported. We tested the heterogeneity among the trials considered for quantitative pooling using a chi-squared test with a significance level of alpha less than or equal to 0.10, and we also examined heterogeneity among studies with an I2 statistic.14 We pooled the mean difference between groups using a random- effects model with the DerSimonian and Laird formula in settings of low heterogeneity (I2 <50%)15 or the profile likelihood estimate when statistical heterogeneity was high.16 For dichotomous outcomes, we calculated pooled odds ratios using a random-effects model with the DerSimonian and Laird formula in settings of low heterogeneity15 or the profile likelihood estimate in settings of high heterogeneity (I2 >50%).16 Sensitivity analyses included sequential study elimination to assess for influential studies. Stratification and metaregression (only if 10 or more studies were included in the meta-analysis) were done to identify and describe sources of heterogeneity and their effects on outcomes when substantial heterogeneity was identified. Strength of the Body of Evidence At the completion of the review, two reviewers sequentially graded the evidence addressing the Key Questions by adapting an evidence grading scheme recommended in the Methods Guide for Effectiveness and Comparative Effectiveness Reviews.17 We generated evidence grades about each intervention comparison for each outcome (Table A) for which there was at least one RCT or three observational studies. We graded the evidence separately for the RCTs and the observational studies.17 The final evidence grade and conclusion were typically based on the RCT grade and could be strengthened by evidence from the observational studies. We separately assessed the strength of evidence for shorter and longer studies (≥2 years); however, we assessed strength of evidence only for longer studies from which we could draw a conclusion. We assessed the study limitations, consistency, directness, precision, and reporting bias. If we conducted a meta-analysis for a body of evidence, we relied on the results of the meta- analysis to rate precision and used the designated minimally important differences as a point of reference for precision. (See full report for details.) We classified evidence pertaining to the Key Questions into four categories: (1) high grade (indicating high confidence that the evidence reflects the true effect, and further research is very unlikely to change our confidence in the estimate of the effect); (2) moderate grade (indicating moderate confidence that the evidence reflects the true effect, but further research could change our confidence in the estimate of the effect and may change the estimate); (3) low grade (indicating low confidence that the evidence reflects the true effect, and further research is likely to change our confidence in the estimate of the effect and is likely to change the estimate); and (4) insufficient (indicating evidence is unavailable or the body of evidence has unacceptable deficiencies, precluding reaching a conclusion).
- 31. ES-7 Applicability We assessed theapplicability of the evidence in terms of the degree to which the study populations, interventions, outcomes, timing, and settings were typical of the treatment of individuals with type 2 diabetes who are receiving treatment in a usual care setting, such as outpatient treatment by internists, family physicians, and endocrinologists. Results In this Executive Summary, results are presented by Key Question and focus on moderate- or high-strength evidence. We also highlight some key areas for which there was low-strength or insufficient evidence. The full results of this synthesis, including detailed results on all evidence, are in the full report. Results of Literature Searches We included 166 publications in our previous review. After excluding studies that no longer had a comparison or an outcome of interest and cohort studies that did not meet our quality criteria, we included 105 of these studies from the prior review (published in 107 articles) in the update. We also retrieved 19,171 unique citations from our updated literature search. After reviewing titles, abstracts, and full text, we included 114 new studies (published in 142 new articles). Ten of the new publications were either extensions or additional analyses of studies included in the previous review. Overall, we included 219 studies, published in 249 articles. Study Duration for All Key Questions (KQ1–KQ4) Of the 177 included RCTs for all Key Questions combined, most studies were less than 1 year in duration (Figure A). Only 4 percent of studies lasted longer than 2 years, making it difficult to draw any firm conclusions about long-term outcomes. Unless stated otherwise in the text or figures, results and conclusions for all the Key Questions are for short-term outcomes. Followup among the 25 observational studies lasted between 3 months and 8 years. Five of the included observational studies lasted 1 year or less. Most (64%) of the cohorts had at least 2 years of followup.
- 32. ES-8 Figure A. Durationof followup for randomized controlled trials comparing the effects of diabetes medications among adults with type 2 diabetes (N = 177) Key Questions 1a and 1b: Intermediate Outcomes Of the 162 RCTs (reported in 189 articles) identified for Key Question 1, 81 percent were less than 1 year long. Only 12 percent of these trials reported having received no industry support, and 14 percent did not report on this at all. Study participants were generally overweight or obese and had a baseline HbA1c between 7 and 9 percent. The exclusion criteria were generally similar for most trials: significant renal, cardiovascular, and hepatic disease. About half of the trials (58%) excluded older subjects (generally older than 75 to 80 years of age). Almost all of the studies reported a diverse male-female mix among the participants. Of the few studies that evaluated longer timeframes (>2 years), most were consistent with the shorter term results. While an occasional longer study conflicted with the shorter study results, the high losses to followup (generally >20%) and frequent use of last observation carried forward analyses made it difficult to draw conclusions about longer term effects. Therefore, results discussed here are for the short term unless otherwise specified in the figures or text. Hemoglobin A1c We found that most diabetes medications as monotherapy (metformin, thiazolidinediones, and sulfonylureas) reduced HbA1c to a similar degree in the short term (Figure B). In the 2011 report,7 the evidence on metformin versus sulfonylurea, which showed no significant between- group differences in HbA1c, was graded as high; therefore, the comparison was not updated in this report. In this report, metformin was more effective in reducing HbA1c than the DPP-4 inhibitors as monotherapy by about 0.4 percent. (All differences for HbA1c represent absolute
- 33. ES-9 percentage points.) Two-drugcombination therapies with metformin (such as metformin plus thiazolidinediones, metformin plus sulfonylureas, metformin plus SGLT-2 inhibitors, and metformin plus DPP-4 inhibitors) were generally more effective in reducing HbA1c than metformin monotherapy by about 1 percent (Figure B). For the combination comparisons, metformin plus a GLP-1 receptor agonist reduced HbA1c more than metformin plus DPP-4 inhibitors by 0.65 percent. Otherwise, most combination therapy comparisons with moderate strength of evidence had either no significant or no clinically meaningful between-group differences (<0.3%) in HbA1c between arms (Figure B). Although we included comparisons with the GLP-1 receptor agonists, we graded the evidence for most of these comparisons as insufficient or low; therefore, we were limited in our ability to draw conclusions about their effectiveness. Despite the clinical interest in comparing metformin plus injectables, there was insufficient or low strength of evidence on glycemic control for the following comparisons: metformin plus the GLP-1 receptor agonists versus metformin plus basal or premixed insulin, and metformin plus premixed insulin versus metformin plus basal insulin. Figure B. Pooled between-group differences in hemoglobin A1c and strength of evidence for monotherapy and metformin-based combination comparisons BL = baseline; CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; ES = effect size (mean between-group difference in HbA1c); GLP1 = glucagon-like peptide-1 agonists; H = high; HbA1c = hemoglobin A1c; M = moderate; Met = metformin; PL = profile likelihood estimate; SGLT2 = sodium-glucose cotransporter-2 inhibitors; SOE = strength of evidence; SU = sulfonylurea; TZD = thiazolidinedione. The width of the horizontal lines represents the 95% confidence intervals for each pooled analysis. Drug 1 is the reference group.
- 34. ES-10 Weight Monotherapy and combinationmedication comparisons generally showed significant between-group differences when comparing medications expected to increase weight (sulfonylureas, thiazolidinediones, and insulin) with medications expected to maintain or decrease weight (metformin, DPP-4 inhibitors, GLP-1 receptor agonists, and SGLT-2 inhibitors). Figure C shows the data from the meta-analyses that could feasibly be conducted. We report between-group differences in the text regarding results where meta-analyses could not be done. DPP-4 inhibitors and GLP-1 receptor agonists both decreased weight more than thiazolidinediones (between-group differences ranging from -2.3 kg to -3.5 kg). In the 2011 report, comparisons of metformin versus thiazolidinedione and metformin versus sulfonylurea were found to favor metformin by about -2.5 kg, with high strength of evidence; therefore, these comparisons were not updated. In this report, several monotherapy and metformin-based combination medications were compared where both arms had medications expected to maintain or decrease weight, or both arms had medications expected to increase weight, with varying effects. Metformin decreased weight more than DPP-4 inhibitors, whereas sulfonylureas caused slightly less weight gain than thiazolidinediones (Figure C). There was moderate strength of evidence that SGLT-2 inhibitors decreased weight more than metformin and more than DPP-4 inhibitors (between-group differences ranging from -1.3 kg to -2.7 kg). The combinations of metformin plus a GLP-1 receptor agonist (Figure C) and metformin plus an SGLT-2 inhibitor (range in between-group differences of -1.8 to -3.6 kg) were both favored over the combination of metformin plus a DPP- 4 inhibitor. Metformin plus a sulfonylurea had more favorable weight effects than the combination of metformin plus a premixed or basal insulin (range in mean between-group differences of -0.5 kg to -1.7 kg), with moderate strength of evidence. Despite the clinical interest in comparing metformin plus injectables, there was low strength of evidence about weight for the following comparisons: metformin plus the GLP-1 receptor agonists versus metformin plus basal or premixed insulin, and metformin plus premixed insulin versus metformin plus basal insulin.
- 35. ES-11 Figure C. Pooledbetween-group differences in weight and strength of evidence for monotherapy and metformin-based combination comparisons BL = baseline; CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; ES = effect size (mean between-group difference in weight); GLP1 = glucagon-like peptide-1 agonists; H = high; M = moderate; Met = metformin; PL = profile likelihood estimate; SGLT2 = sodium-glucose cotransporter-2 inhibitors; SOE = strength of evidence; SU = sulfonylurea; TZD = thiazolidinedione. The width of the horizontal lines represents the 95% confidence intervals for each pooled analysis. Drug 1 is the reference group. Systolic Blood Pressure and Heart Rate Systolic blood pressure and heart rate were evaluated only for the newer medications, SGLT- 2 inhibitors and GLP-1 receptor agonists, owing to the suspected effects of these newer medications on these clinical outcomes based on prior literature.18, 19 The SGLT-2 inhibitors consistently reduced systolic blood pressure by 3 to 5 mmHg in all comparisons for which there were sufficient numbers of studies (Table C). Also, metformin plus a GLP-1 receptor agonist yielded a greater reduction in systolic blood pressure, about 3 mmHg, compared with metformin alone (Table C). For heart rate, only two comparisons had sufficient data to grade the evidence as more than insufficient or low. These comparisons had no or small differences (<2 beats per minute) between groups (Table C). When there were differences in outcomes among comparisons rated as having low strength of evidence, they were less than three beats per minute.
- 36. ES-12 Table C. Summaryof the moderate- to high-strength evidence on the comparative effectiveness and safety of diabetes medications as monotherapy and metformin-based combination therapy for systolic blood pressure and heart rate Outcome Conclusions Strength of Evidence Systolic blood pressure Metformin plus an SGLT-2 inhibitor reduced systolic blood pressure more than— Metformin alone: pooled between-group difference for shorter studies, 4.4 mmHg (95% CI, 2.9 to 6.0 mmHg) Metformin plus SU: pooled between-group difference, 5.1 mmHg (95% CI, 4.2 mmHg to 6.0 mmHg) High Metformin plus an SGLT-2 inhibitor reduced systolic blood pressure more than metformin plus a DPP-4 inhibitor: pooled between-group difference, 4.1 mmHg (95% CI, 3.6 mmHg to 4.6 mmHg). Moderate SGLT-2 inhibitors reduced systolic blood pressure more than metformin: pooled between-group difference, 2.8 mmHg (95% CI, 2.6 mmHg to 3.0 mmHg). Moderate Metformin plus a GLP-1 receptor agonist reduced systolic blood pressure more than metformin: pooled between-group difference, 3.1 mmHg (95% CI, 1.4 to 4.9 mmHg). Moderate Heart rate Increases in heart rate were minimal and similar for metformin and GLP-1 receptor agonist monotherapy. Moderate Combination therapy with metformin plus an SGLT-2 inhibitor resulted in less increase in heart rate than metformin plus an SU: pooled between-group difference in heart rate, 1.5 bpm; 95% CI, 0.6 bpm to 2.3 bpm. Moderate bpm = beats per minute; CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; SGLT-2 = sodium-glucose cotransporter-2; SU = sulfonylurea. Key Questions 2a and 2b: All-Cause Mortality and Macrovascular and Microvascular Outcomes Of 118 studies (reported in 141 publications) identified for Key Question 2, 96 were RCTs and 21 were observational (mainly retrospective cohort) studies. Most studies evaluated all-cause or cardiovascular mortality or cardiovascular morbidity. Of the 96 trials, 33 were at least 1 year in duration. Only 11 had 2 years or more of followup time, and 10 of these had over 20-percent losses to followup. No trial specified mortality or a macrovascular or microvascular outcome as its primary outcome. Mean/median followup of the observational studies ranged from 6 months to 5 years, with 12 lasting at least 2 years. Seven of the observational studies were designed to evaluate cardiovascular outcomes. Because of low event rates and sample size, the pooled studies for most comparisons on these outcomes were underpowered. All-Cause Mortality, Cardiovascular Mortality, and Cardiovascular Morbidity Only one comparison had moderate strength of evidence for any of these outcomes. The rest of the outcomes were rated as low strength of evidence or insufficient. We found moderate strength of evidence that sulfonylurea monotherapy was associated with a 50-percent to 70- percent higher relative risk (absolute risk difference, 0.1% to 2.9% in RCTs; number needed to treat, 20 to 1,000) of cardiovascular mortality compared with metformin monotherapy (Table D). This conclusion was supported by consistent findings from two high-quality RCTs (N = 4,664), with a range in mean/median followup of 2.8 to 4.0 years, and three high-quality observational studies (N =115,105) that used propensity score methodology (2 studies) and multivariate regression (1 study) to account for confounding. Our findings on all cause-mortality and cardiovascular morbidity, drawn from the same RCTs plus additional observational studies (noted in Table D), also favored metformin over sulfonylureas; however, the strength of
- 37. ES-13 evidence was lowfor these outcomes because of less consistency in results across studies. It is of note that losses to followup were greater than 20 percent in both RCTs. Losses to followup were the same (20%) across arms in the study by Hong and colleagues (2013) and therefore not anticipated to bias the comparison of arms.20 In A Diabetes Outcome Progression Trial (ADOPT), losses to followup were higher in the sulfonylurea (44%) than the metformin (38%) arm, with median followup of 3.3 years for the sulfonylurea arm versus 4.0 years for the metformin arm.21 Therefore, study results were likely biased to the null, lending further support to the inference that metformin was favored over sulfonylurea monotherapy. Table D. Comparative effectiveness of sulfonylureas compared with metformin for long-term all- cause mortality and cardiovascular mortality and morbidity—moderate strength of evidence or consistent low-strength evidence Outcome Range in RR From RCTs Range in RD From RCTs Adjusted HR From Observational Studies SOE All-cause mortality 1.0 to 2.1 (N = 2) 0.1% to 5.0% (N = 2) 1.2 to 1.9 (N = 7*) Low CVD mortality 1.5 to 1.7 (N = 2) 0.1% to 2.9% (N = 2) 1.1 to 1.6 (N = 3) Moderate CVD morbidity 0.7 to 1.4 (N = 2) -10.1% to 0.4% (N = 2) 1.1 to 3.3 (N = 5 † ) Low CVD = cardiovascular disease; HR = hazard ratio; RCT = randomized controlled trial; RD = risk difference; RR = relative risk; SOE = strength of evidence. *One additional retrospective cohort study reported an odds ratio of 1.1. †Additionally, 1 case-control study reported an odds ratio of 1.2. Retinopathy, Nephropathy, and Neuropathy While we found more evidence than in the prior report, there were still too few studies to reach firm conclusions; all evidence for these outcomes was of low strength or insufficient. Key Questions 3a and 3b: Comparative Safety Of 145 studies identified for Key Question 3, 137 were RCTs and 8 were observational (mainly retrospective cohort) studies. Most RCTs lasted a year or less, with only about 5 percent lasting more than 2 years. Mean or median followup of the eight observational studies ranged from 3 months to 5 years. The few longer studies were generally consistent with the shorter term results; however, the losses to followup were often high (>20% in the majority of the longer studies), making it difficult to draw firm long-term conclusions. Therefore, most safety comparisons represent shorter term results unless specifically stated in the text or a figure. Hypoglycemia Sulfonylureas alone and in combination with metformin had a higher risk of mild, moderate, or total hypoglycemia than any other monotherapies and metformin-based combinations for which we identified evidence (Figure D). While studies were too heterogeneous for a meta- analysis, sulfonylureas also had greater risk of hypoglycemia than GLP-1 receptor agonists (range in odds ratio [OR], 3.1 to 5.3; range in risk difference [RD], 12% to 21%) and DPP-4 inhibitors (range in OR, 3.8 to 12.4; range in RD, 6% to 15%), with moderate strength of evidence. In addition to the increased risk of hypoglycemia with metformin plus sulfonylurea versus several comparators (Figure D), the combination of metformin plus sulfonylurea also had greater risk of hypoglycemia compared with metformin monotherapy (range in OR, 2 to 17; range in RD, 0% to 35%) and compared with the combination of metformin plus a GLP-1 receptor agonist (for studies lasting 104 to 234 weeks: range in OR, 3.4 to 7.1; range in RD, 15% to 30%). When compared with metformin plus a basal or premixed insulin, metformin plus a
- 38. ES-14 GLP-1 receptor agonisthad less hypoglycemia risk (range in OR, 0.18 to 0.35; range in RD, -3% to -13%), with moderate strength of evidence. The combination of metformin plus basal insulin had a lower risk of hypoglycemia than the combination of metformin plus premixed insulin (range in OR, 0.23 to 0.89; range in RD, -5% to -28%), with moderate strength of evidence. We did not pool these studies owing to high heterogeneity. We found moderate strength of evidence that sulfonylureas had an increased risk of severe hypoglycemia compared with metformin or thiazolidinedione monotherapy (range in OR, 1.4 to 8; range in RD, 0.5% to 23%). Similarly, sulfonylureas in combination with metformin had a greater risk of severe hypoglycemia than the combination of metformin plus DPP-4 inhibitors (range in OR, 6 to 14; range in RD, 0% to 3%) or metformin plus SGLT-2 inhibitors (OR, 7; range in RD, 1% to 3%), with moderate strength of evidence for both comparisons. Figure D. Pooled odds ratios of mild/moderate hypoglycemia and strength of evidence for monotherapy and metformin-based combination comparisons CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; H = high; M = moderate; Met = metformin; OR = odds ratio; PL = profile likelihood estimate; RD = absolute risk difference; SGLT2 = sodium-glucose cotransporter-2 inhibitors; SOE = strength of evidence; SU = sulfonylurea; TZD = thiazolidinediones. The width of the horizontal lines represents the 95% confidence intervals for each pooled analysis. Drug 1 is the reference group. Gastrointestinal Side Effects Metformin and GLP-1 receptor agonists were associated with more gastrointestinal side effects (typically nausea, vomiting, or diarrhea) than any other medications with sufficient studies for comparison, regardless of whether they were used as monotherapy or in combination (Figure E). Although there were insufficient studies for a meta-analysis, GLP-1 receptor agonists
- 39. ES-15 had greater gastrointestinalside effects than sulfonylureas, with moderate strength of evidence (range in OR, 1.4 to 2.4; range in RD, 3% to 9%). Metformin plus a GLP-1 receptor agonist had more gastrointestinal side effects than metformin plus DPP-4 inhibitors (range in OR, 1.0 to 7.7; range in RD, 0% to 23%) and metformin plus thiazolidinediones (range in OR, 2.9 to 6.3; range in RD, 8% to 19%), with moderate strength of evidence. Nausea and vomiting were more common with GLP-1 receptor agonists than with metformin (Figure E), but rates of diarrhea were similar between the groups. The rates of gastrointestinal side effects were similar for metformin monotherapy compared with metformin plus a DPP-4 inhibitor or metformin plus SGLT-2 inhibitors (Figure E). We found high strength of evidence that the rates of gastrointestinal adverse events were similar for thiazolidinediones (range, 2% to 9%) and sulfonylureas (range, 3% to 10%), with a range in RD of -1.2% to 1.7%. The combination of metformin plus a sulfonylurea (range, 1% to 18%) was also similar to the combination of metformin plus a thiazolidinedione (range, 1% to 13%), with a range in RD of -5.0% to 2.1% (moderate strength of evidence).
- 40. ES-16 Figure E. Pooledodds ratios of gastrointestinal adverse events and strength of evidence for monotherapy and metformin-based combination comparisons * CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; GI = gastrointestinal; GLP-1 = glucagon-like peptide-1 receptor agonists; H = high; M = moderate; Met = metformin; OR = odds ratio; RD = absolute risk difference; SGLT-2 = sodium-glucose cotransporter-2 inhibitors; SOE = strength of evidence; SU = sulfonylurea; TZD = thiazolidinediones. The width of the horizontal lines represents the 95% confidence intervals for each pooled analysis. Drug 1 is the reference group. * All results presented in this graph are based on short-term (less than 52 weeks) studies unless otherwise specified. † Based on studies with 104 weeks of followup. Congestive Heart Failure There was only one long-term trial, which lasted 4 years, and only a few observational studies of medium quality with 6 to 8 years of followup that allow an assessment of the comparative safety of diabetes medications regarding congestive heart failure. We found low strength of evidence that the risk of congestive heart failure was 1.2 to 1.6 times as great with thiazolidinediones as with sulfonylureas (pooled OR, 1.6; 95% CI, 0.96 to 2.8; range in RD, 0% to 2%) or metformin (2 RCTs lasting less than a year with no events; 1 4-year RCT with an RD of 3%; and range in hazard ratio of 1.2 to 1.5 in 2 observational studies). Despite recent concerns about congestive heart failure with specific DPP-4 inhibitors, we found low or insufficient strength of evidence on the comparative safety of this drug class for this outcome in studies lasting less than 2 years (5 RCTs reporting no events in the DPP-4 inhibitor arms; 1 RCT with 1 event in the metformin plus DPP-4 inhibitor arm and none in the comparator arm; and 1 RCT of metformin plus DPP-4 inhibitor vs. metformin plus sulfonylurea reporting fewer events in the DPP-4 combination arm compared with the sulfonylurea combination arm [3 vs. 6 events]).
- 41. ES-17 Cancer Evidence was generallylacking or of low strength for cancer outcomes. We found low strength of evidence that the combination of metformin plus a sulfonylurea was favored over the combination of metformin plus a DPP-4 inhibitor for cancer risk (3 RCTs with 104 weeks of followup). An unpublished study (104 weeks of followup) and an unpublished longer term (156 weeks) followup of one of the included published studies22 were consistent with this finding and might have increased the evidence to moderate strength had they been included. A recent RCT with only 52 weeks of followup also found a higher risk of cancer in the DPP-4 inhibitor combination arm compared with the sulfonylurea combination arm.23 Adverse Events Specific to SGLT-2 Inhibitors We evaluated the comparative effectiveness of SGLT-2 inhibitors for specific adverse events of interest: urinary tract infections, genital mycotic infections, renal function impairment, fractures, and volume depletion. We found high strength of evidence that the combination of metformin plus an SGLT-2 inhibitor increased the odds of a genital mycotic infection approximately threefold compared with metformin monotherapy and sixfold compared with the combination of metformin plus a sulfonylurea (Table E). We also found moderate strength of evidence that SGLT-2 inhibitors increased the odds of genital mycotic infection fourfold compared with metformin monotherapy. The evidence was of low strength or insufficient for the other safety outcomes specific to SGLT-2 inhibitors. Other Outcomes The evidence on the outcomes of liver injury, pancreatitis, lactic acidosis, severe allergic reactions, and macular edema and decreased vision was of low strength or insufficient. We could not make any conclusions about these outcomes. Table E. Summary of the moderate- to high-strength evidence on the comparative safety of diabetes medications as monotherapy and metformin-based combination therapy for genital mycotic infections Conclusions Strength of Evidence The rates of genital mycotic infections were higher with metformin plus SGLT-2 inhibitors compared with— Metformin monotherapy: o Pooled OR, 3.0; 95% CI, 1.2 to 7.2 for females o Pooled OR, 2.7; 95% CI, 0.8 to 9.0 for males o Range in between-group risk difference, -2.3% to 9.9% Metformin plus SU: o Pooled OR, 5.2; 95% CI, 3.4 to 8.0 for females o Pooled OR, 7.6; 95% CI, 4.0 to 14.4 for males o Range in between-group risk difference, 7.1% to 17.4% High The rates of genital mycotic infections were higher with SGLT-2 inhibitors compared with metformin monotherapy o Pooled OR, 4.1; 95% CI, 2.0 to 8.3 o Range in between-group risk difference, -0.04% to 15.7% Moderate The rates of genital mycotic infections were higher with metformin plus SGLT-2 inhibitors compared with metformin plus DPP-4 inhibitors Range in between-group risk difference, -2.8% to 8.8% Moderate CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; OR=odds ratio; SGLT-2 = sodium-glucose cotransporter-2; SU = sulfonylurea.
- 42. ES-18 Key Question 4:Subgroups We found little evidence on the comparative effectiveness and safety of diabetes medications in predefined subgroups of age, sex, race/ethnicity, or body mass index. Most of the evidence on subgroups was for the outcome of HbA1c and did not show differential effects of the included comparisons by age, sex, race/ethnicity, or body mass index. Discussion Key Findings in Context Intermediate Outcomes This report builds on prior work by adding more information for HbA1c and weight regarding the metformin-based combination comparisons and comparisons with the newer medications. It also adds new comparative information for the SGLT-2 inhibitors and GLP-1 agonists on both heart rate and blood pressure. While there is controversy about HbA1c targets, better glycemic control (measured by HbA1c levels) is strongly associated with lower risk of microvascular disease,24-26 making it a good proximal outcome to measure. Consistent with the 2011 Comparative Effectiveness Review, most monotherapies were found to be similarly effective in reducing HbA1c, with the exception of DPP-4 inhibitors, which had a smaller effect relative to metformin (Figure B).7 While metformin versus GLP-1 receptor agonists and metformin versus SGLT-2 inhibitors also showed no clear between-group differences in HbA1c, the evidence was graded as low strength because the three studies in each comparison were imprecise and inconsistent. In this update, we found inconsistent findings in the studies of GLP-1 receptor agonists. It may be that the individual GLP-1 receptor agonists have different effects on HbA1c. A 2011 Cochrane systematic review showed small between-group differences in HbA1c, around 0.3 percent, favoring liraglutide and weekly exenatide over daily exenatide.19 Combination therapy with metformin generally reduced HbA1c by 0.7 to 1 absolute percentage points compared with metformin monotherapy. While we found moderate strength of evidence that some combination comparisons were more effective than others, most between- group differences were small (<0.3 percentage points), with questionable clinical relevance. Only one combination comparison with moderate strength of evidence was favored by greater than 0.3 percentage points over any other combination comparison: the combination of metformin plus a GLP-1 receptor agonist reduced HbA1c more than metformin plus a DPP-4 inhibitor by 0.65 percentage points. Two prior network meta-analyses27, 28 showed that most metformin combination comparisons had similar reductions in HbA1c. However, the results of the direct comparisons evaluated in this report are more precise, allowing us to detect smaller between- group differences than the indirect comparisons in the network meta-analyses. Weight gain was small to moderate in the trials in which participants gained weight; even in the longest trials, weight gain was less than 5 kg. However, even small to moderate weight gain (5% to 10% of body weight) may be associated with increased insulin resistance.29 In addition, weight loss and glycemic control were reported as the primary drivers of patient preferences for diabetes medications when compared with treatment burden and side effects in a recent systematic review.30 Drug effects on weight, therefore, have a strong impact on the choice of the drug for second-line combination therapy in a patient not well controlled on a single agent. Our
- 43. ES-19 systematic review buildson prior work by adding more direct comparative data about metformin combination comparisons that further confirm the known weight effects of the individual medications. As monotherapy and in combination with metformin, thiazolidinediones, sulfonylureas, and insulin are associated with weight gain, DPP-4 inhibitors with weight maintenance, and SGLT-2 inhibitors and GLP-1 receptor agonists with weight loss.7, 18, 19, 31 We evaluated systolic blood pressure and heart rate for the newer classes of medications, the SGLT-2 inhibitors and GLP-1 receptor agonists, because of suspected effects of these medications based on prior literature.18, 19 Blood pressure control is essential in adults with diabetes.32-35 The United Kingdom Prospective Diabetes Study showed that for every 10 mmHg decrease in systolic blood pressure, there is a 15-percent decrease in diabetes-related deaths.33 Our findings of modest systolic blood pressure reductions of 3 to 5 mmHg with SGLT-2 inhibitors compared with many other agents are consistent with other reviews18 on these agents, and our review builds on prior work by evaluating direct comparisons of specific medication classes. This is important because thiazolidinediones and GLP-1 receptor agonists have been associated previously with decreases in systolic blood pressure of 3 to 5 mmHg.6, 19 We also found moderate strength of evidence that metformin plus a GLP-1 receptor agonist had a greater reduction in systolic blood pressure than metformin alone (pooled between-group difference, 3.1 mmHg; 95% CI, 1.4 to 4.9 mmHg). While the clinical relevance of these small differences is unclear, a change of 3 to 5 mmHg is about half the effect of a low-sodium diet (around 7 to 11 mmHg) and about one-third the effect of blood pressure medications (around 10 to 15 mmHg).36, 37 Future research should determine if there are any links between these small differences in blood pressure and micro- and macrovascular outcomes, especially given the prevalent use of effective medications to reduce cardiovascular risk (e.g., aspirin, blood pressure and cholesterol medications). Increased heart rate is associated with increased mortality.38 However, whether heart rate is an independent predictor of long-term clinical outcomes, such as mortality, is less clear.39 We wanted to determine if the potential benefits from blood pressure reduction might be offset by a concomitant increase in heart rate. We did not identify any prior systematic reviews that evaluated this outcome for the diabetes comparisons of interest. Only two comparisons had sufficient data to grade the evidence as more than insufficient or low. The SGLT-2 inhibitors in combination with metformin were found to decrease heart rate by 1.5 beats per minute (bpm) (95% CI, 0.6 bpm to 2.3 bpm) when compared with metformin plus a sulfonylurea; metformin and GLP-1 receptor agonists showed no differences in heart rate between groups. Therefore, these early findings support minimal to no effects on heart rate and no increase in heart rate for the newer medications. All-Cause Mortality and Macrovascular and Microvascular Outcomes Additional evidence allowed this report to include firm conclusions regarding metformin versus sulfonylurea monotherapy for cardiovascular mortality. Sulfonylurea monotherapy was associated with a 50-percent to 70-percent higher relative risk of cardiovascular mortality than metformin monotherapy (for sulfonylurea vs. metformin: absolute risk difference, 0.1% to 2.9%; number needed to harm, 34 to 1,000 in RCTs). The low-strength evidence regarding all-cause mortality and cardiovascular morbidity was consistent with this conclusion, also favoring metformin over sulfonylureas. Our results augment findings from prior meta-analyses published in 2012 and 2013, which relied more heavily on observational data or did not report on explicit head-to-head comparisons of metformin and sulfonylurea monotherapy.40, 41 Importantly, we do
- 44. ES-20 not know ifmetformin actually decreases cardiovascular disease mortality or just increases cardiovascular disease mortality less than sulfonylureas; likewise, we do not know if sulfonylureas actually increase cardiovascular disease mortality or just decrease cardiovascular disease mortality less than metformin. We did not find evidence to support substantive conclusions about the comparative effectiveness of thiazolidinediones on long-term cardiovascular risk and therefore could not address the issues raised previously about rosiglitazone and cardiovascular outcomes.42 We did not include the Rosiglitazone Evaluated for Cardiovascular Outcomes in oral agent combination therapy for type 2 Diabetes (RECORD) Trial here because it did not report on macrovascular outcomes stratified by specific medication combinations of interest; however, a reanalysis of data from this study led the FDA to lift its restrictions on the use of rosiglitazone.43 We found little evidence supporting conclusions regarding the comparative effectiveness of most of the newer classes of drugs (DPP-4 inhibitors, GLP-1 receptor agonists, and SGLT-2 inhibitors) and these clinical outcomes. However, three recent large placebo-controlled RCTs not meeting our inclusion criteria (because they did not evaluate direct head-to-head comparisons of interest) evaluated the effects of DPP-4 inhibitors on cardiovascular outcomes: SAVOR-TIMI (Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus Thrombolysis in Myocardial Infarction) 53, EXAMINE (Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care), and TECOS (Trial Evaluating Cardiovascular Outcomes with Sitagliptin). These studies reported noninferiority for DPP-4 inhibitors relative to standard care,44-46 but several limitations prevent conclusions based on these studies: (1) differential diabetes medication use across arms; (2) low power to demonstrate noninferiority; and (3) mixed inconsistent findings on cardiovascular outcomes across trials (N >35,000).44-46 Otherwise, most of the evidence on all-cause mortality and macrovascular and microvascular outcomes came from RCTs that were generally 12 months or shorter in duration with rare or no events; this evidence was of low strength or insufficient, precluding conclusions on the comparative effectiveness of the comparisons of interest for short-term harms. The scant evidence on the comparative effectiveness of diabetes medications and microvascular outcomes (retinopathy, nephropathy, and neuropathy) precluded any substantive conclusions. Safety Outcomes Severe hypoglycemia is associated with increased morbidity (e.g., reduced cognition), increased avoidable health care use (e.g., emergency room visits for hypoglycemia), and increased mortality.47-50 In this report, we confirmed the elevated risk for severe hypoglycemia and nonsevere hypoglycemia with sulfonylureas compared with other drug classes (Figure D).We added to the literature base on SGLT-2 inhibitors by providing more evidence showing that SGLT-2 inhibitors may have less risk of hypoglycemia than metformin, although both medications had low absolute rates of hypoglycemia. We also found that, when compared with metformin plus basal or premixed insulin, metformin plus a GLP-1 receptor agonist had less hypoglycemia risk. For the outcome of gastrointestinal side effects, we also confirmed findings from our 2011 report7 and a prior Cochrane systematic review19 that both metformin and GLP-1 receptor agonists induce more gastrointestinal side effects than most comparators. Our data add information about specific combination comparisons and specific types of gastrointestinal adverse events. The combinations of metformin plus DPP-4 inhibitors did not have worse
- 45. ES-21 gastrointestinal side effectsthan metformin monotherapy or metformin plus a sulfonylurea. We identified new evidence about GLP-1 receptor agonists and SGLT-2 inhibitors: metformin plus a GLP-1 receptor agonist was associated with more gastrointestinal side effects than metformin plus a thiazolidinedione or metformin plus a sulfonylurea. GLP-1 receptor agonists were associated with more vomiting, but similar rates of diarrhea, when compared with metformin monotherapy. SGLT-2 inhibitors did not increase gastrointestinal side effects when added to metformin. There was only one long-term trial lasting 4 years (the rest, less than 2 years) and only a few observational studies of medium quality with 6 to 8 years of followup that assessed the effect of diabetes medications on congestive heart failure. We found 1.2 to 1.6 times increased odds of heart failure with the thiazolidinedione class of medications (low strength of evidence) when compared with metformin or sulfonylureas, a finding also reported in two recent meta- analyses.51, 52 We excluded the RECORD study for this outcome because the active comparator in the analysis was either sulfonylurea or metformin instead of a single active comparator. Consistent with our findings, RECORD showed that the combination of thiazolidinediones and another agent (sulfonylurea or metformin) was associated with a significant doubling in the risk of heart failure compared with the combination of sulfonylurea and metformin.53 Both thiazolidinediones, rosiglitazone and pioglitazone, are contraindicated in patients with serious or severe heart failure (Stage 3 or Stage 4) according to product labels.54, 55 We had low or insufficient strength of evidence for most other medication comparisons for heart failure, including the newer agents. Despite recent concerns about congestive heart failure with DPP-4 inhibitors, we found low or insufficient strength of evidence on the comparative safety of this drug class for this outcome in mainly short studies. Several large double-blind placebo-controlled RCTs evaluating DPP-4 inhibitors on cardiovascular outcomes in adults with moderate to high cardiovascular risk were excluded from our systematic review of head-to-head comparisons but deserve mention because of recent controversy.44-46 Two of these RCTs (comparing either saxagliptin or alogliptin with placebo) reported an increased risk of hospitalization for congestive heart failure in adults at moderate to high cardiovascular risk (range in RD of 0.7% and 0.9%).44, 46 The EXAMINE trial with alogliptin reported these differences solely for the outcome of first hospitalization for heart failure in adults without preexisting congestive heart failure as part of a post hoc subgroup analysis.46 The third placebo- controlled RCT45 compared sitagliptin with placebo on cardiovascular outcomes in adults at elevated risk for these outcomes, and reported no between-group differences in hospitalization for congestive heart failure (3.1% in each arm). It is unclear if differences in these trials result from differences in drug type, chance alone, or other causes. Because of these findings, however, the FDA has requested additional labeling for saxagliptin and alogliptin to reflect concerns about the potential increased risk of hospitalization for congestive heart failure.56 Further research directly comparing specific DPP-4 inhibitors with other active comparators and placebo will be useful in determining the comparative safety of these medications on heart failure risk. Two RCTs of linagliptin are in progress: the Cardiovascular Outcome Study of Linagliptin Versus Glimepiride in Patients with Type 2 Diabetes (CAROLINA) and the Cardiovascular and Renal Microvascular Outcome Study with Linagliptin in Patients with Type 2 Diabetes Mellitus (CARMELINA) studies.57, 58 As in the 2011 report,7 we found little evidence about cancer risk. While animal studies have raised concerns about medullary thyroid cancer with GLP-1 receptor agonists59-62 and in vitro studies have raised concern about pancreatic cancer risk with incretin mimetic therapies,63 we
- 46. ES-22 found no evidenceallowing for substantive conclusions on the association between GLP-1 receptor agonists or DPP-4 inhibitors and cancer. We found low strength of evidence from published RCTs with 104 weeks of followup that the combination of metformin plus a sulfonylurea was favored over the combination of metformin plus a DPP-4 inhibitor for cancer risk; unpublished studies that supported these findings may have strengthened this evidence if they had been included in our review. A newer study with only 52 weeks of followup also corroborated the findings from the longer RCTs. The SAVOR-TIMI 53, TECOS, and EXAMINE trials, mentioned earlier, did not find differences in the risk of pancreatic cancer for DPP-4 inhibitors added to current treatment versus standard care, but other diabetes medication use was differential across arms, thus limiting inferences about effects specific to DPP-4 inhibitors.44-46 Reviews and meta-analyses suggest that metformin decreases the risk of many types of cancer64, 65 and that pioglitazone66 increases the risk of bladder cancer slightly, but we could not include many of the studies supporting those conclusions in our review because of our stringent inclusion criteria for observational studies. We found little evidence from comparative effectiveness studies to substantiate firm conclusions about the risk of pancreatitis for DPP-4 inhibitors and GLP-1 receptor agonists, since we excluded placebo-controlled trials and studies that did not include the specific diabetes medication comparisons of interest for this review. SAVOR-TIMI 53, TECOS, and EXAMINE all reported increased incidence of acute pancreatitis with DPP-4 inhibitors added to standard therapy versus standard therapy alone, with a consistent risk difference of 0.1 percent (number needed to harm for DPP-4 inhibitors, 1,000).44-46 Data across the Liraglutide Effect and Action in Diabetes (LEAD) RCTs also found more pancreatitis with DPP-4 inhibitors.67 We have added additional evidence on specific comparisons based on SGLT-2 inhibitors, confirming the increased risk of genital mycotic infections with this class, which has been described in prior reviews.18, 68 The evidence on SGLT-2 inhibitor comparisons regarding fractures, renal impairment, urinary tract infections, and volume depletion was not conclusive. However, in late 2015, the FDA strengthened its warning of an increased risk of fractures with canagliflozin based on pooled data from nine clinical trials (mean followup, 85 weeks) that showed incidences of fracture of 1.4 and 1.5 per 100 patient-years for canagliflozin 100 mg daily and canagliflozin 300 mg daily, respectively, versus 1.1 per 100 patient-years for the active/placebo combined comparators.69 The labeling for canagliflozin notes that factors that increase fracture risk should be considered when starting canagliflozin.70 The FDA issued a warning on the possible risk of ketoacidosis associated with SGLT-2 inhibitors on May 15, 2015.71 We did not evaluate this outcome, because it was not a concern at the time of the selection of outcomes for this report; the FDA has not changed the labeling for SGLT-2 inhibitors and is currently evaluating emerging data on this issue. A separate analysis of 17,596 participants in canagliflozin trials showed a dose-dependent increased risk of ketoacidosis in participants receiving SGLT-2 inhibitors versus other therapy/placebo; the authors noted that a number of patients with ketoacidosis had evidence of autoimmune diabetes.72 Evidence on other adverse events, including liver injury, lactic acidosis, macular edema or decreased vision, and severe allergic reactions, does not support conclusions. Similarly, the evidence on the comparative effectiveness of diabetes medications in subgroups defined by age, sex, race/ethnicity, and body mass index was generally insufficient for conclusions.
- 47. ES-23 Implications This update providesadditional evidence supporting metformin as the firstline medication therapy to treat type 2 diabetes when tolerated, and it supports a number of treatment options that might be added to metformin based on patient preferences. Not only is metformin favored on many intermediate outcomes, including HbA1c and weight, but also we found more conclusive evidence that cardiovascular mortality is higher with sulfonylureas than metformin. This is consistent with several guidelines, such as those of the American College of Physicians and American Diabetes Association, which recommend metformin as a firstline treatment choice. The alternative to initial therapy with metformin in type 2 diabetes is an important consideration, given that metformin is not currently recommended for use in patients with kidney disease73 (approximately 22% of people with diabetes in the United States)74 or may not be tolerated because of side effects. In addition, the “best” second-line therapy after metformin is still unclear. We evaluated non–metformin-based monotherapy comparisons in this report and demonstrated that the other monotherapies, with the exception of DPP-4 inhibitors, which are not as effective in reducing HbA1c as metformin, generally decrease HbA1c to a similar extent (and comparably to metformin). These other monotherapies’ effects on body weight vary, as do their risks, such as congestive heart failure (increased risk for thiazolidinediones), hypoglycemia (highest risk with sulfonylureas, including for severe hypoglycemia for many comparisons), gastrointestinal side effects (nausea and vomiting with GLP-1 receptor agonists), and genital mycotic infections (increased risk for SGLT-2 inhibitors). Most importantly, we do not have conclusive evidence on the relative long-term effects of non–metformin-based monotherapy comparisons on all-cause mortality or cardiovascular outcomes, microvascular outcomes, and rare serious adverse events (e.g., pancreatitis risk with GLP-1 receptor agonists). The evidence we present on metformin-based combination therapies provides some insight into the selection of add-on therapy to metformin, but it is not definitive because of the uncertainty of long-term outcomes and differential effects on weight and adverse effects. Comparisons of the metformin- based combinations yielded effectiveness and safety results consistent with the metformin monotherapy comparisons described in detail previously. Therefore, the “best” alternative to metformin initial therapy or the “best” second-line therapy choice after metformin remains unclear and should be based on individual patient factors, as suggested in recent guidelines.4 These include clinical factors such as patient age and weight as well as preferences related to differential effects of medications on weight, hypoglycemia, and gastrointestinal and other side effects; tolerance of unknown risks; treatment burden (e.g., oral vs. parenteral administration); and cost. Limitations of the Review Process A few key limitations to our review deserve mention. To focus on comparative effectiveness, we did not include placebo-controlled studies and instead evaluated head-to-head comparisons. We also excluded studies in which participants could take nonstudy drugs for treating diabetes (“background” medications) and the results were not stratified by medication. We used this exclusion to avoid interactions between medications. This was especially important because of our goal of evaluating two-drug combinations. Using these criteria, we excluded several large trials,26, 47, 75-83 because investigators did not stratify their results to allow reporting on the head- to-head comparisons of interest. We also used strict selection criteria for observational studies, mainly based on the control of confounding factors. In this way, we included observational
- 48. ES-24 studies with themost valid results to support conclusions. Also, we focused on interclass (and not intraclass) comparisons in this report. While we did not combine studies in which individual drugs were found to be a clinical or statistical source of heterogeneity, we may have missed smaller intraclass differences. In our 2007 report,6 we found that glyburide/glibenclamide had a higher absolute risk difference of mild, moderate, or total hypoglycemia than other sulfonylureas (pooled RD, 3%; 95% CI, 0.5% to 5%). In this update, which focused on interclass comparisons, the studies that included glyburide/glibenclamide as the sulfonylurea did not consistently have larger between-group differences in hypoglycemia risk than the other sulfonylurea studies. Therefore, these studies were combined with the other sulfonylurea comparisons for hypoglycemia evaluation. For microvascular outcomes, we included studies evaluating more proximal measures, such as change in retinal exam or changes in microalbuminuria, which may be less relevant than other included clinical outcomes of blindness and changes in estimated glomerular filtration rate. However, we were unable to conclude anything about comparative effects on the microvascular outcomes because of lack of sufficient evidence. These distinctions may become more important as more evidence accrues on the different microvascular outcomes. Finally, we did not evaluate patient-reported outcomes, such as quality of life; future research is needed to identify ideal measures to assess treatment-sensitive patient-reported outcomes in diabetes. Applicability Using the PICOTS (populations, interventions, comparisons, outcomes, timing, and setting) framework, the evidence in this report is generally applicable to the population of U.S. adults with type 2 diabetes, with a few notable concerns. Compared with the general population with type 2 diabetes,84 populations in the included studies had fewer elderly adults (e.g., often excluded persons ≥75 years of age), had fewer significant comorbid conditions, and were less racially and ethnically diverse. Regarding the interventions, the majority of studies were less than 2 years long, while patients with diabetes are typically on medications for decades. While many of the longer duration studies were consistent with the short-term findings, more studies lasting longer than 2 years are needed to better understand the durability of the differences reported in shorter term studies. Research Gaps Based on the limitations of the evidence base, we highlight several major gaps in the evidence using the PICOTS framework and provide corresponding recommendations for future research (Table F). The most important gap is the lack of conclusive evidence on the comparative effectiveness and safety of the diabetes medications for all-cause mortality, macrovascular complications, microvascular complications, and rare serious adverse events. Based on the relatively low frequency of these outcomes and long timeframe for development, RCTs are simply not feasible to address this gap because of both cost and the need for evidence now (and not in 5 to 10 years). Therefore, supplementing the rare RCT that can be conducted for these outcomes with high- quality observational studies is paramount. Database requirements for such observational studies include sufficient sample size, followup of patients over time, detailed data on treatments (including doses and duration), and detailed data on confounding variables (e.g., duration of diabetes, comorbid conditions). Study designs will need to handle the following sources of bias: confounding by indication, immortal
- 49. ES-25 time bias, time-and cumulative exposure-varying incidence of outcomes, reverse causation, informative censoring, time-varying drug exposure, and time-dependent confounders.85
- 50. ES-26 Table F. Evidencegaps and future research needs for the comparative effectiveness and safety of diabetes medications for adults with type 2 diabetes Category Evidence Gap Future Research Needs Population Lack of study of older adults, racial/ethnic minorities, and people with comorbid conditions, such as significant renal, cardiovascular, and hepatic impairment Limited evidence on a priori subgroups of interest, such as older adults, racial/ethnic minorities, and subgroups by sex and BMI Studies that include diverse populations Studies with an a priori plan to investigate differences by important subgroups of interest Interventions and comparators (HbA1c, weight, hypoglycemia, and GI adverse events) Limited information on GLP-1 receptor agonist comparisons as monotherapy and in combination with metformin Limited information on metformin plus insulin vs. other metformin-based combinations RCTs evaluating the GLP-1 receptor agonists as monotherapy and in combination with metformin RCTs evaluating metformin plus insulin with other metformin- based combinations, especially metformin plus GLP-1 receptor agonist as injectable add-on therapy to metformin Outcomes All-cause mortality and macrovascular and microvascular outcomes Limited information on macrovascular outcomes and death Underpowered existing evidence Limited number of high-quality observational studies No conclusive evidence on microvascular outcomes No RCTs evaluating these outcomes as a primary outcome Inconsistent outcome definitions, ascertainment, and reporting in each study arm High-quality observational studies* for all comparisons Longer duration RCTs (>2 years) for all comparisons evaluating macrovascular and microvascular events as primary outcomes Standardized definitions for macrovascular and microvascular outcomes Reporting on outcomes in all arms of RCTs Rare safety outcomes Limited evidence on rare safety outcomes Underpowered existing evidence Lack of high-quality observational studies Inconsistent outcome definitions, ascertainment, and reporting in each study arm, especially for pancreatitis and cancer High-quality observational studies* RCTs— o Active ascertainment of all safety outcomes o Standardized definitions for all safety outcomes o Reporting on safety outcomes in all arms o Responsiveness to incorporating evaluation of new safety concerns Timing Most evidence is for short-term outcomes, as few studies lasted more than 2 years Longer duration studies (>2 years) to— o Determine durability of short-term comparative effects on HbA1c and weight o Determine long-term clinical effectiveness and safety Methodological High, and often differential, losses to followup in RCTs Lack of reporting on randomization methods for RCTs Lack of reporting on allocation concealment, blinding, and withdrawals for all studies Lack of appropriate accounting for bias in observational studies Lack of reporting on treatments in observational studies Complete or near-complete followup in RCTs Appropriate methods to account for losses to followup in RCTs Reporting on methods for randomization, allocation concealment, and blinding in RCTs High-quality observational studies* for long-term comparative effectiveness and safety of diabetes medications BMI = body mass index; GI = gastrointestinal; GLP-1 = glucagon-like peptide-1; HbA1c = hemoglobin A1c; RCT = randomized controlled trial. *See text for more detail.
- 51. ES-27 Conclusions The evidence supportsmetformin as a firstline therapy, given its beneficial effects on HbA1c, weight, cardiovascular mortality (vs. sulfonylureas), and relative safety profile. The comparative long-term benefits and harms of other diabetes medications remain unclear. In this report, we provide comprehensive information comparing the benefits and common and serious harms of diabetes medications. In the absence of conclusive findings on long-term clinical and safety outcomes for most medication comparisons, this evidence synthesis can facilitate personalized treatment choices for clinicians and their patients, as well as support decisionmaking by payers and regulators.
- 52. ES-28 References 1. Centers forDisease Control and Prevention. National Diabetes Statistics Report: Estimates of Diabetes and Its Burden in the United States, 2014. Atlanta, GA: U.S. Department of Health and Human Services; 2014. www.cdc.gov/diabetes/pubs/statsreport14/na tional-diabetes-report-web.pdf. Accessed February 26, 2015. 2. American Diabetes Association. Standards of medical care in diabetes-2014. Diabetes Care. 2014;37(Suppl 1):S14-S80. PMID: 24357209. 3. Qaseem A, Humphrey LL, Sweet DE, et al. Oral pharmacologic treatment of type 2 diabetes mellitus: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2012 Feb 7;156(3):218-31. PMID: 22312141. 4. American Diabetes Association. (7) Approaches to glycemic treatment. Diabetes Care. 2015 Jan;38 Suppl:S41-8. PMID: 25537707. 5. Garber A, Abrahamson M, Barzilay J, et al. American Association of Clinical Endocrinologists' comprehensive diabetes management algorithm 2013 consensus statement—executive summary. Endocr Pract. 2013;19(Suppl 2):1-48. PMID: 23816937. 6. Bolen S, Wilson L, Vassy J, et al. Comparative Effectiveness and Safety of Oral Diabetes Medications for Adults with Type 2 Diabetes. Comparative Effectiveness Review No 8. (Prepared by the Johns Hopkins Evidence-based Practice Center under Contract No. 290-02-0018.) AHRQ Publication No. 07-EHC010-EF. Rockville, MD: Agency for Healthcare Research and Quality; 2007. www.effectivehealthcare.ahrq.gov/reports/fi nal.cfm. 7. Bennett WL, Wilson LM, Bolen S, et al. Oral Diabetes Medications for Adults With Type 2 Diabetes: An Update. Comparative Effectiveness Review No. 27. (Prepared by Johns Hopkins University Evidence-based Practice Center under Contract No. 290-02- 0018.) AHRQ Publication No. 11-EHC038- EF. Rockville, MD: Agency for Healthcare Research and Quality; 2011. www.effectivehealthcare.ahrq.gov/reports/fi nal.cfm. 8. Colhoun HM, Livingstone SJ, Looker HC, et al. Hospitalised hip fracture risk with rosiglitazone and pioglitazone use compared with other glucose-lowering drugs. Diabetologia. 2012 Nov;55(11):2929-37. PMID: 22945303. 9. Lu CJ, Sun Y, Muo CH, et al. Risk of stroke with thiazolidinediones: a ten-year nationwide population-based cohort study. Cerebrovasc Dis. 2013;36(2):145-51. PMID: 24029780. 10. Mahaffey KW, Hafley G, Dickerson S, et al. Results of a reevaluation of cardiovascular outcomes in the RECORD trial. Am Heart J. 2013 Aug;166(2):240-9 e1. PMID: 23895806. 11. Mamtani R, Haynes K, Bilker WB, et al. Association between longer therapy with thiazolidinediones and risk of bladder cancer: a cohort study. J Natl Cancer Inst. 2012 Sep 19;104(18):1411-21. PMID: 22878886. 12. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996 Feb;17(1):1-12. PMID: 8721797. 13. Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health. 1998 Jun;52(6):377-84. PMID: 9764259. 14. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ. 2003 Sep 6;327(7414):557-60. PMID: 12958120.
- 53. ES-29 15. DerSimonian R,Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986 Sep;7(3):177-88. PMID: 3802833. 16. Cornell JE, Mulrow CD, Localio R, et al. Random-effects meta-analysis of inconsistent effects: a time for change. Ann Intern Med. 2014 Feb 18;160(4):267-70. PMID: 24727843. 17. Owens DK, Lohr KN, Atkins D, et al. AHRQ series paper 5: grading the strength of a body of evidence when comparing medical interventions--Agency for Healthcare Research and Quality and the Effective Health-Care Program. J Clin Epidemiol. 2010 May;63(5):513-23. PMID: 19595577. 18. Vasilakou D, Karagiannis T, Athanasiadou E, et al. Sodium-glucose cotransporter 2 inhibitors for type 2 diabetes: a systematic review and meta-analysis. Ann Intern Med. 2013 Aug 20;159(4):262-74. PMID: 24026259. 19. Shyangdan DS, Royle P, Clar C, et al. Glucagon-like peptide analogues for type 2 diabetes mellitus. Cochrane Database Syst Rev [serial on the Internet]. 2011;(10). http://onlinelibrary.wiley.com/doi/10.1002/1 4651858.CD006423.pub2/abstract. 20. Hong J, Zhang Y, Lai S, et al. Effects of metformin versus glipizide on cardiovascular outcomes in patients with type 2 diabetes and coronary artery disease. Diabetes Care. 2013 May;36(5):1304-11. PMID: 23230096. 21. Kahn SE, Haffner SM, Heise MA, et al. Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy. N Engl J Med. 2006 Dec 7;355(23):2427-43. PMID: 17145742. 22. Ahren B, Johnson SL, Stewart M, et al. HARMONY 3: 104-week randomized, double-blind, placebo- and active-controlled trial assessing the efficacy and safety of albiglutide compared with placebo, sitagliptin, and glimepiride in patients with type 2 diabetes taking metformin. Diabetes Care. 2014 Aug;37(8):2141-8. PMID: 24898304. 23. Schernthaner G, Duran-Garcia S, Hanefeld M, et al. Efficacy and tolerability of saxagliptin compared with glimepiride in elderly patients with type 2 diabetes: a randomized, controlled study (GENERATION). Diabetes Obes Metab. 2015 Jul;17(7):630-8. PMID: 25761977. 24. Schellhase KG, Koepsell TD, Weiss NS. Glycemic control and the risk of multiple microvascular diabetic complications. Fam Med. 2005 Feb;37(2):125-30. PMID: 15690253. 25. Vijan S, Hofer TP, Hayward RA. Estimated benefits of glycemic control in microvascular complications in type 2 diabetes. Ann Intern Med. 1997 Nov 1;127(9):788-95. PMID: 9382399. 26. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet. 1998 Sep 12;352(9131):837-53. PMID: 9742976. 27. Liu SC, Tu YK, Chien MN, et al. Effect of antidiabetic agents added to metformin on glycaemic control, hypoglycaemia and weight change in patients with type 2 diabetes: a network meta-analysis. Diabetes Obes Metab. 2012 Sep;14(9):810-20. PMID: 22486990. 28. McIntosh B, Cameron C, Singh SR, et al. Second-line therapy in patients with type 2 diabetes inadequately controlled with metformin monotherapy: a systematic review and mixed-treatment comparison meta-analysis. Open Med. 2011;5(1):e35-48. PMID: 22046219. 29. Kahn BB, Flier JS. Obesity and insulin resistance. J Clin Invest. 2000 Aug;106(4):473-81. PMID: 10953022. 30. Purnell TS, Joy S, Little E, et al. Patient preferences for noninsulin diabetes medications: a systematic review. Diabetes Care. 2014 Jul;37(7):2055-62. PMID: 24963113.
- 54. ES-30 31. Richter B,Bandeira-Echtler E, Bergerhoff K, et al. Dipeptidyl peptidase-4 (DPP-4) inhibitors for type 2 diabetes mellitus. Cochrane Database Syst Rev [serial on the Internet]. 2008;(2). http://onlinelibrary.wiley.com/doi/10.1002/1 4651858.CD006739.pub2/abstract. 32. SHEP Cooperative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program (SHEP). JAMA. 1991 Jun 26;265(24):3255-64. PMID: 2046107. 33. UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. BMJ. 1998 Sep 12;317(7160):703-13. PMID: 9732337. 34. Adler AI, Stratton IM, Neil HA, et al. Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36): prospective observational study. BMJ. 2000 Aug 12;321(7258):412-9. PMID: 10938049. 35. Wang JG, Staessen JA, Gong L, et al. Chinese trial on isolated systolic hypertension in the elderly. Systolic Hypertension in China (Syst-China) Collaborative Group. Arch Intern Med. 2000 Jan 24;160(2):211-20. PMID: 10647760. 36. Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. DASH-Sodium Collaborative Research Group. N Engl J Med. 2001 Jan 4;344(1):3-10. PMID: 11136953. 37. Wu J, Kraja AT, Oberman A, et al. A summary of the effects of antihypertensive medications on measured blood pressure. Am J Hypertens. 2005 Jul;18(7):935-42. PMID: 16053990. 38. Nauman J, Janszky I, Vatten LJ, et al. Temporal changes in resting heart rate and deaths from ischemic heart disease. JAMA. 2011 Dec 21;306(23):2579-87. PMID: 22187277. 39. Singh N. Diabetes, heart rate, and mortality. J Cardiovasc Pharmacol Ther. 2002 Apr;7(2):117-29. PMID: 12075400. 40. Phung OJ, Schwartzman E, Allen RW, et al. Sulphonylureas and risk of cardiovascular disease: systematic review and meta- analysis. Diabet Med. 2013 Oct;30(10):1160-71. PMID: 23663156. 41. Monami M, Genovese S, Mannucci E. Cardiovascular safety of sulfonylureas: a meta-analysis of randomized clinical trials. Diabetes Obes Metab. 2013 Oct;15(10):938- 53. PMID: 23594109. 42. Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med. 2007 Jun 14;356(24):2457-71. PMID: 17517853. 43. U.S. Food and Drug Administration. FDA requires removal of some prescribing and dispensing restrictions for rosiglitazone- containing diabetes medicines. 2013. www.fda.gov/downloads/Drugs/DrugSafety/ UCM381108.pdf. Accessed February 25, 2015. 44. Scirica BM, Bhatt DL, Braunwald E, et al. Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus. N Engl J Med. 2013 Oct 3;369(14):1317-26. PMID: 23992601. 45. Green JB, Bethel MA, Armstrong PW, et al. Effect of sitagliptin on cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2015 Jul 16;373(3):232-42. PMID: 26052984. 46. Zannad F, Cannon CP, Cushman WC, et al. Heart failure and mortality outcomes in patients with type 2 diabetes taking alogliptin versus placebo in EXAMINE: a multicentre, randomised, double-blind trial. Lancet. 2015 May 23;385(9982):2067-76. PMID: 25765696. 47. Duckworth W, Abraira C, Moritz T, et al. Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med. 2009 Jan 8;360(2):129-39. PMID: 19092145.
- 55. ES-31 48. Bonds DE,Miller ME, Bergenstal RM, et al. The association between symptomatic, severe hypoglycaemia and mortality in type 2 diabetes: retrospective epidemiological analysis of the ACCORD study. BMJ. 2010;340:b4909. PMID: 20061358. 49. Holman RR, Farmer AJ, Davies MJ, et al. Three-year efficacy of complex insulin regimens in type 2 diabetes. N Engl J Med. 2009 Oct 29;361(18):1736-47. PMID: 19850703. 50. Budnitz DS, Shehab N, Kegler SR, et al. Medication use leading to emergency department visits for adverse drug events in older adults. Ann Intern Med. 2007 Dec 4;147(11):755-65. PMID: 18056659. 51. Lago RM, Singh PP, Nesto RW. Congestive heart failure and cardiovascular death in patients with prediabetes and type 2 diabetes given thiazolidinediones: a meta-analysis of randomised clinical trials. Lancet. 2007 Sep 29;370(9593):1129-36. PMID: 17905165. 52. Singh S, Loke YK, Furberg CD. Long-term risk of cardiovascular events with rosiglitazone: a meta-analysis. JAMA. 2007 Sep 12;298(10):1189-95. PMID: 17848653. 53. Home PD, Pocock SJ, Beck-Nielsen H, et al. Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicentre, randomised, open-label trial. Lancet. 2009 Jun 5;373(9681):2125-35. PMID: 19501900. 54. GlaxoSmithKline. Highlights of Prescribing Information: Avandia (rosiglitazone maleate) tablets. 2014. http://us.gsk.com/products/assets/us_avandia .pdf. Accessed March 2, 2015. 55. Takeda Pharmaceuticals America. Highlights of Prescribing Information: Actos (pioglitazone) tables for oral use. 2013. http://general.takedapharm.com/content/file/ pi.pdf?applicationcode=8a9c4571-a123- 4477- 91deb9cafe7d07e3&filetypecode=actospi. Accessed March 2, 2015. 56. U.S. Food and Drug Administration. FDA panel wants new DPP-4 inhibitor labels - cardiovascular data warrant new risk information for saxagliptin and alogliptin, advisers say. www.medpagetoday.com/PublicHealthPolic y/ClinicalTrials/50990. Accessed July 25, 2015. 57. Clinicaltrials.gov. CAROLINA: Cardiovascular Outcome Study of Linagliptin Versus Glimepiride in Patients With Type 2 Diabetes. 2010. https://clinicaltrials.gov/ct2/show/NCT0124 3424. Accessed July 30, 2015. 58. Clinicaltrials.gov. Cardiovascular and Renal Microvascular Outcome Study With Linagliptin in Patients With Type 2 Diabetes Mellitus (CARMELINA). 2013. https://clinicaltrials.gov/ct2/show/NCT0189 7532. Accessed July 30, 2015. 59. U.S. Food and Drug Administration. Highlights of Prescribing Information: Victoza (liraglutide [rDNA origin] injection, solution for subcutaneous use. 2011. www.accessdata.fda.gov/drugsatfda_docs/la bel/2011/022341s004lbl.pdf. Accessed March 2, 2015. 60. U.S. Food and Drug Administration. Highlights of Prescribing Information: Tanzeum (albiglutide) for injection, for subcutaneous use. 2014. www.accessdata.fda.gov/drugsatfda_docs/la bel/2014/125431s000lbl.pdf. Accessed March 2, 2015. 61. U.S. Food and Drug Administration. Highlights of Prescribing Information: Bydureon (exenatide extended-release) for injectable suspension. 2015. www.fda.gov/safety/medwatch/safetyinform ation/ucm400570.htm. Accessed August 7, 2015. 62. U.S. Food and Drug Administration. Trulicity (dulaglutide) injection, for subcutaneous use. 2015. www.fda.gov/safety/medwatch/safetyinform ation/ucm442202.htm. Accessed August 7, 2015.
- 56. ES-32 63. U.S. Foodand Drug Administration. Incretin mimetic drugs for type 2 diabetes: early communication - reports of possible increased risk of pancreatitis and pre- cancerous findings of the pancreas. 2013. www.fda.gov/Safety/MedWatch/SafetyInfor mation/SafetyAlertsforHumanMedicalProdu cts/ucm343805.htm. Accessed August 1, 2015. 64. Franciosi M, Lucisano G, Lapice E, et al. Metformin therapy and risk of cancer in patients with type 2 diabetes: systematic review. PLoS One. 2013;8(8):e71583. PMID: 23936520. 65. Zhang ZJ, Bi Y, Li S, et al. Reduced risk of lung cancer with metformin therapy in diabetic patients: a systematic review and meta-analysis. Am J Epidemiol. 2014 Jul 1;180(1):11-4. PMID: 24920786. 66. Ferwana M, Firwana B, Hasan R, et al. Pioglitazone and risk of bladder cancer: a meta-analysis of controlled studies. Diabet Med. 2013 Sep;30(9):1026-32. PMID: 23350856. 67. Franks AS, Lee PH, George CM. Pancreatitis: a potential complication of liraglutide? Ann Pharmacother. 2012 Nov;46(11):1547-53. PMID: 23136352. 68. Kawalec P, Mikrut A, Lopuch S. The safety of dipeptidyl peptidase-4 (DPP-4) inhibitors or sodium-glucose cotransporter 2 (SGLT-2) inhibitors added to metformin background therapy in patients with type 2 diabetes mellitus: a systematic review and meta- analysis. Diabetes Metab Res Rev. 2014 May;30(4):269-83. PMID: 24829965. 69. U.S. Food and Drug Administration. FDA Drug Safety Communication: FDA revised label of diabetes drug canagliflozin (Invokana, Invokamet) to include updates on bone fracture risk and new information on decreased bone mineral density. 2015. www.fda.gov/Drugs/DrugSafety/ucm46144 9.htm. Accessed September 16, 2015. 70. U.S. Food and Drug Administration. Highlights of Prescribing Information. Invokana (canagliflozin) tablets, for oral use. 2015. www.accessdata.fda.gov/drugsatfda_docs/la bel/2015/204042s006lbl.pdf. Accessed September 16, 2015. 71. U.S. Food and Drug Administration. FDA Drug Safety Communication: FDA warns that SGLT2 inhibitors for diabetes may result in a serious condition of too much acid in the blood. 2015. www.fda.gov/Drugs/DrugSafety/ucm44684 5.htm. Accessed July 25, 2015. 72. Erondu N, Desai M, Ways K, et al. Diabetic ketoacidosis and related events in the Canagliflozin Type 2 Diabetes Clinical Program. Diabetes Care. 2015 Sep;38(9):1680-6. PMID: 26203064. 73. Bristol-Myers Squibb. GLUCOPHAGE® (metformin hydrochloride) Tablets. GLUCOPHAGE® XR (metformin hydrochloride) Extended-Release Tablets. http://packageinserts.bms.com/pi/pi_glucoph age_xr.pdf. Accessed July 30, 2015. 74. Bailey RA, Wang Y, Zhu V, et al. Chronic kidney disease in US adults with type 2 diabetes: an updated national estimate of prevalence based on Kidney Disease: Improving Global Outcomes (KDIGO) staging. BMC Res Notes. 2014;7:415. PMID: 24990184. 75. Dormandy JA, Charbonnel B, Eckland DJ, et al. Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial In macroVascular Events): a randomised controlled trial. Lancet. 2005 Oct 8;366(9493):1279-89. PMID: 16214598. 76. Nauck M, Frid A, Hermansen K, et al. Efficacy and safety comparison of liraglutide, glimepiride, and placebo, all in combination with metformin, in type 2 diabetes: the LEAD (liraglutide effect and action in diabetes)-2 study. Diabetes Care. 2009 Jan;32(1):84-90. PMID: 18931095. 77. UK Prospective Diabetes Study (UKPDS) Group. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet. 1998 Sep 12;352(9131):854-65. PMID: 9742977.
- 57. ES-33 78. Turner R,Murchison L, Wright AD, et al. United Kingdom Prospective Diabetes Study 24: a 6-year, randomized, controlled trial comparing sulfonylurea, insulin, and metformin therapy in patients with newly diagnosed type 2 diabetes that could not be controlled with diet therapy. Ann Intern Med. 1998;128(3):165-75. PMID: 9454524. 79. U.K. Prospective Diabetes Study. II. Reduction in HbA1c with basal insulin supplement, sulfonylurea, or biguanide therapy in maturity-onset diabetes. A multicenter study. Diabetes. 1985 Aug;34(8):793-8. PMID: 2862087. 80. United Kingdom Prospective Diabetes Study (UKPDS). 13: Relative efficacy of randomly allocated diet, sulphonylurea, insulin, or metformin in patients with newly diagnosed non-insulin dependent diabetes followed for three years. BMJ. 1995 Jan 14;310(6972):83-8. PMID: 7833731. 81. Gerstein HC, Miller ME, Byington RP, et al. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med. 2008 Jun 12;358(24):2545-59. PMID: 18539917. 82. Patel A, MacMahon S, Chalmers J, et al. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med. 2008 Jun 12;358(24):2560-72. PMID: 18539916. 83. Nissen SE, Nicholls SJ, Wolski K, et al. Comparison of pioglitazone vs glimepiride on progression of coronary atherosclerosis in patients with type 2 diabetes: the PERISCOPE randomized controlled trial. JAMA. 2008 Apr 2;299(13):1561-73. PMID: 18378631. 84. Wong HK, Ong KL, Cheung CL, et al. Utilization of glucose, blood pressure, and lipid lowering medications among people with type II diabetes in the United States, 1999-2010. Ann Epidemiol. 2014 Jul;24(7):516-21 e1. PMID: 24935464. 85. Patorno E, Patrick AR, Garry EM, et al. Observational studies of the association between glucose-lowering medications and cardiovascular outcomes: addressing methodological limitations. Diabetologia. 2014 Nov;57(11):2237-50. PMID: 25212258.
- 58. 1 Introduction Background Type 2 diabetescurrently affects 9.3 percent of the US population, or 29.1 million people.1 The proportion of affected individuals in the US varies greatly by race and ethnicity: 16 percent of American Indian/ Alaska Natives, 13 percent of non-Hispanic black Americans and Hispanic Americans, 9 percent of Asian Americans, and 7 percent of non-Hispanic white Americans are afflicted with diabetes. The vast majority of these cases are type 2 diabetes.2 Within these racial categories, rates also vary substantially within sub-populations (e.g., South Asian-Americans and East Asian-Americans).2 Estimates of diabetes incidence that include laboratory-diagnosed diabetes, in addition to self-report, are higher than those reported by the US Centers for Disease Control and Prevention.3 Encouragingly, most reports in the US and Europe suggest that the incidence of disease has not been rising over the past decade.4 Similarly, the age at diagnosis has been relatively stable at 55 years in non-Hispanic whites, and 49 years in non-Hispanic blacks and Hispanics.5 Diabetes and its complications are a substantial public health burden, as they contribute significantly to mortality, morbidity, and health care costs.1, 6 Costs related to diabetes were approximately $245 billion in 2012.1 Complications of longstanding diabetes include the microvascular complications of retinopathy and blindness, neuropathy, nephropathy, and end- stage kidney disease. Diabetes is the most prevalent cause of new-onset blindness and new-onset end-stage renal disease in adults in the US. Diabetes also contributes importantly to macrovascular complications, including coronary artery disease, peripheral arterial disease, and carotid artery disease, and increases the risk of cardiovascular-related death nearly two-fold.7 Lifestyle modification and pharmacologic therapy are the cornerstones of the management of hyperglycemia for type 2 diabetes.8 Results from randomized controlled trials have established that the risk of microvascular complications, particularly retinopathy, can be reduced with glycemic control in patients with type 2 diabetes.9, 10 However, studies in the past decade have suggested that using diabetes medications to achieve intensive glycemic control [hemoglobin A1c (HbA1c) less than 7%] does not benefit cardiovascular morbidity and mortality11, 12 and may harm patients, including those with important co-morbid conditions.13 Recent work also suggests that the effects of intensive glucose lowering may vary across racial and ethnic groups.14 These mixed results on the benefits and safety of glycemic control through pharmacologic therapy suggest the need for further research, including investigation of the long-term impact of glucose lowering therapies. Even if questions about intensity of control are resolved, clinicians and other stakeholders need to determine the optimal agent for glucose lowering. Given the ever-increasing literature about type 2 diabetes medications and the recent approval of many new medications, an updated systematic review evaluating the effects of these medications on intermediate and long-term effectiveness and safety outcomes will be valuable to clinicians, patients, investigators, funders, guideline developers, and payers. In this era of intensive, direct-to-consumer marketing of new drugs, clinicians need a trustworthy source of comprehensive information about the comparative effectiveness and safety of medications. This review seeks to provide information about treatment options to a diverse set of clinicians, including family practitioners, general internists, nurse practitioners, physician assistants, nurses, pharmacists, endocrinologists, cardiologists, nephrologists, and others. Guideline developers may also find this review to be informative for clinical practice guideline preparation. Patients and patient advocates will find the information
- 59. 2 valuable when makingdecisions about treatment options. Finally, investigators will be able to use the results of this review to identify gaps in the literature and formulate original research questions to fill these knowledge gaps. Rationale for Update of Review on Comparative Effectiveness of Diabetes Medications The Effective Health Care (EHC) Program of the Agency for Healthcare Research and Quality (AHRQ) has published two systematic reviews comparing monotherapies and medication combinations for adults with type 2 diabetes.15, 16 In 2007, the AHRQ published its first systematic review, including 216 studies, on this topic.15 This review concluded that most diabetes medications approved by the U.S. Food and Drug Administration (FDA) had similar effects on reducing HbA1c, and most drugs, except for metformin and acarbose, caused at least modest increases in body weight. The sulfonylurea class was associated with an increased risk of hypoglycemia, metformin with gastrointestinal problems, and the thiazolidinediones with heart failure. Importantly, the literature was too sparse to support any conclusions about differential effects of the oral diabetes medications on all-cause mortality, cardiovascular mortality and morbidity, and microvascular complications. When asked by AHRQ to update that review in 2011, we identified an additional 140 randomized controlled trials and 26 observational studies.16 We found that most medications lowered HbA1c by 1 absolute percentage point, on average, but metformin was more effective for HbA1c-lowering than the dipeptidyl-peptidase 4 (DPP-4) inhibitors, a newer class of diabetes medications approved since the initial report. Mostly, the two-drug combinations had similar effects on HbA1c reduction. Compared with metformin, thiazolidinediones and sulfonylureas contributed to more weight gain. Sulfonylureas had a four- fold higher risk of mild/moderate hypoglycemia compared with metformin alone, and, in combination with metformin, had more than a five-fold increased risk of hypoglycemia when compared with metformin plus thiazolidinediones. The risk of congestive heart failure was higher with thiazolidinediones than with sulfonylureas, and the risk of bone fractures was higher with thiazolidinediones than with metformin. Thus, the evidence continued to support use of metformin as a first line agent, based on its effects on HbA1c and weight and side effect profile. The risk of adverse effects was the main determinant of the risk-benefit balance for the two-drug combinations. Despite the addition of important evidence on the HbA1c-lowering and adverse effects of the FDA-approved diabetes medications in 2011, data on the then recently-approved medication classes (glucagon-like peptide-1 (GLP-1) agonists and DPP-4 inhibitors) were sparse, and data on long-term outcomes for both older and newer medications were still lacking.17, 18 Based on these prior systematic reviews, metformin has strong evidence to support its use as an initial pharmacologic treatment for most patients with type 2 diabetes;7 It’s use as a first-line therapy has been widely promoted by clinical practice guidelines.19-21 Not all patients, however, can successfully use metformin due to contraindications to its use or intolerance of its side effects. The evidence base regarding alternative monotherapies for these patients continues to evolve. Since January 2010, one new medication class [the sodium-glucose cotransporter 2 (SGLT-2) inhibitors, with three new medications] and several new DPP-4 inhibitors and GLP-1 receptor agonists have been approved by the FDA. Also since 2010, additional data on previously- approved medications have emerged that could change the balance of benefit and risk attributable to these drugs or could alter the strength of evidence about some of the drug comparisons previously reviewed.22-25 Including insulin, there are 10 medication classes with
- 60. 3 approval by theFDA for treatment of type 2 diabetes. We limited the add-on insulins to premixed or basal insulins in the 2011 report since these are often used as a second line agent after metformin. We have included most, although not all medication classes, in this updated systematic review (Table 1). Table 1. Characteristics of medications included in this report Class Main Mechanism of Action Drug Trade Name Dosing Biguanides Inhibit glucose production by the liver Metformin Glucophage®, Glucophage XR® Oral: 500 to 2550 mg divided doses (qd to tid) Max dose: 2550 mg; 2000 mg for XR Thiazolidinediones Increase glucose uptake by skeletal muscle Pioglitazone Actos® Oral: 15 to 30 mg qd Max dose: 45 mg qd Rosiglitazone Avandia® Oral: 4 to 8 mg qd or 2 to 4 mg bid Max dose: 8 mg qd or 4 mg qd with insulin or sulfonylurea Sulfonylureas Increase insulin secretion by pancreatic beta cells Glimepiride Amaryl® Oral: 1 to 8 mg qd Max dose: 8 mg qd Glipizide Glucotrol®, Glucotrol XL® Oral: 5 to 15 mg qd or 5 to 20 mg bid Max dose: 20 mg bid, 20 mg qd for XL Glyburide or glibenclamide DiaBeta®, Glynase® PresTab®, Micronase® Oral: 2.5 to 20 mg qd or bid Max dose: 20 mg qd DPP-4 inhibitors Increase incretin hormone activity which increases insulin release and decreases inappropriate glucagon production by the pancreatic islet cells* Alogliptin Nesina® Oral: 6.25 to 25 mg qd Recommended dose: 25 mg qd Linagliptin Tradjenta® Oral: 5 mg qd Recommended dose: 5 mg qd Saxagliptin Onglyza® Oral: 2.5 to 5 mg qd Recommended dose: 2.5 or 5 mg qd Sitagliptin Januvia® Oral: 25 to 100 mg qd Recommended dose: 100 mg qd SGLT-2 inhibitors Increases urinary excretion of glucose Canagliflozin Invokana® Oral: 100 to 300 mg Max dose: 300 mg Dapagliflozin Farxiga® Oral: 5 to 10 mg qd Max dose: 10 mg qd Empagliflozin Jardiance® Oral: 10 to 25 mg qd Max dose: 25 mg qd
- 61. 4 Table 1. Characteristicsof medications included in this report (continued) Class Main Mechanism of Action Drug Trade Name Dosing GLP-1 receptor agonists Increase glucose- dependent insulin release and decrease inappropriate glucagon production by the pancreatic islet cells* Albiglutide injection Tanzeum® SC injection: 30 mg qw Max dose: 50 mg qw Dulaglutide injection Trulicity® SC injection: 0.75 to 1.5 mg/0.5 mL Max dose: 1.5 mg/0.5 mL Exenatide injection Byetta® SC injection: 5 to 10 mcg SC bid Liraglutide injection Victoza® SC injection: 1.6 to 1.8 mg SC qd Basal insulin Increases long- acting insulin NPH insulin Humulin N®, Novolin N® NA Insulin detemir Levemir® NA Insulin glargine Lantus® NA Premixed insulin Increases short and long-acting insulin 50% NPH and 50% regular insulin Humulin® 50/50 NA 70% NPH and 30% regular insulin Humulin® 70/30 Novolin® 70/30 NA 50% lispro protamine suspension and 50% lispro Humalog Mix® 50/50 NA 75% lispro protamine suspension and 25% lispro Humalog Mix® 75/25 NA 70% aspart protamine suspension and 30% aspart NovoLog Mix® 70/30 NA bid = twice daily; DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; mcg = microgram; mg = milligrams; mL = milliliter; NA = not applicable since there is no maximum dose for these insulins; NPH = neutral protamine Hagedorn; qd = once daily; qw = once weekly; SC = subcutaneous; SGLT-2 = sodium-glucose co-transporter-2; tid = three-times daily; XL = extended release; XR = extended release. * Decreased glucagon production decreases glucose production by the liver. Analytic Framework Our analytic framework describes the decisions that patients and their providers face when managing type 2 diabetes pharmacologically (Figure 1). It highlights the comparisons and outcomes of interest that correspond to each of the Key Questions in our review. When beginning medical treatment, patients usually start with one of five drug classes (Table 1), which have all been FDA-approved for monotherapy. These include biguanides, thiazolidinediones, sulfonylureas, DPP-4 inhibitors, SGLT-2 inhibitors, and GLP-1 receptor agonists. Clinical guidelines of the American Diabetes Association recommend monitoring HbA1c to determine the need for changing the medication dose or adding another agent to improve glycemic control.26 If the HbA1c is not adequately controlled, clinicians typically add an additional oral
- 62. 5 diabetes medication, orthey may add insulin or a noninsulin injectable medication like a GLP-1 receptor agonist. Clinicians also monitor other intermediate outcomes, such as weight and short- term and long-term safety and adverse effects of the drugs, which vary by drug class. The ultimate goal is to improve long-term outcomes while maximizing quality of life.
- 63. 6 Figure 1. Analyticframework BMI = body mass index; DPP-4 = dipeptidyl peptidase-4; GLP1 = glucagon-like peptide-1; HbA1c = hemoglobin A1c; KQ=Key Question; NPH = neutral protamine Hagedorn; SGLT-2 inhibitor = sodium-glucose co-transporter 2
- 64. 7 Scope This review updatesthe 2011 review on oral diabetes medications for adults with type 2 diabetes.16 In this review, we have chosen to focus on head-to-head drug class comparisons for which there are evidence gaps (see Table 2). We have included a new FDA-approved class of oral diabetes medications, the SGLT-2 inhibitors, including empagliflozin, dapagliflozin, and canagliflozin. We have included new DPP-4 inhibitors approved since the last review, linagliptin and alogliptin, and GLP-1 receptor agonists approved since the last review, albiglutide and dulaglutide. After discussion with our technical expert panel, we excluded head-to-head intraclass drug comparisons and excluded placebo-controlled trials since these comparisons were considered lower priority given the large number of head-to-head studies. Since most guidelines recommend metformin as first-line therapy,19-21 we have chosen to focus Key Questions 1b, 2b, and 3b on metformin-based combination comparisons to assess second-line therapy options after metformin. Given the unique and emerging safety concerns of some of these medications, we have included additional safety outcomes in the review, including impaired renal function, urinary tract infections, genital infections, volume depletion, and bone fractures for studies that include a comparison with SGLT-2 inhibitors. We have also included systolic blood pressure and heart rate as intermediate outcomes for studies including either SGLT-2 inhibitors or GLP-1 receptor agonists. We have chosen to exclude meglitinides as interventions of interest as they are uncommonly used in current clinical practice (<1% of hypoglycemic prescriptions).27, 28 We evaluated meglitinides in our two earlier systematic reviews and found that this class has similar effects on HbA1c and similar rates of hypoglycemia as sulfonylureas. The 2011 update included little new information on meglitinides, and we expected to find little additional evidence for this class of medication. Similarly, we are no longer reporting on lipid levels as intermediate outcomes of interest. LDL targets are no longer universally the primary factor guiding the use of cholesterol-lowering therapy. Current guidelines suggest that 10-year global cardiovascular disease (CVD) risk should be used to determine statin usage and intensity, and this global risk score does not actually include low-density lipoprotein cholesterol.29 Furthermore, triglycerides and high-density lipoprotein are not usual targets of cholesterol therapy. Statin usage is recommended for all patients 40 years of age and older with diabetes in the US.30 Based on these new approaches to lipids, we did not feel that evidence of the impact of diabetes medications on lipid levels would be substantially informative to clinical care to warrant inclusion in this report. Key Questions Key Question 1a: In adults age 18 or older with type 2 diabetes mellitus, what is the comparative effectiveness of the specified monotherapy FDA-approved diabetes medications (see Table 2) for the intermediate outcomes of hemoglobin A1c, weight, systolic blood pressure (for comparisons including SGLT-2 inhibitors or GLP-1 receptor agonists), and heart rate (for comparisons including SGLT-2 inhibitors or GLP-1 receptor agonists)? Key Question 1b: In adults age 18 or older with type 2 diabetes mellitus, what is the comparative effectiveness of the specified metformin-based combinations of FDA-approved diabetes medications (see Table 2) for the intermediate outcomes of hemoglobin A1c, weight,
- 65. 8 systolic blood pressure(for comparisons including SGLT-2 inhibitors or GLP-1 receptor agonists), and heart rate (for comparisons including SGLT-2 inhibitors or GLP-1 receptor agonists)? Key Question 2a: In adults age 18 or older with type 2 diabetes mellitus, what is the comparative effectiveness of the specified monotherapy FDA-approved diabetes medications (see Table 2) for the long-term clinical outcomes of all-cause mortality, cardiovascular and cerebrovascular morbidity and mortality, retinopathy, nephropathy, and neuropathy? Key Question 2b: In adults age 18 or older with type 2 diabetes mellitus, what is the comparative effectiveness of the specified metformin-based combinations of FDA-approved diabetes medications (see Table 2) for the long-term clinical outcomes of all-cause mortality, cardiovascular and cerebrovascular morbidity and mortality, retinopathy, nephropathy, and neuropathy? Key Question 3a: In adults age 18 or older with type 2 diabetes mellitus, what is the comparative safety of the specified monotherapy FDA-approved diabetes medications (see Table 2) regarding liver injury, lactic acidosis, pancreatitis, hypoglycemia, congestive heart failure, cancer, severe allergic reactions, macular edema or decreased vision, and gastrointestinal side effects; and for comparisons including SGLT-2 inhibitors, urinary tract infections, impaired renal function, genital mycotic infections, fracture, and volume depletion? Key Question 3b: In adults age 18 or older with type 2 diabetes mellitus, what is the comparative safety of the specified metformin-based combinations of FDA-approved diabetes medications (see Table 2) regarding liver injury, lactic acidosis, pancreatitis, hypoglycemia, congestive heart failure, cancer, severe allergic reactions, macular edema or decreased vision, and gastrointestinal side effects; and for comparisons including SGLT-2 inhibitors, urinary tract infections, impaired renal function, genital mycotic infections, fracture, and volume depletion? Key Question 4: Do the comparative safety and effectiveness of these treatments differ across subgroups defined by the age, sex, race/ethnicity, and body mass index (BMI) of adults with type 2 diabetes?
- 66. 9 Table 2. Prioritymedication comparisons included for each Key Question Intervention Main Intervention Class (Generic Individual Drug Names) Comparisons Monotherapy as main intervention Biguanides (metformin) Thiazolidinediones* Sulfonylureas † DPP-4 inhibitors SGLT-2 inhibitors GLP-1 receptor agonists ‡ Combination of metformin plus thiazolidinedione Combination of metformin plus sulfonylurea Combination of metformin plus DPP-4 inhibitor Combination of metformin plus SGLT-2 inhibitor Combination of metformin plus GLP-1 receptor agonist Thiazolidinediones (rosiglitazone, or pioglitazone) Sulfonylureas DPP-4 inhibitors SGLT-2 inhibitors GLP-1 receptor agonists Sulfonylureas (glimepiride, glyburide ¶ , glibenclamide ¶ , or glipizide) DPP-4 inhibitors SGLT-2 inhibitors GLP-1 receptor agonists DPP-4 inhibitors (alogliptin, linagliptin, saxagliptin, or sitagliptin) SGLT-2 inhibitors GLP-1 receptor agonists SGLT-2 inhibitors (canagliflozin, dapagliflozin, or empagliflozin) GLP-1 receptor agonists Combination therapy as main intervention Combination of metformin plus (thiazolidinedione or sulfonylurea or DPP-4 inhibitor or SGLT-2 inhibitor or GLP-1 receptor agonist or basal insulin) Combination of metformin plus (sulfonylurea or DPP-4 inhibitor or SGLT-2 inhibitor or GLP-1 receptor agonist or basal insulin ‡ or premixed insulin ‡ ) DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; HbA1c = hemoglobin A1c; SGLT-2 = sodium-glucose co- transporter 2 * For studies comparing thiazolidinediones with metformin, we reviewed only HbA1c, long-term outcomes, and selected safety outcomes given the high strength of evidence from our prior Comparative Effectiveness Review for other outcomes (specifically fracture and weight).16 † For studies comparing sulfonylureas with metformin, we reviewed only the long-term outcomes and cancer given the high strength of evidence on the other outcomes from our prior Comparative Effectiveness Review.16 ‡ The generic individual drug names for the GLP-1 receptor agonists are exenatide, liraglutide, dulaglutide, and albiglutide. The generic individual drug names for basal insulin are insulin glargine, insulin detemir, and neutral protamine Hagedorn (NPH) insulin. The generic individual drug names for premixed insulin are NPH/regular 50/50, NPH/regular 70/30, insulin lispro 50/50, insulin lispro 75/25, and insulin aspart 70/30. ¶ Glyburide and glibenclamide are the same drug.
- 67. 10 Methods Topic Refinement andReview Protocol This review updates the 2011 review on oral diabetes medications for adults with type 2 diabetes.16 We recruited a Technical Expert Panel (TEP) to review a draft of the protocol and a summary of the revisions from the 2011 review (see the Scope and Key Questions section from the Introduction). The TEP included endocrinologists, general internists, biostatisticians, and representatives from government agencies. The TEP reviewed our protocol and provided feedback on the proposed methods for addressing the Key Questions. With the feedback from the TEP and the Agency for Healthcare Research and Quality (AHRQ) representatives, we finalized the protocol and posted it on AHRQ Effective Health Care Program’s Web site (www.effectivehealthcare.ahrq.gov). Literature Search Strategy Search Strategy The 2011 review searched the following databases for the dates listed: MEDLINE® (1966 to April 2010), Embase® (1974 to April 2010), and the Cochrane Central Register of Controlled Trials (CENTRAL). Per AHRQ’s guidance, our new search dates overlapped the prior search by more than 1 year.31 We ran the search developed for the 2011 review with the date restrictions of April 2009 through April 2015 (see Appendix A). An additional expanded search included medical subject headings (MeSH) and text words for all of the new medications included in this updated report. The expanded search did not have any date restrictions. We handsearched the reference lists of all newly included articles and relevant systematic reviews. Additionally, we searched ClinicalTrials.gov to identify relevant registered trials. We also reviewed the Web site of the Food and Drug Administration (FDA) for any unpublished additional studies relevant to the topic as part of our grey literature search. We also provided an opportunity for manufacturers of interventions to submit unpublished data. Study Selection All of the review authors participated in the study selection. Two independent reviewers conducted title scans. For a title to be eliminated at this level, both reviewers needed to indicate that the study was ineligible. If the reviewers disagreed, the article was advanced to the next level, which was abstract review. The abstract review phase was designed to identify studies reporting the effectiveness or safety of the medications and medication combinations of interest. Abstracts were reviewed independently by two investigators and were excluded if both investigators agreed that the article met one or more of the exclusion criteria (see the inclusion and exclusion criteria listed in Table 3). Differences between investigators regarding the inclusion or exclusion of abstracts were tracked and resolved through consensus adjudication. Articles promoted on the basis of the abstract review underwent another independent parallel review to determine if they should be included in the final qualitative and quantitative systematic
- 68. 11 review and meta-analysis.The differences regarding article inclusion were tracked and resolved through consensus adjudication. Table 3. Inclusion and exclusion criteria PICOTS Inclusion Criteria Exclusion Criteria Population We included studies of adult humans with type 2 diabetes, non-insulin dependent diabetes mellitus, or adult- onset diabetes. We excluded studies of patients with type 1 diabetes, impaired glucose tolerance, metabolic syndrome, maturity onset diabetes of youth, and gestational diabetes. We excluded studies if they included only pregnant women or subjects 17 years of age or younger. We excluded studies where everyone was required to have at least one of the following comorbid conditions: ESLD, ESRD, cancer, new onset diabetes after organ transplant, or a recent cardiovascular event within the 3 months prior to study start. Interventions We included studies that evaluated a diabetes medication of interest or drug combination of interest (see Table 2). We excluded studies that did not specify the adjunctive medications, such as those stating use of “any oral hypoglycemic,” or if the study listed several possible medications without stratification of the results by treatment. Comparisons We included studies that evaluated a comparison of interest (see Table 2). We excluded studies that did not have a comparison group or that used a placebo comparison or non-pharmacological comparison. We excluded intraclass head-to-head comparisons.
- 69. 12 Table 3. Inclusionand exclusion criteria (continued) PICOTS Inclusion Criteria Exclusion Criteria Outcomes* We included studies addressing the following intermediate outcomes for KQ1: Hemoglobin A1c^ Weight † Systolic blood pressure ‡ Heart rate ‡ We included studies addressing the following microvascular, macrovascular, and mortality outcomes for KQ2: All-cause mortality Cardiovascular and cerebrovascular morbidity and mortality Retinopathy Nephropathy Neuropathy We included studies addressing the following safety outcomes for KQ3: Liver injury^ Impaired renal function § Lactic acidosis^ Pancreatitis^ Hypoglycemia^ Gastrointestinal side effects^ Congestive heart failure^ Cancer Macular edema or decreased vision^ Fractures § Urinary tract infections § Genital mycotic infections § Volume depletion § KQ4 included studies considering any of the above outcomes. Type of study For KQ1, we included only RCTs. For KQ2 and KQ3, we included RCTs, non-randomized experimental studies with a comparison group, and high- quality observational studies with a comparison group. We included randomized trials utilizing a crossover design with some exceptions. ǁ We excluded studies not written in English ¶ and excluded articles with no original data. We excluded meeting abstracts. Timing and setting We excluded studies in which the observed intervention or exposure period was less than 3 months, 12 weeks, or 90 days. ESLD = end-stage liver disease; ESRD = end-stage renal disease; KQ = Key Question; PICOTS = populations, interventions, comparisons, outcomes, timing, and settings; RCT = randomized controlled trial * Of note, some outcomes could be classified as either safety or long-term clinical outcomes (e.g., myocardial infarction and cancer). ^ We did not evaluate this outcome for metformin vs. sulfonylurea comparisons as the evidence was high from the prior report. † We did not evaluate this outcome for metformin vs. thiazolidinedione or metformin vs. sulfonylurea comparisons as the evidence was high from the prior report. ‡ We evaluated this outcome only for comparisons that included a GLP-1 receptor agonist or a SGLT-2 inhibitor. § We evaluated this outcome only for comparisons that included a SGLT-2 inhibitor. ǁ For crossover randomized trials, we abstracted data on all outcomes at the end of the first period prior to the crossover. If data were not presented at the end of the first period, then we excluded the article for the following outcomes where we would be
- 70. 13 unable to drawconclusions about causality: long-term outcomes (KQ2), fractures, cancer, intermediate outcomes in studies where there was a washout period of less than 3 months; and safety outcomes in studies where the washout period was less than a month except for hypoglycemia, gastrointestinal side effects, and liver injury. ¶ We decided to include non-English language articles through the full text article review phase of the updated search and assess the volume and content of these articles along with workload to determine if abstracting data from these articles would add value to the review. Data Extraction We used a systematic approach to extract all data to minimize the risk of bias in this process. We used standardized forms from the previous reviews as templates for data extraction and pilot tested them for the new medications and outcomes (Appendix B). By creating standardized forms for data extraction, we sought to maximize consistency in identifying all pertinent data available for synthesis. We double-reviewed all data abstracted from the studies. The second reviewer confirmed the first reviewer’s abstracted data for completeness and accuracy. Reviewer pairs were formed to include personnel with both clinical and methodological expertise. A third reviewer audited a random sample of articles to ensure consistency in the data abstraction of the articles. Reviewers were not masked to the authors of the articles, their respective institutions, nor the journals in which their articles were published. For all articles, the reviewers extracted information on the general study characteristics (e.g., study design, study period, and followup); study participants (e.g., age, sex, race, weight/body mass index [BMI], hemoglobin A1c levels, and duration of diabetes); interventions (e.g., initial, maximum, and mean doses, frequency of use, duration of use, and permissibility of treatment intensification with additional therapies), comparisons; the method of ascertainment of safety outcomes; and the outcome results, including measures of variability. We also collected data on outcomes for the subgroups of interest: age, sex, race/ethnicity, and BMI. For continuous outcomes, we extracted the mean difference between groups and a measure of dispersion. If the between-group difference was not reported, we calculated the point estimate of the difference using the mean difference from baseline for each group. If the mean difference from baseline was not reported, we calculated this from the baseline and final values for each group.32 If there were no measures of dispersion for the mean difference from baseline for each group, we calculated the variance using the standard deviation of the baseline and final values, assuming a correlation between baseline and final values of 0.5. We entered all information from the article review process into a DistillerSR database (Evidence Partners Inc., Ottawa, Canada). Reviewers entered comments into the system whenever applicable. The DistillerSR database was used to maintain the data and to create detailed evidence tables and summary tables. Data will later be uploaded into the Systematic Review Data Repository. Risk of Bias Assessment of Individual Studies Two independent reviewers assessed study quality. We assessed the risk of bias in individual randomized controlled trials (RCTs) using the Jadad criteria consistent with the prior report.33 Although newer quality assessment tools exist, we felt that continuing to use the Jadad criteria would be adequate and consistent with our previous methods. We used the Downs and Black tool for assessment of risk of bias for non-randomized trials and observational studies.34 Given that observational studies with a high risk of bias add little value to a systematic review of effectiveness,35 we included only medium- and high-quality observational studies as determined
- 71. 14 by assessment ofeach study’s risk of bias. For inclusion, we required that observational studies account for the following potential confounders: age, sex, either race or socioeconomic status, and co-morbid conditions (quantified with a co-morbidity scale or index, or by inclusion of other medical conditions or medications used by the patient, or with valid methods to adjust for confounding by indication or restricted to one race or age group making adjustment unnecessary). If the study met the confounding criteria, the observational study was considered eligible for inclusion in the review. We also applied the Downs and Black tool and other inclusion criteria for nonrandomized trials and observational studies to the non-randomized trials and observational studies that had been included in the prior report.16 Data Synthesis For each Key Question, we created a set of detailed evidence tables containing all information extracted from eligible studies, including those from the prior evidence reports. We included both the results of individual studies included in the prior report and the results of newly-identified studies. We conducted meta-analyses when there were sufficient data (at least three trials) and studies were sufficiently homogenous with respect to key variables (population characteristics, study duration, and drug dose). For trials having more than one dosing arm, we chose the arm for inclusion that had dosing most consistent with the other trials considered for inclusion in the meta-analysis. When more than one followup interval was reported, we used the data from the followup most similar to the other trials. While there is no definitive cut-point for long-term versus short-term, we considered trials lasting 2 years or longer to be “long-term” since a multifactorial intervention in adults with type 2 diabetes has shown changes in morbidity starting as early as 2 years.36 We tested the heterogeneity among the trials considered for quantitative pooling using a standard chi-squared test using a significance level of alpha less than or equal to 0.10. We also examined heterogeneity among studies with an I-squared statistic, which describes the variability in effect estimates that is due to heterogeneity rather than random chance. We considered a value greater than 50 percent to indicate substantial heterogeneity.37 We pooled the mean difference between groups using a random-effects model with the DerSimonian and Laird formula in settings of low heterogeneity (I-squared <50%).38 We pooled studies using the profile likelihood estimate when we detected high statistical heterogeneity (I-squared >50%).39 When data were insufficient or inappropriate to combine in a meta-analysis, we summarized the outcomes by reporting the ranges of values for mean differences from baseline or mean differences between groups, when available. Since we anticipated that most drugs would have similar physiologic effects within a class, we combined studies of unique medications within classes when reporting outcomes except where known differences exist (e.g., the effects of pioglitazone and rosiglitazone on cardiovascular outcomes). If we saw substantial heterogeneity (I-squared >50%) in pooled estimates for any outcome, we stratified studies by medication within a class and repeated the pooled analyses and recalculated measures of heterogeneity. Additionally, when there were at least 10 studies for a given comparison and outcome and evidence of statistical heterogeneity, we attempted to determine other reasons for heterogeneity by evaluating study-level characteristics, such as baseline values of the outcome, study duration, quality measures, or dosing differences between study arms using metaregression techniques. We also conducted sensitivity analyses by omitting one study at a time to assess the influence of any single study on the pooled estimates.
- 72. 15 For the outcomeof hypoglycemia, we conducted separate analyses for: (a) severe hypoglycemia and (b) mild or moderate or total hypoglycemia. The categories were based on the definitions of hypoglycemia provided in the studies. For hypoglycemia and all other dichotomous outcomes, we calculated pooled odds ratios using a random-effects model with the DerSimonian and Laird formula in settings of low heterogeneity38 or the profile likelihood estimate in settings of high heterogeneity.39 Reporting Bias Assessment We assessed reporting biases in the included RCTs as follows:40 1. Publication bias was evaluated by: a. Using the Begg and Mazumdar41 and the Egger42 test to quantitatively assess for publication bias when there were at least 10 studies for a given comparison and outcome pair b. Comparing ClinicalTrials.gov entries and actual publications for evidence of absence of published literature c. Comparing FDA medical and statistical reviews and actual publications for evidence of absence of published literature (results are detailed in Appendix E). 2. Selective Outcomes Reporting bias was evaluated by comparing differences in reporting on the outcomes of hemoglobin A1c (HbA1c), hypoglycemia, and all-cause mortality in the actual publications to the FDA medical and statistical reviews. 3. Selective Analysis Reporting bias was evaluated by assessing the precision of outcome data reporting by determining the number of studies which reported on an outcome of interest (e.g., HbA1c) but did not report a measure of dispersion completely or at all. We assessed this for the outcomes of HbA1c, hypoglycemia, and all-cause mortality. For dichotomous outcomes (hypoglycemia and all-cause mortality), we evaluated the number of studies reporting the n of events uniformly across all arms. We reviewed this for the studies included for the update only. Strength of the Body of Evidence At the completion of our review, two reviewers sequentially graded the available evidence addressing the Key Questions by adapting an evidence grading scheme recommended by the Guide for Conducting Comparative Effectiveness Reviews.43 We applied evidence grades to the bodies of evidence about each intervention comparison for each outcome that were addressed by at least one RCT or three observational studies. We separately assessed the strength of evidence for shorter and longer studies (2 years or greater); however, we only assessed strength of evidence for longer studies where we could draw a conclusion. We assessed the study limitations, consistency, directness, precision, and reporting bias. We assessed the study limitations of individual studies using the tools described in the Risk of Bias of Individual Studies section. We started with the assumption that randomized controlled trials would have “low” study limitations and observational studies would have “medium” study limitations. We downgraded the study limitations score based on the items in the quality assessment tools.
- 73. 16 We rated thebody of evidence as “consistent” if most of the studies (about 75%) showed the same direction of effect. We rated the consistency of comparison-outcome dyads for which there was only a single study as “unknown.” All other bodies of evidence were rated as “inconsistent.” We rated the bodies of evidence for all outcomes as “direct,” except for heart rate and liver injury. We rated the bodies of evidence for heart rate as “indirect,” because the association between heart rate and clinically important outcomes such as mortality is less strong in adults with diabetes.44 We rated the bodies of evidence for the outcome of liver injury as “indirect,” since most of the studies used liver injury enzyme elevation as the indicator of injury. If we conducted a meta-analysis for a body of evidence, we relied on the results of the meta- analysis to rate precision and used the designated minimally important differences as a point of reference for precision. For continuous outcomes, we rated the body of evidence as “imprecise” if one-half of the width of the confidence interval for the meta-analysis was wider than the minimally important difference. We defined the minimally important difference to be 0.3% for HbA1c, 1 kg for weight, and 3 mmHg for systolic blood pressure. While there are no strict definitions of what should be considered clinically relevant differences, we used minimally important differences that clinical experts suggested are clinically relevant and that are supported, in part, in the literature.45 If there was no meta-analysis, we rated precision by evaluating the narrowness of the confidence intervals or the magnitude of the P-value. For dichotomous outcomes, we evaluated precision using the optimal information size for that outcome. If the total sample size across both arms of the studies was greater than the optimal information size, then we rated the body of evidence as “precise.” Otherwise, it was rated as “imprecise.” We estimated rough optimal information sizes using the Mantel Hanszel model for relative odds and incorporating the approximate baseline rate of the outcome and the desired minimum detectable relative odds (Table 4).46 Table 4. Optimal information size for one arm and classification of dichotomous outcomes for optimal information size “Low” Detectable OR, 1.05 “Medium” Detectable OR, 1.5 “High” Detectable OR, 2.0 “Low” baseline risk, 0.01 654,548 (All-cause mortality, cardiovascular mortality, cardiovascular morbidity, cancer, diabetic nephropathy) 8,364 (Liver injury, pancreatitis, severe allergic reaction, renal impairment, congestive heart failure, microalbuminuria, volume depletion) 2,597 “Medium” baseline risk, 0.15 51,168 (Severe hypoglycemia) 690 (Urinary tract infections, genital infections) 225 “High” baseline risk, 0.3 31,296 446 (Hypoglycemia) 153 (Gastrointestinal events) OR = odds ratio We rated reporting bias by evaluating publication bias, selective outcomes reporting bias, and selective analysis reporting bias (described in the Reporting Bias Assessment section). If any of these domains was rated as “suspected,” then we rated the body of evidence as having “suspected” reporting bias. Otherwise, we rated reporting bias as “undetected.” We classified evidence pertaining to the Key Questions into four categories: (1) “high” grade (indicating high confidence that the evidence reflects the true effect and further research is very unlikely to change our confidence in the estimate of the effect); (2) “moderate” grade (indicating moderate confidence that the evidence reflects the true effect but further research could change
- 74. 17 our confidence inthe estimate of the effect and may change the estimate); (3) “low” grade (indicating low confidence that the evidence reflects the true effect and further research is likely to change our confidence in the estimate of the effect and is likely to change the estimate); and (4) “insufficient” grade (indicating evidence is unavailable or the body of evidence has unacceptable deficiencies, precluding reaching a conclusion). We provided a conclusion regarding whether a given drug was favored over another (or if neither was favored) when the evidence permitted this. For all-cause mortality, cardiovascular mortality, cardiovascular morbidity, and safety outcomes, if we concluded that neither arm was favored (i.e., benefit or harm excluded), we did not rate the evidence as “moderate” in strength if the evidence was underpowered (rated as “imprecise”). We graded the evidence separately for the RCTs and the observational studies.43 The final evidence grade and conclusion was typically based on the RCT grade and could be strengthened by evidence from the observational studies. We noted differences between RCT and observational evidence in the text, when present. Applicability We discussed the applicability of studies in terms of the degree to which the study population (e.g., age, sex, race/ethnicity, and co-morbid conditions), interventions (e.g., dose, frequency, rescue therapy, and duration of exposure), outcomes (e.g., outcome definition and reporting), and settings are typical of the treatment of individuals with type 2 diabetes who are receiving treatment in a usual care setting (conceived as outpatient treatment by internists, family physicians, and endocrinologists). Peer Review and Public Commentary Experts in endocrinologists, general internists, epidemiologists, biostatisticians, and representatives from government agencies were invited to provide external peer review of this systematic review; AHRQ and an associate editor also provided comments. The draft report was posted on the AHRQ Web site for 4 weeks to elicit public comment. We addressed all reviewer comments, revising the text as appropriate, and documented everything in a disposition of comments report that will be made available 3 months after the Agency posts the final systematic review on the EHC Web site.
- 75. 18 Results Results of LiteratureSearches We included 166 publications in our previous review. After excluding studies without a comparison or an outcome relevant to this update, and cohort studies not meeting our revised quality criteria, we included 105 studies (published in 107 articles) in this update. We retrieved 19,171 unique citations from our updated literature search (Figure 2). After reviewing titles, abstracts, and full text, we included 114 new studies (published in 142 new articles). Ten of the new publications were either extensions or additional analyses of studies included in the previous review. In total, we include in this review 219 studies, published in 249 articles.
- 76. 19 Figure 2. Summaryof the search (number of articles) FDA = Food and Drug Administration * Total may exceed number in corresponding box, as articles could be excluded for more than one reason at this level. † Comorbid condition restrictions were end-stage renal disease, end-stage liver disease, cancer, new onset diabetes after transplant, or a cardiovascular event within 3 months (e.g., acute coronary syndrome, acute myocardial infarction, post-coronary artery bypass graft surgery, or with drug-eluting stents) Study Duration of RCTs for All Key Questions (KQ1–KQ4) Of the 177 included randomized controlled trials (RCTs) for all Key Questions combined, most studies were less than a year (Figure 3). Only 4 percent of studies lasted over 2 years, making it difficult to draw any firm conclusions about long-term outcomes. Unless stated otherwise in the text or figures below, results and conclusions for all the Key Questions are for short-term outcomes. Electronic databases MEDLINE® (8053) EMBASE® (21708) Cochrane (1919) Retrieved 31680 Title review 19171 Duplicates 12509 Abstract review 6477 Excluded 12694 Excluded 4838 Article review 1805 Excluded 1495 (update) 61 (previous) Included 219 studies (249 publications) Reasons for exclusion at abstract review* No original data: 1987 No human data: 117 No adults: 17 No patients with type 2 diabetes: 142 No control group: 647 No comparison of interest: 1916 Not an FDA-approved formulation: 15 Followup less than 1 month: 284 Does not apply: 1006 Placebo-controlled trial: 37 Other: 241 Included in previous review 166 Reasons for exclusion at article review* No original data: 98 Meeting abstract: 747 Study population not exclusively patients with type 2 diabetes: 12 Does not meet the study design criteria: 73 Not a comparison of interest: 349 Placebo-controlled trial: 61 Not an FDA-approved formulation: 1 Patients allowed on background medications: 283 No outcome of interest: 77 Followup less than 3 months: 22 Study population was required to have a comorbid condition † : 29 Does not apply: 41 Non-randomized study that does not report on a long-term outcome or adverse event: 21 Non-randomized study that does not account for confounding: 79 Non-English Language: 20 Head-to-head intraclass comparison: 4 Other: 18
- 77. 20 Figure 3. Durationof followup for randomized controlled trials comparing the effects of diabetes medications among adults with type 2 diabetes (N = 177) Key Questions 1a and 1b: Intermediate Outcomes Study Design and Population Characteristics One hundred sixty-two RCTs (reported in 189 articles) evaluated intermediate clinical outcomes for adults with type 2 diabetes and met our inclusion criteria (Appendix D, Tables D1 to D4). All trials were parallel arm RCTs, except one which also used a crossover design47 and one which also used a factorial design.48 About half of the trials answering Key Question 1 occurred partly or exclusively in the United States (US) (n = 26), Japan (n=13), Italy (n = 12), and/or were multi-national (n = 56); the rest of the trials occurred in developed or newly industrialized countries. These RCTs lasted from 12 weeks to 5.5 years; however, most studies (81%) lasted less than 1 year, and only six studies lasted more than 2 years (including the well- known Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycemia in Diabetes (RECORD), and A Diabetes Outcome Progression Trial (ADOPT)).49-54 Only 12 percent of studies (n=20) reported receiving no pharmaceutical support, while about 14 percent of RCTs (n = 22) did not describe whether or not they received pharmaceutical support. The number of studies not describing their pharmaceutical support dropped from 25 percent of the studies included in the last diabetes medication comparative effectiveness report16 to only 5 percent of the newly included 87 studies in this update. The use of rescue therapy (i.e., the addition of another diabetes medication when the blood sugar was not controlled on the randomized treatment regimen) was not reported in 41 of the 87 studies included (47.1%), was not allowed in
- 78. 21 20 studies (23.0%),and was allowed in 26 studies (29.9%). In the studies where rescue therapy was allowed, 12 studies did not specify which medications were used and, when reported, the medications varied greatly. Study participants were mainly middle-aged, overweight, or obese adults who had had diabetes for 3 to 7 years. The exclusion criteria were generally similar for most trials: significant renal, cardiovascular, and hepatic disease. About half of the trials (58%) excluded older subjects (generally over the age of 75 to 80). Almost all of the studies included men and women. About 28% of the RCTs did not report race/ethnicity. In this update, the percent not reporting race/ethnicity increased from 20% of the 119 studies in the prior report16 to 38% of the 89 studies in the newly included studies. In these studies, when race was reported, most subjects were Caucasian, but between 10% and 20% of the enrolled population was of other races. The mean baseline HbA1c among study subjects varied from 6 to 12 absolute percentage points, with most subjects having a mean baseline HbA1c between 7 and 9 absolute percentage points. Risk of Bias All of the studies included in this section were described as randomized (Figure 4). Fifty percent described their randomization scheme; 66 percent described their study as double- blinded. About one-third (36%) of all double-blinded RCTs also described the steps taken to ensure blinding. The majority of trials (86%) described the withdrawals and dropouts. Twelve of the fifteen studies with at least 2 years of followup had over 20% losses to followup. Figure 4. Summary of the risk of bias of randomized controlled trials evaluating intermediate outcomes 100% 50% 66% 36% 86% 4% 30% 6% 14% 46% 57% 4% 0% 20% 40% 60% 80% 100% Randomized Randomization scheme appropriate Double-blind Blinding method appropriate Dropouts described Yes No Not described Not reported/Can't tell
- 79. 22 Key Points andEvidence Grades for Intermediate Outcomes Hemoglobin A1c Monotherapy Comparisons Most oral diabetes medications had similar efficacy in achieving reductions in hemoglobin A1c (HbA1c). o In the prior report, the strength of evidence was graded as high that metformin was similar to sulfonylurea (pooled between-group difference of 0.1%; 95% confidence interval [CI], -0.1% to 0.3%). Therefore, we did not update this comparison for HbA1c in this review. o The strength of evidence (SOE) was graded as high that metformin was similar to thiazolidinedione (pooled between-group difference of -0.04%; 95% CI, -0.11% to 0.03%). o Thiazolidinediones performed similarly to sulfonylureas (pooled between-group difference of -0.04%; 95% CI, -0.13% to 0.06%). (SOE: High) o The SOE was graded as low or insufficient for all the monotherapy comparisons of the newer classes of sodium-glucose cotransporter (SGLT-2) inhibitors and glucagon- like peptide-1 (GLP-1) agonists, and will warrant further study. The one exception was that metformin had a greater reduction in HbA1c compared with dipeptidyl peptidase-4 (DPP-4) inhibitors (pooled between-group difference of -0.4%; 95% CI, -0.5% to -0.3%). (SOE: High) Metformin-Based Combination Comparisons The combination of metformin plus GLP-1 receptor agonists reduced HbA1c more than metformin plus DPP-4 inhibitors, with a pooled between-group difference of -0.65% (95% CI, -0.75% to -0.54%) in the short-term. (SOE: Moderate) Most other combination therapy comparisons had either no significant or no clinically meaningful (<0.3%) between-group differences in HbA1c between arms. The evidence was graded as moderate for the following comparisons: metformin plus a thiazolidinedione versus metformin plus a sulfonylurea, metformin plus a thiazolidinedione versus metformin plus a DPP-4 inhibitor, metformin plus a sulfonylurea versus metformin plus an SGLT-2 inhibitor, metformin plus a DPP-4 inhibitor versus metformin plus an SGLT-2 inhibitor, and metformin plus a DPP-4 inhibitor versus metformin plus a GLP-1 receptor agonist. Despite the clinical interest in comparing metformin plus injectables, there was insufficient or low strength of evidence on glycemic control for the following comparisons: metformin plus the GLP-1 receptor agonists versus metformin plus basal or premixed insulin, and metformin plus premixed insulin versus metformin plus basal insulin. Weight Monotherapy Comparisons In the 2011 report, metformin had greater weight reduction than thiazolidinediones (pooled mean between-group difference of -2.6 kg; 95% CI, -4.1 kg to -1.2 kg) or
- 80. 23 sulfonylureas (pooled meanbetween-group difference of -2.7 kg; 95% CI, -3.5 kg to -1.9 kg) with high strength of evidence. Therefore, we did not update these two comparisons in this report. Metformin had greater weight reduction than DPP-4 inhibitors (pooled mean between- group difference, -1.3 kg; 95% CI, -1.6 kg to -1.0 kg). (SOE: High) SGLT-2 inhibitors had greater weight reduction when compared with metformin or DPP- 4 inhibitors (between-group differences ranging from -1.3 kg to -2.7 kg). (SOE: Moderate for both comparisons) DPP-4 inhibitors and GLP-1 receptor agonists both decreased weight more than thiazolidinediones (between-group differences ranging from -2.3 kg to -3.5 kg). (SOE: Moderate for both comparisons) GLP-1 receptor agonists decreased weight more than sulfonylureas (pooled mean between-group difference, -2.3 kg; 95% CI, -3.3 kg to -1.2 kg). (SOE: Moderate) Sulfonylureas caused slightly less weight gain when compared with thiazolidinediones (between-group difference of -1.2 kg; 95% CI, -1.8 kg to -0.6 kg). (SOE: Moderate) Metformin Versus Metformin-Based Combination Comparisons Metformin monotherapy reduced weight more than the combination of metformin plus a thiazolidinedione (pooled mean between-group difference, -2.2 kg; 95% CI, -2.6 kg to - 1.9 kg) or metformin plus a sulfonylurea (pooled mean between-group difference, -2.2 kg, 95% CI, -3.4 kg to -1.0 kg). (SOE: High for both comparisons) When compared with metformin monotherapy, the combination of metformin plus o SGLT-2 inhibitor had greater weight reduction (pooled mean between-group difference, -2.0 kg; 95% CI, -2.5 kg to -1.5 kg). (SOE: High) o GLP-1 receptor agonist had greater weight reduction (pooled mean between-group difference, -2.0 kg; 95% CI, -2.7 kg to -1.3 kg). (SOE: Moderate) Metformin monotherapy had no significant differences in weight when compared with the combination of metformin plus DPP-4 inhibitors (pooled mean between-group difference, -0.1 kg; 95% CI, -0.2 kg to 0.03 kg). (SOE: Moderate) Metformin-Based Combination Comparisons The combinations of metformin plus a sulfonylurea, metformin plus a GLP-1 receptor agonist, and metformin plus a DPP-4 inhibitor all had a more favorable effect on weight compared with metformin plus a thiazolidinedione (range in between-group differences, - 0.9 kg to -5.1 kg). (SOE: Moderate for all comparisons) When compared with the combination of metformin plus a sulfonylurea, the combination of metformin plus o DPP-4 inhibitors had more favorable effects on weight (pooled mean between-group difference, -2.2 kg; 95% CI, -1.8 kg to -2.5 kg). (SOE: High) o SGLT-2 inhibitors had more favorable effects on weight (pooled mean between- group difference, -4.7 kg; 95% CI, -4.4 kg to -5.0 kg). (SOE: High) o GLP-1 receptor agonist had more favorable effects on weight (range in mean between-group differences, -2.4 kg to -12.3 kg). (SOE: Moderate) o Premixed insulin or basal insulin had less favorable effects on weight (range in mean between-group differences, 0.5 kg to 1.7 kg). The strength of evidence was low for both comparisons, due to the small number of studies. However, taken together, the
- 81. 24 strength of evidencewould be moderate favoring metformin plus sulfonylurea over metformin plus a premixed or long-acting insulin. When compared with metformin plus a DPP-4 inhibitor, the combination of metformin plus o GLP-1 receptor agonist had greater reductions in weight (pooled mean between-group difference, -1.8 kg; 95% CI, -1.1 kg to -2.5 kg). (SOE: Moderate) o SGLT-2 inhibitors had greater reductions in weight (between-group differences of around -2.5 kg). (SOE: Moderate) Despite the clinical interest in comparing metformin plus injectables, there was low strength of evidence on weight for the following comparisons: metformin plus the GLP-1 receptor agonists versus metformin plus basal or premixed insulin, and metformin plus premixed insulin versus metformin plus basal insulin. Systolic Blood Pressure (for Comparisons That Include SGLT-2 Inhibitors or GLP-1 Receptor Agonists) Monotherapy Comparisons SGLT-2 inhibitors had a greater reduction in systolic blood pressure compared with metformin, (pooled between-group difference of -2.8 mmHg; 95% CI, -2.6 mmHg to -3.0 mmHg). (SOE: Moderate) The strength of evidence was graded low or insufficient for the following comparisons: o SGLT-2 inhibitors versus DPP-4 inhibitors, and o GLP-1 receptor agonists versus metformin, thiazolidinediones, sulfonylureas, and DPP-4 inhibitors. Metformin Versus Metformin-Based Combination Comparisons Metformin plus a SGLT-2 inhibitor reduced systolic blood pressure more than metformin alone (pooled between-group difference of -4.4 mmHg; 95% CI, -2.9 to -6.0 mmHg) for shorter studies. (SOE: High) Metformin plus a GLP-1 receptor agonist reduced systolic blood pressure more than metformin alone (pooled between-group difference of -3.1 mmHg; 95% CI, -1.4 to -4.9 mmHg). (SOE: Moderate) Metformin-Based Combination Comparisons Metformin plus a SGLT-2 inhibitor reduced systolic blood pressure more than metformin plus a sulfonylurea (pooled between-group difference, -5.0 mmHg; 95% CI, -4.2 mmHg to -6.0 mmHg) or metformin plus a DPP-4 inhibitor (pooled between-group difference, - 4.1 mmHg; 95% CI, -3.6 mmHg to -4.6 mmHg). (SOE: High and Moderate, respectively) Heart Rate (for Comparisons That Include SGLT-2 Inhibitors or GLP-1 Receptor Agonists) Monotherapy Comparisons Metformin compared with a GLP-1 receptor agonist yielded no differences in heart rate between arms. (SOE: Moderate)
- 82. 25 Metformin Versus Metformin-BasedCombination Comparisons There was low or insufficient evidence for all metformin combination therapies compared with metformin alone. Metformin-Based Combination Comparisons Combination therapy with metformin plus a SGLT-2 inhibitor resulted in less increase in heart rate compared with metformin plus a sulfonylurea (pooled between group difference in heart rate, -1.5 bpm; 95% CI, -0.6 bpm to -2.3 bpm). (SOE: Moderate) Evidence for Hemoglobin A1c Monotherapy Comparisons Metformin Versus Thiazolidinediones Twenty-three RCTs, each lasting approximately one year or less, directly compared metformin with a thiazolidinedione, and showed no between-group differences in HbA1c (pooled between-group difference of -0.04%; 95% CI, -0.11% to 0.03%) (Figure 5).55-77 We tested the effect of each individual study on the combined point estimate. No single study influenced the pooled results. No substantial heterogeneity was identified. Three additional trials examined this comparison but were excluded from the pooled results, one with a median study duration of 4 years,50 one which reported median HbA1c instead of means,78 and one study where the mean difference between groups could not be calculated.79 The 4-year, double-blind RCT (known as the ADOPT study), with around a 60% loss to followup, was designed to compare long-term glycemic control between metformin, rosiglitazone, and glyburide monotherapy as initial treatment for adults with type 2 diabetes.50 The authors found a statistically significant but small difference between groups favoring rosiglitazone (mean difference between groups 0.1%; 95% CI, 0.05% to 0.2%). Of note, the HbA1c decreased in all groups for the first 6 months and then increased in all groups over the rest of the study. The other two short duration RCTs excluded from the meta-analysis were consistent with the pooled results. One study reported no between-group differences in median HbA1c.78 The second study was missing the number in each arm needed to calculate the between-group difference. Since this was an RCT, we calculated the between-group difference with the assumption of equal numbers in each arm which showed no statistically significant differences between-groups in HbA1c.79 (SOE: High; Neither drug favored)
- 83. 26 Figure 5. Pooledmean between-group difference in hemoglobin A1c comparing metformin with thiazolidinediones CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus Sulfonylureas In the prior report, we graded the evidence as high showing no differences in HbA1c between groups for this comparison. Therefore, we did not re-evaluate this comparison for HbA1c. Metformin Versus DPP-4 Inhibitors Six short duration RCTs (reported in nine articles) compared metformin with DPP-4 inhibitors (sitagliptin, alogliptin, linagliptin and saxagliptin).73, 80-87 These studies reported greater reductions in HbA1c with metformin (pooled between-group difference in HbA1c of - 0.4%; 95% CI, -0.5% to -0.3%) (Figure 6). No single study strongly influenced the meta-analysis results. In the three studies using both low and high metformin dosages compared with the
- 84. 27 maximum dose DPP-4inhibitor, we included the maximum dose metformin arm in the meta- analysis to make the drug dosages most comparable. The lower dose metformin arms (1000 mg) compared with maximum dose DPP-4 showed no statistically significant between-group differences in HbA1c.84-86 Two RCTs (in five articles) were reported as extension studies.80, 81, 83, 85, 87 The shorter duration results were included in the meta-analysis, since their study durations were more similar to the other studies in the meta-analysis. The first RCT comparing metformin 1000 mg twice daily with sitagliptin 100 mg daily reported HbA1c at 24 weeks,80 54 weeks,81 and 104 weeks.85 The between-group difference in HbA1c of -0.5 percent favored metformin over sitagliptin at both 24 and 54 weeks of followup. At week 104, there was no significant difference between groups in HbA1c, but there were high and differential losses to followup among the arms (74% loss to followup in the sitagliptin arm and 48% in the metformin arm). The second 76-week study87 was an RCT initially reported at 24 weeks comparing metformin up to 1000 mg twice daily with saxagliptin 10 mg daily. In this study, the between-group difference of -0.3 in HbA1c non-significantly favored metformin at 24 weeks83 and statistically significantly favored metformin at 76 weeks (mean difference between-groups in HbA1c, -0.2%; 95% CI, -0.5% to -0.03%),87 which is consistent with the meta-analysis results. (SOE: High; Metformin favored) Figure 6. Pooled mean between-group difference in hemoglobin A1c comparing metformin with DPP-4 inhibitors CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus SGLT-2 Inhibitors Three short duration and one longer duration RCTs (reported in three articles) compared metformin with an SGLT-2 inhibitor, showing no consistent between-group differences in
- 85. 28 HbA1c among thestudies.88-90 We did not combine these studies in a meta-analysis due to dosing and study duration differences. Two of the short duration studies reported in one article compared metformin XR titrated to 2000 mg with dapagliflozin 5 mg in the first study, and compared metformin XR titrated to 2000 mg with dapagliflozin 10 mg in the second study.88 Both studies, each lasting 24 weeks, reported no significant between-group differences in HbA1c.88 The study comparing metformin XR to the lower dose dapagliflozin arm of 5 mg had a mean difference between-groups in HbA1c which favored metformin by 0.16 percent although non-significantly, and the study comparing metformin XR to the higher dose dapagliflozin arm of 10 mg did not favor either arm. The third study comparing a lower dose of metformin XR of 1500 mg daily with dapagliflozin 10 mg daily for 12 weeks favored the dapagliflozin arm (calculated mean between-group difference in HbA1c of 0.12%; 95% CI, 0.08% to 0.16%).89 The 90-week RCT comparing metformin 1000 mg twice daily with empagliflozin 10 mg daily and 25 mg daily reported no significant differences between groups in HbA1c.90 (SOE: Low; Neither drug favored) Metformin Versus GLP-1 Receptor Agonists Three studies, each lasting one year or less, compared metformin versus a GLP-1 receptor agonist, with no consistent between-group differences in HbA1c.73, 91, 92 We did not combine the studies in a meta-analysis due to study duration and dosing differences. Each study, lasting 24 to 52 weeks in duration, compared metformin at 1500 mg or higher to a GLP-1 receptor agonist (exenatide twice daily in one study, exenatide weekly in a second study, and dulaglutide weekly in a third study). Only one study had a borderline significant result, favoring dulaglutide 1.5 mg weekly over metformin titrated to 2000 mg daily after 52 weeks (calculated mean between-group difference in HbA1c of 0.2%; 95% CI, 0.0% to 0.4%).91 This same study also had a lower-dose dulaglutide arm at 0.75 mg weekly, which showed no significant difference in HbA1c when compared with metformin titrated to 2000 mg daily.91 (SOE: Low; Neither drug favored) Thiazolidinediones Versus Sulfonylureas Thiazolidinediones (pioglitazone and rosiglitazone) and sulfonylureas (glibenclamide, glimepiride, and glyburide) had similar effects on HbA1c in 15 short duration RCTs (pooled mean between-group difference of -0.04%; 95% CI, -0.13% to 0.06%) (Figure 7).60, 61, 63, 74, 93-103 In a sensitivity analysis, we found no single study influenced the results , and there was no substantial heterogeneity between studies. We excluded one short duration RCT from the meta- analysis, since it did not report a number for analysis in each arm.79 This open-label 12-week RCT compared rosiglitazone titrated to 4-8 mg daily with glipizide titrated to 5-15 mg daily and reported a greater reduction in HbA1c in the thiazolidinedione arm (-0.9%) compared with the glipizide arm (-0.3%).79 We excluded the ADOPT study from the meta-analysis due to its long duration (median followup of 4 years).50 As mentioned previously, this double-blind RCT evaluated the long-term glycemic control between metformin, rosiglitazone, and glyburide monotherapy as initial treatment for type 2 diabetic adults, and had a 62 percent, 63 percent, and 56 percent loss to followup, respectively, in each treatment arm. The between-group difference between rosiglitazone and glyburide favored rosiglitazone after 4 years (mean difference between-groups of -0.4%; 95% CI, -0.5% to -0.3%). Of note, glyburide reduced HbA1c more than rosiglitazone, initially. HbA1c then rose higher in the glyburide arm than in the rosiglitazone arm after 1.5 years. (SOE: High; Neither drug favored in the short-term. SOE: Insufficient for the long-term.)
- 86. 29 Figure 7. Pooledmean between-group difference in hemoglobin A1c comparing thiazolidinediones with sulfonylureas CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Thiazolidinediones Versus DPP-4 Inhibitors Three RCTs, each lasting less than 26 weeks, compared pioglitazone with the DPP-4 inhibitors alogliptin and sitagliptin with no clear between-group differences in HbA1c (range in between-group differences of -0.48% to 0.23%).48, 73, 104 We did not combine the studies due to dosing differences among the studies. The one RCT with maximal dosing in both arms (pioglitazone titrated to 45 mg daily in one arm and sitagliptin 100 mg daily in the other arm) favored pioglitazone over sitagliptin (between-group difference in HbA1c of -0.5%; 95% CI - 0.7% to -0.3%).73 The other two RCTs used maximum dose DPP-4 inhibitors compared with moderately-dosed pioglitazone at 30 mg daily and reported no significant between-group differences in HbA1c.48, 104 (SOE: Insufficient) Thiazolidinediones Versus GLP-1 Receptor Agonists Two comparably-dosed RCTs compared pioglitazone with exenatide in differing dosing regimens, with mixed results.73, 105 One double-blind, moderately-sized RCT compared pioglitazone titrated to 45 mg daily with exenatide 2 mg weekly.73 After 26 weeks, the mean
- 87. 30 between-group difference inHbA1c was -0.1% with a reported 98.3% CI of -0.15% to 0.35%.73 The second open-label RCT compared pioglitazone at 45 mg daily with exenatide 10 ug twice daily. After 48 weeks, the calculated mean between-group difference in HbA1c favored exenatide by 0.3% (95% CI, 0.0% to 0.6%).105 (SOE: Insufficient) Sulfonylureas Versus DPP-4 Inhibitors Three RCTs, each lasting 54 weeks or less, compared a sulfonylurea (glipizide or glimepiride) with a DPP-4 inhibitor (sitagliptin or linagliptin) with no clear between-group differences in HbA1c.106-108 We did not combine these studies in a meta-analysis due to dosing differences and study population differences. Two RCTs non-significantly favored sulfonylureas over the DPP-4 inhibitor arms (between-group differences in HbA1c of -0.22% and -0.28%).106, 108 The third RCT enrolled patients with moderate or severe renal insufficiency at baseline and compared glipizide (mean dose 7.7 mg) with sitagliptin at 25 or 50 mg daily, depending on the participant’s renal function.107 This study showed no significant between-group differences in HbA1c.107 (SOE: Insufficient) Sulfonylureas Versus GLP-1 Receptor Agonists Four RCTs (reported in five articles) compared sulfonylureas directly with a GLP-1 receptor agonist (all studies using liraglutide).109-113 Three of the four studies favored liraglutide over sulfonylureas.109, 110, 112, 113 We did not combine these trials in a meta-analysis due to dosing differences between studies. Only two of the four studies used comparable dosing in the two arms. The first reported no statistically significant differences between the two arms.111 The second RCT favored the GLP-1 arm (between-group difference in HbA1c of 0.6%; 95% CI, 0.4% to 0.8%, at 52 weeks, and 0.3%; 95% CI, 0.2% to 0.4%, at the 104-week followup).112, 113 The two other RCTs, lasting 24 and 52 weeks, significantly favored the liraglutide arm by 0.5% each;109, 110 yet both of these studies used relatively lower doses in the sulfonylurea arm compared with the liraglutide arm, making it difficult to discern drug differences versus dosing differences.109, 110 (SOE: Insufficient) DPP-4 Inhibitors Versus SGLT-2 Inhibitors Only one double-blind, moderately-sized RCT, lasting 24 weeks, compared the DPP-4 inhibitor sitagliptin at 100 mg daily with the SGLT-2 inhibitor empagliflozin at 10 mg and 25 mg daily.114 The lower dose empagliflozin arm showed no significant between-group differences in HbA1c when compared with sitagliptin 100 mg daily. The higher dose empagliflozin 25 mg arm was favored slightly, but not significantly, over sitagliptin 100 mg (between-group difference in HbA1c of 0.01%; 95% CI, -0.03% to 0.3%).114 (SOE: Insufficient) DPP-4 Inhibitors Versus GLP-1 Receptor Agonists Two short duration RCTs compared a DPP-4 inhibitor with a GLP-1 receptor agonist, favoring the GLP-1 receptor agonists.73, 115 The first double-blind, moderately-sized RCT compared sitagliptin at 100 mg daily with exenatide 2 mg weekly for 26 weeks (calculated between-group difference in HbA1c of 0.4%; 95% CI, 0.07% to 0.49%) favoring exenatide.73 A second open-label RCT, with 40 participants and lasting 24 weeks, compared sitagliptin at 50 mg daily with liraglutide titrated to 0.9 mg daily (calculated mean between-group difference in HbA1c of 1.3%; 95% CI, -0.6% to 3.2%) non-significantly favoring liraglutide.115 (SOE: Low; GLP-1 receptor agonists favored)
- 88. 31 Metformin Versus Metformin-BasedCombination Comparisons Metformin Versus a Combination of Metformin Plus a Thiazolidinedione Fourteen studies lasting less than one year compared metformin with the combination of metformin plus a thiazolidinedione (eight studies with rosiglitazone and six studies with pioglitazone)55, 59, 67, 116-126 and showed a greater improvement in HbA1c with the combination therapy, in all the studies. The pooled between-group difference for all the studies combined had marked heterogeneity, but the meta-regression and stratified meta-analysis results showed consistent superiority of combination therapy (Table 5). The baseline HbA1c and dosing differences between arms were significant sources of heterogeneity. Studies with higher baseline HbA1c (HbA1c > 8%) had greater between-group differences than studies with lower baseline HbA1c (HbA1c < 8%). Studies with smaller dosing differences between study arms had smaller between-group differences in HbA1c than studies with larger dosing differences between arms. One long study,127 with 80 weeks of followup, compared metformin titrated to 2000 mg daily with metformin plus rosiglitazone titrated to 2000/8 mg daily. In that study, with around 5 percent loss to followup, the adjusted mean between-group difference in HbA1c favored combination therapy by 0.5 percent, consistent with the results in the shorter studies. (SOE: High; Combination of metformin plus a thiazolidinedione favored) Table 5. Pooled mean between-group difference in HbA1c comparing metformin with a combination of metformin plus a thiazolidinedione stratified by baseline HbA1c and dosing differences Variables N of Studies WMD (95% CI) I 2 Summary Baseline HbA1c <8% 7 0.43% (0.23% to 0.63%) 79% Favored metformin + thiazolidinedione Baseline HbA1c >=8% 7 0.88% (0.73% to 1.04%) 18% Favored metformin + thiazolidinedione Small dosing differences between study arms* 4 0.25% (0.16% to 0.34%) 0% Favored metformin + thiazolidinedione Large dosing differences between study arms* 10 0.79% (0.64% to 0.95%) 57% Favored metformin + thiazolidinedione CI = confidence interval; HbA1c = hemoglobin A1c; WMD = weighted mean difference *Studies were grouped together that had similar between-group differences in study dosing between arms. This led to two categories: those studies with smaller and larger between-group differences in drug dosing. We used the DerSimonian and Laird random effects point estimate for the weighted mean difference of the large dosing differences since profile likelihood estimate results would not converge. Metformin Versus a Combination of Metformin Plus a Sulfonylurea Fifteen RCTs, each lasting less than one year, compared metformin with the combination of metformin plus a sulfonylurea, with all of the studies favoring the combination arm over monotherapy (pooled between-group difference, 0.9%; 95% CI, 0.7% to 1.2%) (Figure 8).47, 55, 128-140 No single study markedly influenced the results. Meta-regression was conducted due to substantial heterogeneity, but none of the a priori variables were found to be significant, including study duration, dosing differences, appropriate randomization, double blinding, baseline HbA1c, or whether the study reported on withdrawals and dropouts. The study by Blonde et al. showed the greatest between-group differences; this study used a high-dose combination and started with the highest baseline HbA1c compared with other studies.131 The
- 89. 32 study with thesmallest between-group difference underdosed the metformin arm substantially in the metformin plus sulfonylurea arm.55 Three of the six dose-response studies showed a dose- response gradient favoring greater reductions in HbA1c with a higher dose combination than with a lower dose combination.131, 132, 134 One crossover study initially showed a difference between groups at the first crossover and then a negative rebound effect when changing the combination to monotherapy.47 A study by Ahren et al. was excluded from the meta-analysis since the study duration was longer than the other studies.141 This study, lasting 104 weeks, compared metformin at > 1500 mg daily to the combination of metformin at > 1500 mg daily plus glimepiride (up to 4 mg daily), and showed a between-group difference in HbA1c of 0.63 percent, favoring the combination arm, which was consistent with the results of the shorter studies included in the meta-analysis. (SOE: High; combination of metformin plus a sulfonylurea favored) Figure 8. Pooled mean between-group difference in hemoglobin A1c comparing metformin with a combination of metformin plus a sulfonylurea CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 90. 33 Metformin Versus aCombination of Metformin Plus a DPP-4 Inhibitor Twenty-seven RCTs, each lasting one year or less, directly compared metformin with the combination of metformin plus a DPP-4 inhibitor, with all favoring the combination arm (pooled between-group difference of 0.65%; 95% CI, 0.60% to 0.70%) (Figure 9).51, 80, 83, 84, 86, 118, 126, 139, 142-160 No single study markedly influenced the results, and no substantial heterogeneity was identified. Three short studies were not included in the meta-analysis due to dosing differences in two studies161, 162 and median HbA1c being reported in the other study.163 Two RCTs had 1000 mg more metformin in the monotherapy arm compared with the combination arm; therefore, these two studies161, 162 had smaller between-group differences than the other studies. The 12-week study163 reporting median HbA1c described a non-significant between-group difference in median HbA1c of 0.9% (p=0.1) favoring the combination arm of metformin (>1000 mg daily) plus sitagliptin (100 mg daily) over metformin alone (>1000 mg daily). Four longer studies (two of which were extension studies), each lasting 76 to 104 weeks with 30 percent to 50 percent losses to followup, also compared metformin with metformin plus a DPP-4 inhibitor, with results consistent with the shorter studies.85, 87, 141, 164 All four favored the combination arm (pooled between-group difference in HbA1c of 0.53%; 95% CI, 0.47% to 0.59%) (Figure 9). No single study markedly influenced the results, and no substantial heterogeneity was found. (SOE: High; Combination of metformin plus a DPP-4 inhibitor favored in the shorter duration studies) (SOE: Moderate; Combination of metformin plus a DPP-4 inhibitor favored in the longer duration studies)
- 91. 34 Figure 9. Pooledmean between-group difference in hemoglobin A1c comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Nine RCTs (reported in eight articles), each lasting less than one year, compared metformin alone with metformin plus an SGLT-2 inhibitor, with all studies favoring the combination arm (pooled between-group difference in HbA1c, 0.61%; 95% CI, 0.52% to 0.71%) (Figure 10).88, 153, 156, 158, 165-168 No single study markedly influenced the results. Heterogeneity was identified
- 92. 35 attributable to theSchumm-Draeger study which had the smallest between-group difference in HbA1c. No clear design differences exist between this study and the other studies in the meta- analysis, so we included it in the meta-analysis. Consistent with the meta-analysis results, two additional RCTs, each lasting 102 weeks, had statistically significant between-group differences in HbA1c of 0.4 percent and 0.8 percent, favoring the combination arms.169, 170 (SOE: High; Combination of metformin plus a SGLT-2 inhibitor favored) Figure 10. Pooled mean between-group difference in hemoglobin A1c comparing metformin with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Five short RCTs, each lasting less than one year, and one RCT, lasting 2 years, compared metformin with metformin plus a GLP-1 receptor agonist (albiglutide, liraglutide, dulaglutide, and exenatide), with all studies significantly favoring the combination arm over the monotherapy arm (range in between-group differences in HbA1c of 0.5% to 1.3%).141, 159, 171-174 We did not combine these studies in a meta-analysis due to differences in baseline HbA1c, study duration, and drug dosing. The two studies with low mean baseline HbA1c of 6.3 percent and 7.2 percent had between-group differences in HbA1c of 0.5 percent,172, 173 and the four studies with higher mean baseline HbA1c of around 8.0 percent had between-group differences in HbA1c ranging from 0.8 percent to 1.3 percent.141, 159, 171, 174 The one study with a lower dose and higher dose
- 93. 36 combination arm showeda dose-response relationship, with a smaller between-group difference in HbA1c of 0.5 percent in the lower dose combination and a larger between-group difference in HbA1c of 0.9 percent with the higher dose combination arm.174 (SOE: Moderate; Combination of metformin plus a GLP-1 receptor agonist favored) Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea Eight comparably-dosed RCTs, each lasting less than one year, directly compared the combination of metformin plus a thiazolidinedione with metformin plus a sulfonylurea (pooled between-group difference in HbA1c of -0.06%; 95% CI, -0.19% to 0.06%) (Figure 11).175-182 No single study markedly influenced the results, and no substantial heterogeneity was found. We excluded four studies from the meta-analysis due to dosing concerns within the studies.55, 183-185 Two studies used lower doses in the metformin plus sulfonylurea arms than in the comparator arms and found between-group differences in HbA1c favoring the metformin plus thiazolidinedione arms (-0.3% in both studies).55, 183 Two additional studies used submaximal sulfonylurea in the metformin plus sulfonylurea arm; one of the two studies favored the metformin plus thiazolidinedione arm. A sensitivity analysis including these four studies in the meta-analysis showed no marked differences in the pooled estimate and confidence interval, but more heterogeneity. In the meta-analysis, we included the 18-month results from the Rosiglitazone Evaluated for Cardiovascular Outcomes in Oral Agent Combination Therapy for Type 2 Diabetes (RECORD) study, since the study duration was comparable to the other included studies.176 The RECORD study was a multicenter, open-label RCT evaluating 4,447 patients with type 2 diabetes and uncontrolled glycemia already on metformin or sulfonylurea monotherapy.49, 176 The investigators randomly assigned subjects to the addition of rosiglitazone or to a combination of metformin and sulfonylurea. They reported glycemic control at a mean of 18 months for the first set of participants and a mean of 5.5 years after the start of the study for all included subjects not lost to followup.49, 176 The between-group difference in HbA1c of -0.07 percent was small and not significant for the first 516 subjects with 18-month followup.176 In the article reporting on the mean followup of 5.5 years in 2,222 subjects, the between-group difference in HbA1c of -0.29 percent significantly favored metformin plus rosiglitazone over metformin plus sulfonylurea.49 (SOE: Moderate; Neither drug combination favored)
- 94. 37 Figure 11. Pooledmean between-group difference in hemoglobin A1c comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a sulfonylurea CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c; Met = metformin; SU = sulfonylurea; TZD = thiazolidinedione Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor Five short RCTs compared metformin plus rosiglitazone with the combination of metformin plus sitagliptin and slightly favored the metformin plus thiazolidinedione arms (pooled between- group difference in HbA1c, -0.12%; 95% CI, -0.21% to -0.02%) (Figure 12).118, 126, 186-188 No substantial heterogeneity was identified in the meta-analysis. Removing the study by Bergenstal et al.188 changed the confidence interval to non-significant (95% CI with Bergenstal et al. removed, -0.19% to 0.01%). This study188 was not qualitatively different than the other studies, so we included it in the overall meta-analysis. This meta-analysis may underestimate the effect of metformin plus thiazolidinedione over metformin plus DPP-4 inhibitors, since two of the studies used lower drug doses in the metformin plus thiazolidinedione arms.186, 187 (SOE: Moderate; Combination of metformin plus a thiazolidinedione favored)
- 95. 38 Figure 12. Pooledmean between-group difference in hemoglobin A1c comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c; Met = metformin; TZD = thiazolidinedione Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two short RCTs, with adequate dosing in both arms, compared metformin plus thiazolidinediones (pioglitazone or rosiglitazone) with metformin plus a GLP-1 receptor agonist (exenatide) and had conflicting results.188, 189 The 20-week RCT comparing a combination of metformin and rosiglitazone with the combination of metformin and exenatide showed no significant between-group differences in HbA1c (between-group difference, -0.1%; P = 0.7).189 The 26-week RCT comparing the combination of metformin and pioglitazone with the combination of metformin and weekly exenatide favored the metformin plus exenatide arm (mean difference in HbA1c, 0.3%; 95 CI, 0.05% to 0.55%).188 (SOE: Insufficient) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor Nine studies (reported in ten articles) compared the combination of metformin plus sulfonylurea with metformin plus a DPP-4 inhibitor. We combined four RCTs, each lasting 1 year or less, comparing metformin plus a sulfonylurea with metformin plus a DPP-4 inhibitor, and found no significant between-groups differences in HbA1c (pooled between-group difference, -0.09%; 95 CI, -0.21% to 0.03%) (Figure 13).190-193 However, all four RCTS used a moderate dose of sulfonylurea in the metformin plus sulfonylurea arms while using the maximum dose of the DPP-4 inhibitors. If we exclude the study by Nauck and colleagues,192 the
- 96. 39 pooled result wouldthen significantly favor the combination of metformin plus sulfonylurea (pooled between group difference -0.13%, 95% CI -0.24% to -0.02%). However, there is no clear difference between this study and the other studies. No other study substantially changed the meta-analysis results. One additional short RCT was excluded from the meta-analysis, since we were unable to calculate a measure of variability.139 This study reported a mean change from baseline in HbA1c that significantly favored the metformin plus sulfonylurea arm over the metformin plus DPP-4 inhibitor arm of -0.2%, despite a lower dose of the sulfonylurea. Five longer studies, lasting 104 weeks and with over 20 percent loss to followup, also compared the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor, and showed no significant pooled between-group difference in HbA1c (- 0.03%; 95% CI, -0.15% to 0.09%) (Figure 13).141, 194-197 However, all five RCTs titrated the sulfonylurea to a moderate dose and compared this to a fixed maximum dose of a DPP-4 inhibitor. One of the longer studies was an extension of a study included in the meta-analysis of the shorter studies.192 No single study strongly influenced the results. (SOE: Low; Neither drug combination favored for both shorter and longer duration studies when comparing moderate dose sulfonylureas plus metformin with maximum dose DPP-4 inhibitors plus metformin) Figure 13. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor, stratified by study duration CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c; Met = metformin; pl = profile likelihood estimate; SU = sulfonylurea Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95% CI for each study. The line at the bottom of the graph indicates the 95% CI for the profile likelihood pooled estimate.
- 97. 40 Combination of MetforminPlus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three RCTs (reported in five articles), each lasting 1 to 4 years, compared the combination of metformin plus a sulfonylurea with the combination of metformin plus a SGLT-2 inhibitor (empagliflozin, dapagliflozin, or canagliflozin).54, 198-201 All three studies lasting 2 years with 20 percent to 30 percent losses to followup favored the combination of metformin plus an SGLT-2 inhibitor (pooled between-group difference in HbA1c of 0.17%; 95% CI, 0.14% to 0.20%) (Figure 14).199-201 No single study markedly influenced the results, and no substantial heterogeneity was identified. While all three studies used the maximum fixed dose of the SGLT- 2 inhibitor, the sulfonylurea arms were all uptitrated to a moderate dose (mean glimepiride dose of 3 mg in one study, mean glimepiride dose of 5.6 mg in a second study, and a mean glipizide dose of 16 mg in the third study). One of the three studies also compared the combination of metformin plus a lower dose SGLT-2 inhibitor arm of canagliflozin 100 mg daily with the combination of metformin plus glimepiride (mean dose of 5.6 mg daily), reporting no significant between-group differences in HbA1c of 0.01%.198 The 1-year and 4-year study findings were consistent with the 2-year results shown in the meta-analysis.54, 198 (SOE: Moderate; Combination of metformin plus a SGLT-2 inhibitor favored) Figure 14. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c; Met = metformin; SGLT-2 = sodium-glucose co-transporter-2; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Six RCTs compared metformin plus sulfonylurea with metformin plus a GLP-1 receptor agonist, with conflicting results.53, 141, 202-205 While no clear source of heterogeneity was
- 98. 41 identified, intraclass differencescould be part of the reason for the conflicting results. Three short-duration RCTs, each lasting one year or less, compared metformin plus sulfonylurea with metformin plus exenatide, all favoring the combination of metformin plus sulfonylurea despite submaximal doses of sulfonylureas being compared with maximal doses of daily exenatide (pooled between-group difference in HbA1c, -0.26%; 95% CI, -0.48% to -0.03%) (Figure 15).202, 203, 205 No single study strongly influenced the results, and no substantial heterogeneity was identified. An additional longer duration RCT, excluded from the meta-analysis due to dosing and study duration differences, compared metformin plus low dose glimepiride (mean daily dose: 2 mg) with metformin plus high dose exenatide (mean daily dose: 17 micrograms) with about a 75 percent loss to followup among the treatment groups over 48 months. The primary outcome was time to treatment failure which was not clearly defined except to state that they were in line with the American Diabetes Association recommendations for requiring alternative treatment due to inadequate glycemic control. They also evaluated HbA1c using a mixed model repeated measures analysis at different time points and reported no significant between-group differences at 1 year. At 2 years, they reported a significant between-group difference in HbA1c, favoring the metformin plus exenatide group by 0.2 percent which was maintained at 3 years.53 Two RCTs also compared metformin plus a sulfonylurea with metformin plus other types of GLP-1 receptor agonists (albiglutide or liraglutide), with conflicting results.141, 204 These were excluded from the meta-analysis due to dosing, drug type, and study duration differences. The first 16-week RCT compared metformin plus glimepiride (titrated to 4 mg daily) with similarly dosed metformin plus liraglutide (titrated to 1.8 micrograms daily), favoring the combination of metformin plus sulfonylurea (mean between-group difference, -0.3%; 95% CI, -0.34% to - 0.27%).204 The 104-week RCT, with over 30 percent loss to followup, compared metformin plus submaximal dose glimepiride (titrated to 4 mg daily) with the combination of metformin plus maximum dose albiglutide (titrated to 50 mg weekly), favoring the metformin plus albiglutide arm (mean between-group difference in HbA1c, 0.3%; 95% CI, 0.1% to 0.5%).141 (SOE: Low; Combination of metformin plus exenatide favored; SOE: Insufficient for combination of metformin plus other GLP-1 receptor agonists)
- 99. 42 Figure 15. Pooledmean between-group difference in hemoglobin A1c comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus daily exenatide CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); GLP-1 agonists = glucagon-like peptide-1 receptor agonist (here, all exenatide); HbA1c = hemoglobin A1c; Met = metformin; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Basal Insulin One small, open-label RCT, lasting 48 weeks, compared the combination of metformin plus a sulfonylurea with the combination of metformin plus a basal insulin, showing no significant between-group differences in HbA1c of 0.1% (95% CI, -0.5% to 0.7%).206 Patients were kept on their prior metformin doses and were randomized to uptitration of glimepiride (mean daily dose of 4 mg) versus uptitration of insulin glargine (mean daily dose of 23 units). Uptitration was stopped after reaching fasting plasma glucose titration goals.206 (SOE: Insufficient) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Premixed Insulin Two 16-week RCTs compared metformin plus glibenclamide with the combination of metformin plus a premixed insulin analogue – insulin aspart 70/30 in one study and insulin lispro 75/25 in the other study, with different results.207, 208 These differences may have been due to differences in dosing of the medications. The RCT207 that showed no significant between-group differences in HbA1c (-0.11%, p = 0.238) reported the mean total dose for each combination arm, while the other RCT, which significantly favored the metformin plus premixed insulin analogue (insulin aspart 70/30) arm over the metformin plus sulfonylurea arm (between-group difference of 0.46%, p = 0.027), did not clearly report mean total or maximum doses.208 Another possible difference may have been the type of premixed insulin analogue. (SOE: Insufficient)
- 100. 43 Combination of MetforminPlus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Four short, sufficiently-dosed RCTs, each lasting one year or less, compared the combination of metformin plus a DPP-4 inhibitor (sitagliptin or saxagliptin) with the combination of metformin plus an SGLT-2 inhibitor (canagliflozin, empagliflozin, or dapagliflozin). The studies significantly favored the combination of metformin plus an SGLT-2 inhibitor (pooled between- group difference in HbA1c, 0.17%; 95% CI, 0.08% to 0.26%) (Figure 16).153, 156, 158, 209 No single study strongly influenced the results, and no substantial heterogeneity was found in the meta- analysis. One longer RCT, lasting 90 weeks and with less than 10 percent loss to followup, comparing metformin plus sitagliptin with metformin plus empagliflozin at maximum doses was consistent with the shorter studies’ pooled results, favoring slightly the metformin plus empagliflozin arm (mean between-group difference in HbA1c, 0.2%; 95% CI, 0.0% to 0.5%).90 (SOE: Moderate; Combination of metformin plus a SGLT-2 inhibitor favored) Figure 16. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c; Met = metformin; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three adequately-dosed RCTs, lasting one year or less, compared the combination of metformin plus sitagliptin with the combination of metformin plus a GLP-1 receptor agonist (liraglutide or exenatide). All three RCTs significantly favored the combination of metformin plus a GLP-1 receptor agonist (pooled between-group difference in HbA1c, 0.65%; 95% CI,
- 101. 44 0.54% to 0.75%)(Figure 17).159, 188, 210 No single study markedly influenced the meta-analysis results, and no substantial heterogeneity was identified. One longer study, lasting 104 weeks, compared metformin plus sitagliptin with metformin plus albiglutide, significantly favoring the metformin plus albiglutide arm by 0.4 percent, consistent with the shorter studies’ pooled results.141 (SOE: Moderate; Combination of metformin plus a GLP-1 receptor agonist favored) Figure 17. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); GLP-1 = glucagon-like peptide-1; HbA1c = hemoglobin A1c; Met = metformin Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 for the random-effects pooled estimate. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a Basal Insulin One moderately-sized, 24-week RCT compared metformin plus sitagliptin (100 mg daily) with metformin plus insulin glargine titrated to 0.5 units per kg, significantly favoring the metformin plus insulin glargine arm (mean between-group difference in HbA1c of 0.59%; 95% CI, 0.42% to 0.76%).211 (SOE: Low; Combination of metformin plus a basal insulin favored) Combination of Metformin Plus a GLP-1 Receptor Agonist Versus a Combination of Metformin Plus a Basal Insulin One 26-week RCT compared metformin plus exenatide (2 mg weekly) with metformin plus glargine insulin, with a reported between-group difference in HbA1c favoring the combination of metformin plus exenatide by -0.2% (95% CI, -0.3% to -0.02%).212 (SOE: Insufficient)
- 102. 45 Combination of MetforminPlus a GLP-1 Receptor Agonist Versus a Combination of Metformin Plus a Premixed Insulin One moderately-sized RCT, lasting 26 weeks, compared the combination of metformin plus exenatide (titrated to 20 micrograms) with the combination of metformin plus premixed insulin (titrated to glucose target, mean dose 28 units), showing no significant between-group difference in HbA1c of 0.14% (95% CI, -0.003% to 0.29%).213 (SOE: Insufficient) Combination of Metformin Plus a Basal Insulin Versus a Combination of Metformin Plus a Premixed Insulin Three RCTs directly compared the combination of metformin plus basal insulin with the combination of metformin plus premixed insulin, showing no between-group differences in HbA1c (pooled between-group difference, 0.3%; 95% CI, -0.3% to 0.9%) (Figure 18).214-216 No single study strongly influenced the results, and no substantial heterogeneity was found. (SOE: Low; Neither drug favored) Figure 18. Pooled mean between-group difference in hemoglobin A1c comparing a combination of metformin plus a basal insulin with a combination of metformin plus a premixed insulin CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); HbA1c = hemoglobin A1c; Met = metformin Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Strength of Evidence for Hemoglobin A1c The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 6, Table 7, and Table 8 and summarized in the Key Points. All studies were RCTs. Study limitations for most comparisons were low or medium with only three comparisons having high study limitations due to lack of blinding, lack of description of withdrawals and dropouts, or very high losses to followup. Where quality influences the results,
- 103. 46 we describe thatunder the appropriate comparisons. In general, we did not find strong differences in outcomes in the lower versus higher quality studies. We did not find any evidence of publication bias using the Begg’s and Egger’s tests in most of the comparisons for HbA1c. A few of the monotherapy versus combination therapy comparisons had a statistically significant results on the publication bias test; however, these comparisons are likely to be missing both large and small negative studies favoring monotherapy. Therefore, we did not feel these statistically significant results represented a true publication bias. We also did not find any evidence of publication bias or reporting bias in the grey literature review which would change the overall conclusions. The grey literature was consistent with our findings for each of the comparisons, except for two comparisons (metformin versus DPP-4 inhibitors and metformin plus sulfonylurea versus metformin plus SGLT-2 inhibitors) where each had one study with results conflicting with the published results. These two studies under-dosed one of the study arms, making it more likely that conflicting results were from differing doses as opposed to publication bias. Only three studies did not report a measure of dispersion; therefore, we were able to combine most of the studies in meta-analyses, where appropriate.
- 104. 47 Table 6. Strengthof evidence domains for monotherapy comparisons in terms of hemoglobin A1c among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. TZD 25 (7365) Medium Consistent Direct Precise Undetected High Neither drug favored; -0.04% (-0.11% to 0.03%) Metformin vs. SU ‡ NA NA NA NA NA NA High Neither drug favored; 0.1% (-0.1% to 0.3%) Metformin vs. DPP-4 inhibitors 6 (6700) Low Consistent Direct Precise Undetected High Metformin favored-0.43% (-0.55% to -0.31%) Metformin vs. SGLT-2 inhibitors 3 (1633) Low Inconsistent Direct Imprecise Undetected Low Neither drug favored Metformin vs. GLP-1 receptor agonists 3 (1089) Low Inconsistent Direct Imprecise Undetected Low Neither drug favored TZD vs. SU 17 (6212) Medium Consistent Direct Precise Undetected High Neither drug favored; -0.04% (-0.13% to 0.06%) TZD vs. DPP-4 inhibitors 3 (1686) Low Inconsistent Direct Imprecise Suspected † Insufficient ¶ Unable to determine TZD vs. GLP-1 receptor agonists 2 (1048) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine SU vs. DPP-4 inhibitors 3 (1271) Low Inconsistent Direct Imprecise Suspected ‡ Insufficient ǁ Unable to determine SU vs. GLP-1 receptor agonists 4 (2056) Medium Inconsistent Direct Imprecise Undetected Insufficient § Unable to determine DPP-4 inhibitors vs. SGLT-2 inhibitors 1 (899) Low Unable to determine Direct Imprecise Undetected Insufficient Unable to determine DPP-4 inhibitors vs. GLP- 1 receptor agonists 2 (860) Medium Consistent Direct Imprecise Undetected Low GLP-1 receptor agonist favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; NA = not applicable; SGLT-2 inhibitors = sodium-glucose co- transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating hemoglobin A1c. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 105. 48 ‡ We didnot re-evaluate hemoglobin A1c for the comparison of metformin with sulfonylureas, because we previously rated this comparison as having high strength of evidence.16 ¶ For thiazolidinediones versus DPP-4 inhibitors, we graded the strength of evidence as insufficient, since there was only one comparably-dosed study which used maximum doses in each arm. We suspected reporting bias, since one study was found in the grey literature which favored thiazolidinediones while two of the three published studies showed no significant difference between-groups in hemoglobin A1c but underdosed the thiazolidinedione arms compared to the DPP-4 inhibitor arms. ǁ For sulfonylureas versus DPP-4 inhibitors, we graded the strength as insufficient, since two of the three studies minimally favored sulfonylureas while one study did not favor either medication. Two additional studies found in the grey literature minimally favored sulfonylurea. Probably, sulfonylurea is mildly favored overall. We will be able to form a more formative opinion by the final report, since these two studies will be included in the updated search we will do between the draft and final report. § For sulfonylurea versus GLP-1 receptor agonist, only two comparably-dosed studies were identified and each showed different results. The two non-comparably-dosed studies favored the GLP-1 receptor agonist.
- 106. 49 Table 7. Strengthof evidence domains for metformin versus metformin-based combination comparisons in terms of hemoglobin A1c among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + TZD 15 (6090) Medium Consistent Direct Precise Undetected High Metformin +TZD favored; range in pooled mean between-group differences in HbA1c, 0.3% to 0.9% Metformin vs. metformin + SU 17 (5210) Low Consistent Direct Precise Undetected High Metformin + SU favored; 0.9% (0.7% to 1.2%) Metformin vs. metformin + DPP-4 inhibitors (shorter duration studies) 30 (18,056) Medium Consistent Direct Precise Undetected High Metformin + DPP-4 inhibitor favored; 0.65% (0.6% to 0.7%) Metformin vs. metformin + DPP-4 inhibitors (longer duration studies) 4 (4013) Medium Consistent Direct Precise Undetected Moderate Metformin + DPP-4 inhibitor favored; 0.5% (0.47% to 0.6%) Metformin vs. metformin + SGLT-2 inhibitors 9 (5778) Low Consistent Direct Precise Undetected High Metformin + SGLT-2 inhibitor favored; 0.6% (0.5% to 0.7%) Metformin vs. metformin + GLP-1 receptor agonists 5 (2556) Medium Inconsistent Direct Precise Undetected Moderate Metformin + GLP-1 receptor agonist; range in between- group differences in HbA1c, 0.5% to 1.3% DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; HbA1c = hemoglobin A1c; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating hemoglobin A1c. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 107. 50 Table 8. Strengthof evidence domains for metformin-based combination comparisons in terms of hemoglobin A1c among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + TZD vs. metformin +SU 14 (3294) Medium Consistent Direct Precise Undetected Moderate Neither drug combination favored; -0.1% (-0.2% to 0.1) Metformin + TZD vs. metformin +DPP-4 inhibitors 5 (2413) Medium Consistent Direct Precise Undetected Moderate Metformin + TZD favored; -0.1% (-0.2% to -0.02%) Metformin + TZD vs. metformin +GLP-1 receptor agonists 2 (604) Low Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Metformin + SU vs. metformin +DPP-4 inhibitors (shorter duration studies) 5 (3300) Medium Inconsistent Direct Precise Undetected Low Neither drug combination favored Metformin + SU vs. metformin +DPP-4 inhibitors (longer duration studies) 5 (7270) High Consistent Direct Precise Undetected Low Neither drug combination favored Metformin + SU vs. metformin +SGLT-2 inhibitors (longer duration studies) 3 (3815) Low Consistent Direct Precise Undetected Moderate Metformin + SGLT-2 inhibitor favored; 0.2% (0.1% to 0.2%) Metformin + SU vs. metformin +GLP-1 receptor agonists 7 (4375) Medium 1. Consistent for Met + SU vs Met + exenatide 2. Inconsistent for Met + SU vs Met + other GLP-1 receptor agonist Direct Precise Undetected Low for #1 and insufficient for #2 1. Metformin + exenatide favored 2. Unable to determine Metformin + SU vs. metformin + basal insulin 1 (75) High Unable to determine Direct Imprecise Undetected Insufficient Unable to determine
- 108. 51 Table 8. Strengthof evidence domains for metformin-based combination comparisons in terms of hemoglobin A1c among adults with type 2 diabetes (continued) Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + SU vs. metformin + premixed insulin 2 (827) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Metformin + DPP-4 inhibitors vs. metformin +SGLT-2 inhibitors 4 (3423) Medium Consistent Direct Precise Undetected Moderate Metformin + SGLT-2 inhibitor favored; 0.2% (0.1% to 0.3%) Metformin + DPP-4 inhibitors vs. metformin +GLP-1 receptor agonists 4 (3322) Medium Consistent Direct Precise Undetected Moderate Metformin + GLP-1 receptor agonist favored; 0.7% (0.5% to 0.8%) Metformin + DPP-4 inhibitors vs. metformin + basal insulin 1 (515) Medium Unable to determine Direct Precise Undetected Low Metformin + basal insulin favored Metformin + GLP-1 receptor agonists vs. metformin + basal insulin 1 (321) Medium Unable to determine Direct Imprecise Undetected Insufficient Unable to determine Metformin + GLP-1 receptor agonists vs. metformin + premixed insulin 1 (363) High Unable to determine Direct Imprecise Undetected Insufficient Unable to determine Metformin + basal insulin vs. metformin + premixed insulin 3 (530) Medium Consistent Direct Imprecise Undetected Low Neither treatment favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating hemoglobin A1c. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 109. 52 Evidence for Weight MonotherapyComparisons Metformin Versus Thiazolidinediones In the prior report, we graded the evidence as high16 that metformin was significantly favored, with weight gain in the thiazolidinedione arms and weight loss in the metformin arms. Therefore, we did not re-evaluate this comparison for weight. (SOE: High; Metformin favored) Metformin Versus Sulfonylureas In the prior report, we graded the evidence as high16 that metformin was significantly favored, with weight gain in the sulfonylurea arms and mild weight loss in the metformin arms. Therefore, we did not re-evaluate this comparison for weight. (SOE: High; Metformin favored) Metformin Versus DPP-4 Inhibitors Six short RCTs (reported in nine articles) compared metformin with DPP-4 inhibitors, reporting greater reductions in weight with metformin (pooled between-group difference, -1.3 kg; 95% CI, -1.6 kg to -1.0 kg) (Figure 19).73, 80-87 No substantial heterogeneity was found in the meta-analysis, and no single study markedly influenced the results. Two RCTs (in three articles) were reported as extension studies.81, 85, 87 The extension studies, lasting 76 weeks and 104 weeks and with losses to followup ranging between 20 percent to 76 percent, all favored metformin over the DPP-4 inhibitors (between-group differences of -0.7 kg to -2.9 kg), consistent with the meta-analysis results from the shorter studies. Three RCTs had a lower dose and higher dose metformin arm.84-86 The higher dose metformin arms in two of the studies which compared metformin with alogliptin and sitagliptin both showed greater reductions in weight than the studies comparing lower dose metformin arms with alogliptin and sitagliptin.84, 85 The third RCT, comparing a low dose and high dose metformin arm with linagliptin, did not show this dose response.86 (SOE: High; Metformin favored)
- 110. 53 Figure 19. Pooledmean between-group difference in weight comparing metformin with DPP-4 inhibitors CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus SGLT-2 Inhibitors Two 24-week (reported in the same article) and one 90-week RCTs compared metformin with an SGLT-2 inhibitor (dapagliflozin or empagliflozin), showing greater reductions in weight with the SGLT-2 inhibitors (range of between-group differences in weight of -1.3 kg to -1.4 kg).88, 90 These between-group differences were statistically significant in two of the three RCTs.88 (SOE: Moderate; SGLT-2 inhibitors favored) Metformin Versus GLP-1 Receptor Agonists Three studies, each lasting one year or less, compared metformin with a GLP-1 receptor agonist, with conflicting effects on weight.73, 91, 92 We did not combine the studies in a meta- analysis due to study duration and dosing differences. Each of the three studies, lasting 24 to 52 weeks, compared metformin at 2000 to 2500 mg with a GLP-1 receptor agonist at maximum doses (exenatide 20 micrograms daily in the first study, exenatide 2 mg weekly in the second study, and dulaglutide 1.5 mg weekly in the third study). The first comparably-dosed, 26-week RCT of metformin titrated to 2500 mg daily compared with a fixed dose of 2 mg of exenatide weekly reported a mean between-group difference in weight of 0 kg (95% CI, -0.6 kg to 0.6 kg).73 The second 26-week RCT compared metformin titrated to 2000 mg with exenatide (10 micrograms twice daily).92 This RCT reported a mean between-group difference of 2.0 kg, favoring the GLP-1 receptor agonist arm (95% CI, 1.2 kg to 2.8 kg).92 The last 52-week RCT compared metformin titrated to 2000 mg with dulaglutide of 1.5 mg weekly, and reported a mean
- 111. 54 between-group difference inweight favoring the metformin arm of -0.7 kg (95% CI, -1.4 kg to - 0.03 kg).91 (SOE: Insufficient) Thiazolidinediones Versus Sulfonylureas Seven studies, each lasting one year or less, compared a thiazolidinedione to a sulfonylurea, showing higher weight gain in the thiazolidinedione arms, with a pooled between-group difference of 1.2 kg (95% CI, 0.6 kg to 1.8 kg) (Figure 20).61, 74, 94, 95, 100, 103, 217 No single study markedly influenced the results, and no substantial heterogeneity was found. One study showed a dose-response relationship between rosiglitazone and weight; patients treated with 4 mg per day of rosiglitazone gained 1.8 kg and those treated with 8 mg per day gained 3.0 kg, over 52 weeks compared with the glibenclamide arm which gained 1.9 kg.94 We excluded two RCTs from the meta-analysis due to their longer durations of 3 to 4 years.50, 52 Both RCTs had results consistent with the meta-analysis. As mentioned previously, the ADOPT study, with >50 percent loss to followup, evaluated the long-term glycemic control of metformin, rosiglitazone, and glyburide monotherapy as initial treatment for adults with type 2 diabetes, with weight as a secondary outcome.50 The between-group difference between rosiglitazone and glyburide was consistent with the results of the meta-analysis of the shorter studies, favoring sulfonylureas after approximately 5 years of followup (mean between-group difference, 2.5 kg; 95% CI, 2.0 kg to 3.1 kg). Of note, individuals in the glyburide arm gained weight over the first year and then stabilized, while those in the rosiglitazone arm had continued weight gain throughout the study. The second large, 3-year, multicenter study comparing pioglitazone with glibenclamide, also having > 50 percent losses to followup, showed a 5.2 kg weight gain in the pioglitazone-treated group and a 0.9 kg weight gain in the glibenclamide- treated group.52 (SOE: Moderate; Sulfonylurea favored)
- 112. 55 Figure 20. Pooledmean between-group difference in weight comparing thiazolidinediones with sulfonylureas CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Thiazolidinediones Versus DPP-4 Inhibitors Two 26-week RCTs compared thiazolidinediones with DPP-4 inhibitors; both studies significantly favored DPP-4 inhibitors with a mean between-group difference of 2.3 kg and 2.5 kg.73, 104 The thiazolidinedione arms increased weight by around 1.8 kg while the DPP-4 inhibitor arms decreased weight by around 0.5 kg. (SOE: Moderate; DPP-4 inhibitors favored) Thiazolidinediones Versus GLP-1 Receptor Agonists Two comparably-dosed RCTs compared thiazolidinediones (pioglitazone) with GLP-1 receptor agonists (exenatide), with both favoring exenatide by 3.5 kg. One double-blind moderately-sized RCT compared pioglitazone titrated to 45 mg daily with exenatide 2 mg weekly.73 After 26 weeks, the calculated between-group difference in weight favored exenatide by 3.5 kg (95% CI, 2.8 kg to 4.2 kg).73 The pioglitazone arm increased weight by 1.5 kg, and the exenatide arm decreased weight by 2 kg. The second open-label RCT compared pioglitazone at 45 mg daily with exenatide 10 ug twice daily. After 48 weeks, the calculated mean between- group difference in weight favored exenatide by 3.5 kg (95% CI, 2.4 kg to 4.6 kg).105 (SOE: Moderate; GLP-1 receptor agonists favored)
- 113. 56 Sulfonylureas Versus DPP-4Inhibitors Three RCTs, each lasting 54 weeks or less, compared a sulfonylurea (glipizide or glimepiride) with a DPP-4 inhibitor (sitagliptin or linagliptin) and favored the DPP-4 inhibitor arms (range in mean between-group differences of 0.9 kg to 1.8 kg).106-108 This difference was significant in two of the three studies; one study did not provide sufficient data to assess.108 We did not combine these studies due to dosing and study population differences. Sulfonylureas increased weight by about 1.2 kg, and the DPP-4 inhibitors decreased weight by around 0.4 kg, in these studies. (SOE: Low; DPP-4 inhibitors favored) Sulfonylureas Versus GLP-1 Receptor Agonists Four RCTs comparing sulfonylureas directly with liraglutide showed greater weight gain with a sulfonylurea (pooled mean between-group difference, 2.3 kg; 95% CI, 1.2 kg to 3.3 kg) (Figure 21).109-112 No single study strongly influenced the results. Substantial heterogeneity was found. Potential sources of heterogeneity were dosing differences, study duration differences, and differences in baseline weight. The one study with the largest between-group difference in weight112 lasted at least 24 weeks longer than the other two studies, used medications titrated to the maximum dose in both arms, and started with a higher baseline BMI. (SOE: Moderate; GLP- 1 receptor agonists favored) Figure 21. Pooled mean between-group difference in weight comparing sulfonylureas with GLP-1 receptor agonists CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); GLP-1 = glucagon-like peptide-1; kg = kilogram; pl = profile likelihood estimate Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The line at the bottom of the graph indicates the 95 percent confidence interval for the profile likelihood pooled estimate.
- 114. 57 DPP-4 Inhibitors VersusSGLT-2 Inhibitors One double-blind, moderately-sized, 24-week RCT compared the DPP-4 inhibitor, sitagliptin at 100 mg daily, with the SGLT-2 inhibitor, empagliflozin at 10 mg and 25 mg daily. The results significantly favored the empagliflozin arms (calculated mean between-group difference of 2.5 kg and 2.7 kg for the low dose and high dose empagliflozin arms, respectively).114 The sitagliptin-treated patients maintained weight, and the empagliflozin-treated patients decreased weight, over the 24 weeks. (SOE: Moderate; SGLT-2 inhibitors favored) DPP-4 Inhibitors Versus GLP-1 Receptor Agonists Two RCTs compared a DPP-4 inhibitor (sitagliptin) with a GLP-1 receptor agonist (exenatide or liraglutide), with both favoring the GLP-1 receptor agonist. The first double-blind RCT compared sitagliptin at 100 mg daily with exenatide 2 mg weekly for 26 weeks, with greater weight reduction in the exenatide arm (calculated mean between-group difference in weight of 1.2 kg; 95% CI, 0.5 kg to 1.9 kg).73 A second open-label RCT, with 40 subjects and lasting 24 weeks, compared sitagliptin at 50 mg daily with liraglutide titrated to 0.9 mg daily, with no significant difference in weight between groups (calculated mean between-group difference in weight of 1.5 kg; 95% CI, -24 kg to 27 kg).115 (SOE: Low; GLP-1 receptor agonists favored) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a Thiazolidinedione We combined six studies which directly compared metformin monotherapy with the combination of metformin plus a thiazolidinedione (mostly rosiglitazone), showing a pooled between-group difference in weight of -2.2 kg (95% CI, -2.6 kg to -1.9 kg) favoring metformin (Figure 22).59, 117, 118, 120, 123, 125 No single study markedly affected the results, and there was no significant heterogeneity between studies. All six studies showed that the metformin arms had weight loss while the combination arms had weight gain. Four studies were excluded from the meta-analysis due to insufficient quantitative data to combine the studies67, 116, 126 or due to study duration differences.127 All four of these studies reported modest weight gain in the combination arms, which was consistent with the studies included in the meta-analysis.67, 116, 126, 127 (SOE: High; Metformin favored)
- 115. 58 Figure 22. Pooledmean between-group difference in weight comparing metformin with a combination of metformin plus a thiazolidinedione CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus a Combination of Metformin Plus a Sulfonylurea Ten short RCTs compared metformin with the combination of metformin plus a sulfonylurea, favoring metformin monotherapy, with a pooled between-group difference of -2.2 kg (95% CI, - 3.4 kg to -1.0 kg) (Figure 23).128-130, 132-137, 139 No single study markedly influenced the results. While heterogeneity existed, all studies favored the metformin arm over the combination arm, with minimal between-group differences among the studies. In meta-regression, baseline weight was identified as a significant source of heterogeneity; dosing differences, double blinding, study duration, and appropriate randomization were not identified as significant. Baseline weight explained 55 percent of the between-study heterogeneity (adjusted r-squared = 55%). We present the stratified meta-analyses in Table 9. One 104-week study with 30 percent to 40 percent loss to followup, depending on the treatment arm, was excluded from the meta-analysis due to its long duration.141 The mean between-group difference in weight favored the metformin monotherapy arm, non-significantly, by 2.2 kg.141 (SOE: High; Metformin favored)
- 116. 59 Figure 23. Pooledmean between-group difference in weight comparing metformin with a combination of metformin plus a sulfonylurea CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The line at the bottom of the graph indicates the 95 percent confidence interval for the profile likelihood pooled estimate. Table 9. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus a sulfonylurea, stratified by baseline weight Variables N of Studies WMD (95% CI) I 2 Summary Baseline weight ≥ 90 kg* 5 -3.2 kg (-4.6 kg to -1.6 kg) 56% Favors metformin Baseline weight < 90 kg 5 -1.2 kg (-1.8 kg to -0.6 kg) 0% Favors metformin CI = confidence interval; kg = kilogram; WMD = weighted mean difference * Analysis was calculated using a profile likelihood estimate. Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor We combined twenty RCTs, each lasting 1 year or less, comparing metformin with the combination of metformin plus a DPP-4 inhibitor (pooled between-group difference of -0.1 kg; 95% CI, -0.24 kg to 0.03 kg) (Figure 24).51, 81, 83, 84, 86, 118, 139, 145, 146, 148, 150-155, 157-160 No substantial heterogeneity was found. In a standard sensitivity analysis, the removal of the study by Lavalle- Gonzalez and colleagues significantly changed the pooled estimate to significantly favor metformin monotherapy slightly. However, there were no clear qualitative differences to prompt removal of this study. Two studies147, 162 were excluded from the meta-analysis due to higher doses of metformin in the monotherapy arms by 500 to 1000 mg daily compared with the combination arms. In these two studies, greater weight loss was seen in the metformin
- 117. 60 monotherapy arms.147, 162 Sixstudies with similar results to the other studies were excluded from the short duration meta-analysis due to absence of data needed to quantitatively combine the studies.126, 143, 144, 149, 156, 163 We also pooled three longer studies (two of which were extension studies), each lasting 76 to 104 weeks and with greater than 20 percent losses to followup, that compared metformin with a metformin plus a DPP-4 inhibitor. Consistent with the short studies, these trials showed no significant difference in weight (pooled between-group difference in weight of 1.1 kg; 95% CI, - 2.3 kg to 0.07 kg) (Figure 24).85, 87, 141 No single study markedly influenced the results, and no substantial heterogeneity was found. (SOE: Moderate; Neither favored for studies 52 weeks or shorter) (SOE: Low; Neither drug favored for studies 1.5 to 2 years) Figure 24. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus a DPP-4 inhibitor, stratified by study duration CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 118. 61 Metformin Versus aCombination of Metformin Plus an SGLT-2 Inhibitor Seven RCTs (reported in six articles), each lasting 26 weeks or less, compared metformin alone with the combination of metformin plus an SGLT-2 inhibitor, with greater weight reductions in the combination arm (pooled between-group difference in weight of 2.0 kg; 95% CI, 1.5 kg to 2.5 kg) (Figure 25).88, 153, 158, 165, 166, 168 No single study markedly influenced the results. There was substantial statistical heterogeneity among the studies, yet the individual between-group differences were similar among the studies. One additional short RCT was excluded, since it only reported percent change in weight, but it also favored the combination of metformin plus an SGLT-2 inhibitor.156 Two 102-week RCTs were excluded from the meta- analysis due to study duration differences.169, 170 These longer RCTs, with 20 percent to 40 percent losses to followup, also significantly favored the combination of metformin plus an SGLT-2 inhibitor, with the range in between-group differences in weight of 2.4 kg to 3.1 kg. (SOE: High; Combination of metformin plus a SGLT-2 inhibitor favored) Figure 25. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram; pl = profile likelihood estimate; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The line at the bottom of the graph indicates the 95 percent confidence interval for the profile likelihood pooled estimate. Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Five RCTs, each lasting 48 weeks or less, compared metformin with the combination of metformin plus a GLP-1 receptor agonist, with all five studies showing greater weight reduction in the combination metformin plus GLP-1 receptor agonist arm (pooled between-group
- 119. 62 difference of 2.0kg; 95% CI, 1.3 kg to 2.7 kg) (Figure 26).159, 171-174 No one study strongly influenced the pooled results, and no substantial heterogeneity was identified. The one short study which had two dosing arms of the combination showed a smaller between-group difference in weight when comparing the metformin monotherapy arm to the lower dose combination arm versus the higher dose combination arm.159 One 104-week RCT excluded from the meta-analysis showed a non-significant greater weight reduction in the combination arm of 0.2 kg, but had 30 percent to 40 percent losses to followup, depending on the treatment arm.141 (SOE: Moderate; Combination of metformin plus a GLP-1 receptor agonist favored) Figure 26. Pooled mean between-group difference in weight comparing metformin with a combination of metformin plus a GLP-1 receptor agonist CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); GLP-1 = glucagon-like peptide-1; kg = kilogram Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea We combined six comparably-dosed studies, each lasting less than one year, that directly compared metformin plus a thiazolidinedione with metformin plus a sulfonylurea.175-178, 181, 218 The pooled mean between-group difference favored the combination of metformin plus sulfonylurea by 0.9 kg (95% CI, 0.4 kg to 1.3 kg) (Figure 27). No one study markedly influenced the results, and no substantial heterogeneity was found. In the meta-analysis, we included the short-term results from the large RECORD study.176 The RECORD study was a multicenter, open-label RCT evaluating 4,447 patients with type 2
- 120. 63 diabetes and uncontrolledglycemia already on metformin or sulfonylurea monotherapy, with less than 20 percent losses to followup.49, 176 Body weight increased significantly with rosiglitazone plus metformin compared with sulfonylurea plus metformin, with a mean between-group difference of 1.2 kg (95% CI, 0.4 kg to 2.0 kg) after 18 months.176 The mean between-group difference increased to 3.8 kg after 5 years of followup.49 We excluded three short RCTs from the meta-analysis, since the dosing was not comparable with the other studies.180, 184, 185 One180 used a lower dose of metformin in the metformin plus sulfonylurea arm compared with a higher dose of metformin in the metformin plus thiazolidinedione arm. The two other short RCTs used low doses of a sulfonylurea in the metformin plus sulfonylurea arm.184, 185 As a sensitivity analysis, we included these three RCTs in the meta-analysis and noted no meaningful change in results (pooled mean between-group difference in weight of 0.8 kg).180, 184, 185 (SOE: Moderate; Combination of metformin plus a sulfonylurea favored) Figure 27. Pooled mean between-group difference in weight comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a sulfonylurea CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); Met = metformin; kg = kilogram; SU = sulfonylurea; TZD = thiazolidinedione Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor Four short-duration RCTs compared metformin plus a thiazolidinedione with the combination of metformin plus a DPP-4 inhibitor, favoring the combination of metformin plus a DPP-4 inhibitor (pooled mean between-group difference of 2.7 kg; 95% CI, 0.8 kg to 4.5 kg)
- 121. 64 (Figure 28).118, 186-188 Thepatients in the metformin plus DPP-4 inhibitor arms had a mean weight loss, and the patients in the metformin plus thiazolidinedione arms had a mean weight gain. No single study markedly influenced the results. Substantial heterogeneity was identified; however, all four studies favored the metformin plus DPP-4 inhibitor arms by 1.4 to 3.6 kg. We were unable to quantitatively explore heterogeneity due to the small numbers of studies, but there were differences in baseline weight and drug types. One additional short-duration RCT comparing metformin plus pioglitazone with metformin plus alogliptin was excluded from the meta-analysis, since it did not have sufficient quantitative data.126 The study reported a decrease in weight of 0.7 kg in the metformin plus alogliptin arms but only stated qualitatively that there was an increase in weight in the metformin plus pioglitazone arm, consistent with the direction of weight change in the other four studies.126 (SOE: Moderate; Combination of metformin plus a DPP-4 inhibitor favored) Figure 28. Pooled mean between-group difference in weight comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); Met = metformin; kg = kilogram; pl = profile likelihood estimate; TZD = thiazolidinedione Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The line at the bottom of the graph indicates the 95 percent confidence interval for the profile likelihood pooled estimate. Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two short, comparably-dosed RCTs compared metformin plus thiazolidinediones (pioglitazone or rosiglitazone) with metformin plus a GLP-1 receptor agonist (exenatide).188, 189 Both studies significantly favored the combination of metformin plus a GLP-1 receptor agonist (range in mean between-group differences in weight, 2.7 kg to 5.1 kg). Both studies had weight gain in the metformin plus thiazolidinedione arms and weight loss in the metformin plus GLP-1
- 122. 65 receptor agonist arms.(SOE: Moderate; Combination of metformin plus a GLP-1 receptor agonist favored) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor Nine RCTS (reported in ten articles) compared the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor.139, 141, 190-197 Both the shorter (1 year or less) and longer (2-year) studies favored the metformin plus DPP-4 inhibitor arms (Figure 29). The metformin plus sulfonylurea arms all had weight gain, and the metformin plus DPP-4 inhibitor arms all had weight loss or weight maintenance. The five RCTs, each lasting 1 year or less, had a pooled between-group difference in weight of 2.1 kg (95% CI, 1.8 kg to 2.3 kg)139, 190-193 and the five RCTs, each lasting 2 years and with greater than 30 percent losses to followup, had a pooled between-group difference of 2.4 kg (95% CI, 1.9 kg to 2.9 kg).141, 194-197 No single study markedly influenced the results. While substantial statistical heterogeneity was identified in the meta-analysis of the longer studies, all studies favored the combination with similar effects. (SOE: High; Combination of metformin plus a DPP-4 inhibitor favored)
- 123. 66 Figure 29. Pooledmean between-group difference in weight comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor, stratified by study duration CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram; Met = metformin; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three RCTs, each lasting 52 to 104 weeks, compared the combination of metformin plus a sulfonylurea (glimepiride or glipizide) with the combination of metformin plus an SGLT-2 inhibitor (canagliflozin, dapagliflozin, or empagliflozin).198-200 The combination of metformin plus an SGLT-2 inhibitor was strongly favored (pooled mean between-group difference in weight of 4.7 kg; 95% CI, 4.4 kg to 5.0 kg) (Figure 30). No single study markedly influenced the results, and no substantial heterogeneity was found. One of the RCTs, which had a lower and higher dose metformin plus canagliflozin arm, demonstrated a small dose-response effect in weight reduction, with a smaller mean between-group difference in weight in the lower dose arm of 4.4 kg versus 4.7 kg in the higher dose arm.198 Two extension studies, lasting 2 years201 and 4 years54 with 30 percent to 60 percent losses to followup, were also identified, showing consistent findings to the shorter studies (calculated mean between-group differences in weight of 4.4 kg
- 124. 67 favoring the combinationof metformin plus an SGLT-2 inhibitor). (SOE: High; Combination of metformin plus a SGLT-2 inhibitor favored) Figure 30. Pooled mean between-group difference in weight comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram; Met = metformin; SGLT-2 = sodium-glucose co-transporter-2; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Four RCTs, three lasting less than 1 year and one lasting 2 years, compared metformin plus sulfonylurea with metformin plus a GLP-1 receptor agonist.141, 203-205 These favored the combination of metformin and GLP-1 receptor agonist (range in mean between-group differences of 2.4 kg to 12.3 kg). All four RCTs showed weight loss with the combination of metformin and GLP-1 receptor agonists and weight gain with the combination of metformin and sulfonylureas. We did not combine these studies in a meta-analysis due to differences in drug dosing, drug type, and study duration. One RCT compared metformin plus a sulfonylurea with three different dosing arms of metformin plus liraglutide. The arms with lower doses of liraglutide had smaller between-group differences in weight relative to the higher dose arm.204 (SOE: Moderate; Combination of metformin plus a GLP-1 receptor agonist favored) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Basal Insulin One small, open-label RCT, lasting 48 weeks, compared the combination of metformin plus a sulfonylurea with the combination of metformin plus a basal insulin, favoring metformin plus sulfonylurea (mean between-group difference in weight of -1.7 kg; 95% CI, -3.1 kg to -0.3
- 125. 68 kg).206 Patients were kepton their prior metformin doses and were randomized to uptitration of glimepiride (mean daily dose of 4 mg) versus uptitration of insulin glargine (mean daily dose of 23 units) until a fasting plasma glucose target was reached. Individuals in the metformin plus sulfonylurea arm had no change in weight, and those in the metformin plus insulin glargine arm gained weight. (SOE: Low; Combination of metformin plus a sulfonylurea favored) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Premixed Insulin Two short RCTs compared metformin plus glibenclamide with the combination of metformin plus a premixed insulin analogue: insulin aspart 70/30 in one study and insulin lispro 75/25 in the other study.207, 208 Both studies favored the metformin plus sulfonylurea arms (range in between- group differences of -0.7 kg to -0.5 kg). One of the two studies showed a statistically significant difference. There was not a mean decrease in weight in any of the study arms. Of note, if we combine the three studies comparing metformin plus sulfonylurea with metformin plus a premixed or basal insulin (adding in the study described in the prior comparison), metformin plus sulfonylurea is favored, with a weighted mean between-group difference in weight of -0.67 kg (95% CI, -0.83 kg to -0.51 kg). No single study influenced the results and no substantial heterogeneity was identified. Since premixed and basal insulins may have similar effects on weight, it may be reasonable to combine these categories. (SOE: Low; Combination of metformin plus a sulfonylurea favored) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Five RCTs compared metformin plus a DPP-4 inhibitor with the combination of metformin plus an SGLT-2 inhibitor (range in mean between-group differences in weight of 1.8 kg to 3.6 kg).90, 153, 156, 158, 209 All five studies significantly favored the combination of metformin plus an SGLT-2 inhibitor. We did not combine these studies due to differences in study duration and reporting of the outcome. The first 12-week RCT reported a mean percent change in weight of - 0.6 percent in the metformin plus DPP-4 arm versus -3.4 percent in the metformin plus SGLT-2 inhibitor arm.156 The second 12-week RCT reported a mean between-group difference in weight of 1.8 kg (95% CI, 1.9 kg to 2.7 kg).153 The third 24-week RCT had a calculated mean between- group difference in weight of 2.4 kg (95% CI, 1.7 kg to 3.1 kg), favoring the metformin plus dapagliflozin arm over the metformin plus alogliptin arm.209 The two longer RCTs, lasting 52 to 90 weeks and with less than 20 percent losses to follow-up, reported mean between-group differences of 2.9 kg and 3.6 kg, favoring the metformin plus SGLT-2 inhibitor arms.90, 158 Two of the three studies with lower dose and higher dose combinations of metformin plus SGLT-2 inhibitor arm demonstrated a dose-response effect in weight reduction, with smaller between- group differences in the lower dose arms and larger between-group differences in the higher dose arms.90, 158 (SOE: Moderate; Combination of metformin plus a SGLT-2 inhibitor favored) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist We combined three short RCTs comparing the combination of metformin plus a DPP-4 inhibitor with the combination of metformin plus a GLP-1 receptor agonist.159, 188, 210 All three studies significantly favored the combination of metformin plus a GLP-1 receptor agonist (pooled mean between-group difference, 1.8 kg; 95% CI, 1.1 kg to 2.5 kg) (Figure 31).
- 126. 69 Individuals in botharms lost weight but the metformin plus GLP-1 receptor agonist decreased weight more than the metformin plus DPP-4 treatment. No single study markedly influenced the results. Moderate heterogeneity was identified, although studies and point estimates were relatively similar. The two studies with both low dose and high dose arms of the GLP-1 receptor agonist showed smaller between-group differences in weight for the lower dose arms (between-group differences for lower dose arms of 1.9 kg and 1.1 kg compared with the higher dose arm of 2.5 kg and 1.5 kg).159, 210 One 104-week RCT, with 30 percent to 40 percent losses to followup depending on the treatment arm, reported a non-significant between-group difference in weight of 0.4 kg (95% CI, -4.7 kg to 5.4 kg).141 (SOE: Moderate; Combination of metformin plus a GLP-1 receptor agonist favored) Figure 31. Pooled mean between-group difference in weight comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); GLP-1 = glucagon-like peptide-1; kg = kilogram; Met = metformin; pl = profile likelihood estimate Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The line at the bottom of the graph indicates the 95 percent confidence interval for the profile likelihood pooled estimate. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a Basal Insulin One 24-week RCT compared metformin plus sitagliptin (100 mg daily) with metformin plus insulin glargine titrated to 0.5 units per kg. The RCT significantly favored the metformin plus DPP-4 arm (mean between-group difference in weight of -1.5 kg; 95% CI, -2.1 kg to -0.9 kg).211 The metformin plus DPP-4 arm decreased weight, and the metformin plus insulin glargine arm increased weight. (SOE: Low; Combination of metformin plus a DPP-4 inhibitor favored)
- 127. 70 Combination of MetforminPlus a GLP-1 Receptor Agonist Versus a Combination of Metformin Plus a Basal Insulin One 26-week, comparably-dosed RCT compared metformin plus exenatide (2 mg weekly) with metformin plus glargine insulin (titrated based on blood sugars). This RCT showed a mean between-group difference of -4.7 kg, favoring the combination of metformin plus a GLP-1 receptor agonist.212 (SOE: Insufficient) Combination of Metformin Plus a GLP-1 Receptor Agonist Versus a Combination of Metformin Plus a Premixed Insulin One moderately-sized RCT, lasting 26 weeks, compared the combination of metformin plus exenatide (titrated to 20 micrograms) with the combination of metformin plus premixed insulin (titrated to glucose target, mean dose 28 units).213 The RCT significantly favored metformin plus exenatide (mean between-group difference in weight of -5.1 kg; 95% CI, -5.7 kg to -4.5 kg). The metformin plus GLP-1 receptor agonist arm decreased weight, and the metformin plus premixed insulin arm increased weight. (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored) Combination of Metformin Plus a Basal Insulin Versus a Combination of Metformin Plus a Premixed Insulin Three RCTs directly compared the combination of metformin plus basal insulin with the combination of metformin plus premixed insulin, showing no between-group differences in weight (pooled mean between-group difference of -1.8 kg; 95% CI, -7.8 kg to 4.2 kg) (Figure 32).214-216 No single study strongly influenced the results, and no substantial heterogeneity was found. (SOE: Low; Neither treatment favored)
- 128. 71 Figure 32. Pooledmean between-group difference in weight comparing a combination of metformin plus a basal insulin with a combination of metformin plus a premixed insulin CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); kg = kilogram; Met = metformin Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Strength of Evidence for Weight The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 10, Table 11, and Table 12 and summarized in the Key Points. All studies were RCTs. Study limitations for most comparisons were low or medium, with only two comparisons having high study limitations due to lack of blinding, lack of description of withdrawals and dropouts, and very high losses to followup. Where quality influences results, we describe that under the appropriate comparisons. In general, we did not find strong differences in outcomes in the lower-quality versus higher-quality studies. We did not find any evidence of publication bias in any of the comparisons for weight which would have impacted the results. We also did not find any evidence of publication bias or reporting bias in the grey literature review. Eleven studies did not report a measure of dispersion; however, addition of these studies would not have importantly changed our conclusions or the strength of evidence assessment. We considered weight a direct outcome, since patients care about weight independent of its cardiovascular effects. Several of the comparisons were downgraded due to imprecision. The comparisons were considered imprecise mainly due to the small and not clinically relevant between-group differences in weight of 2 pounds or less.
- 129. 72 Table 10. Strengthof evidence domains for monotherapy comparisons in terms of weight among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. TZD ‡ NA NA NA NA NA NA High Metformin favored; -2.6 kg (-4.1 to -1.2 kg) Metformin vs. SU ‡ NA NA NA NA NA NA High Metformin favored; -2.7 kg (-3.5 to -1.9 kg) Metformin vs. DPP- 4 inhibitors 6 (6700) Medium Consistent Direct Precise Undetected High Metformin favored; -1.3 kg (-1.6 to -1.0 kg) Metformin vs. SGLT-2 inhibitors 3 (1903) Medium Consistent Direct Imprecise Undetected Moderate SGLT-2 inhibitors favored; range of between-group differences, -1.3 to -1.4 kg Metformin vs. GLP- 1 receptor agonists 3 (1089) Low Inconsistent Direct Imprecise Undetected Insufficient Unable to determine TZD vs. SU 9 (6766) Medium Inconsistent Direct Precise Undetected Moderate SU favored; 1.2 kg (0.6 to 1.8 kg) TZD vs. DPP-4 inhibitors 2 (1475) Low Consistent Direct Precise Undetected Moderate DPP-4 inhibitors favored; range in between-group differences, 2.3 to 2.5 kg TZD vs. GLP-1 receptor agonists 2 (1048) Low Consistent Direct Precise Undetected Moderate GLP-1 receptor agonists favored; between-group differences for both studies, 3.5 kg SU vs. DPP-4 inhibitors 3 (1271) Low Consistent Direct Imprecise Undetected Low DPP-4 inhibitors favored SU vs. GLP-1 receptor agonists 4 (1157) Medium Consistent Direct Precise Undetected Moderate GLP-1 receptor agonists favored; 2.3 (1.2 to 3.3 kg) DPP-4 inhibitors vs. SGLT-2 inhibitors 1 (899) Low Unknown Direct Precise Undetected Moderate SGLT-2 inhibitors favored; between- group difference, 2.5 to 2.7 kg DPP-4 inhibitors vs. GLP-1 receptor agonists 2 (860) Low Consistent Direct Imprecise Undetected Low GLP-1 receptor agonists favored DPP-4 = dipeptidyl-peptidase 4; GLP-1 = glucagon-like peptide-1; SGLT-2 = sodium-glucose co-transporter 2; SU = sulfonylurea; TZD = thiazolidinedione * Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome.
- 130. 73 † Unless otherwisespecified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence. ‡ We did not re-evaluate weight for the comparisons of metformin with thiazolidinediones and metformin with sulfonylureas because we previously rated these comparisons as having high strength of evidence.16
- 131. 74 Table 11. Strengthof evidence domains for metformin versus metformin-based combination comparisons in terms of weight among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + TZD 10 (5102) Medium Consistent Direct Precise Undetected High Metformin favored; -2.2 kg (-2.6 to -1.9 kg) Metformin vs. metformin + SU 11 (3692) Medium Consistent Direct Precise Undetected High Metformin favored; -2.2 kg (-3.4 to -1.0 kg) Metformin vs. metformin + DPP-4 inhibitors (studies 1 year or shorter) 28 (16,837) Medium Consistent Direct Imprecise Undetected Moderate Neither treatment favored; -0.1 kg (-0.2 to 0.03 kg) Metformin vs. metformin + DPP-4 inhibitors (studies 1.5 to 2 years) 3 (3446) High Consistent Direct Imprecise Undetected Low Neither treatment favored Metformin vs. metformin + SGLT-2 inhibitors 10 (5978) Low Consistent Direct Precise Undetected High Metformin + SGLT-2 inhibitors favored; 2.0 kg (1.5 to 2.5 kg) Metformin vs. metformin + GLP-1 receptor agonists 6 (2882) Medium Consistent Direct Precise Undetected Moderate Metformin + GLP-1 receptor agonists favored; 2.0 kg (1.3 to 2.7 kg) DPP-4 = dipeptidyl-peptidase 4; GLP-1 = glucagon-like peptide-1; SGLT-2 = sodium-glucose co-transporter 2; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence. ‡ We did not re-evaluate weight for the comparisons of metformin with thiazolidinediones and metformin with sulfonylureas because we previously rated these comparisons as having high strength of evidence.16
- 132. 75 Table 12. Strengthof evidence domains for metformin-based combination comparisons in terms of weight among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + TZD vs. metformin + SU 9 (2928) Medium Consistent Direct Precise Undetected Moderate Metformin + SU favored; 0.9 kg (0.4 to 1.3 kg) Metformin + TZD vs. metformin + DPP-4 inhibitors 5 (2413) Medium Consistent Direct Imprecise Undetected Moderate Metformin + DPP-4 inhibitors favored; 2.7 kg (0.8 to 4.5 kg) Metformin + TZD vs. metformin + GLP-1 receptor agonists 2 (604) Low Consistent Direct Precise Undetected Moderate Metformin + GLP-1 receptor agonists favored; range in mean between-group differences, 2.7 to 5.1 kg Metformin + SU vs. metformin + DPP-4 inhibitors (studies < 1 year) 5 (3300) Low Consistent Direct Precise Undetected High Metformin + DPP-4 inhibitors favored; 2.1 kg (1.8 to 2.3 kg) Metformin + SU vs. metformin + DPP-4 inhibitors (2-year studies) 5 (7270) High Consistent Direct Precise Undetected Low Metformin + DPP-4 inhibitors favored Metformin + SU vs. metformin + SGLT-2 inhibitors (longer duration studies) 3 (3815) Low Consistent Direct Precise Undetected High Metformin + SGLT-2 inhibitors favored; 4.7 kg (4.4 to 5.0 kg) Metformin + SU vs. metformin + GLP-1 receptor agonists 4 (3304) Medium Consistent Direct Imprecise Undetected Moderate Metformin + GLP-1 receptor agonists favored; range in between-group differences 2.4 kg to 12.3 kg Metformin + SU vs. metformin + basal insulin 1 (75) Medium Unknown Direct Precise Undetected Low** Metformin + SU favored Metformin + SU vs. metformin + premixed insulin 2 (819) Medium Consistent Direct Imprecise Undetected Low** Metformin + SU favored
- 133. 76 Table 12. Strengthof evidence domains for metformin-based combination comparisons in terms of weight among adults with type 2 diabetes (continued) Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors 5 (3423) Medium Consistent Direct Precise Undetected Moderate Metformin + SGLT-2 inhibitors favored; range in between-group differences 1.8 to 3.6 kg Metformin + DPP-4 inhibitors vs. metformin + GLP-1 receptor agonists 4 (3322) Medium Consistent Direct Precise Undetected Moderate Metformin + GLP-1 receptor agonists favored; 1.8 kg (1.1 to 2.5 kg) Metformin + DPP-4 inhibitors vs. metformin + basal insulin 1 (515) Medium Unknown Direct Precise Undetected Low Metformin + DPP-4 inhibitors favored Metformin + GLP-1 receptor agonists vs. metformin + basal insulin 1 (321) Medium Unknown Direct Imprecise Undetected Insufficient Unable to determine effect Metformin + GLP-1 receptor agonists vs. metformin + premixed insulin 1 (363) High Unknown Direct Precise Undetected Low Metformin + GLP-1 receptor agonists favored Metformin + basal insulin vs. metformin + premixed insulin 3 (530) Medium Consistent Direct Imprecise Undetected Low Neither treatment favored DPP-4 = dipeptidyl-peptidase 4; GLP-1 = glucagon-like peptide-1; SGLT-2 = sodium-glucose co-transporter 2; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence. ** While each of these comparisons are rated low strength of evidence due to low numbers of studies, this would move to moderate if all three studies were combined in a row of metformin plus sulfonylurea versus metformin plus premixed or basal insulin, since all three studies were consistent, direct, precise, study limitations were medium, and reporting bias was undetected. Since premixed and basal insulins may have similar effects on weight, it may be reasonable to combine these categories.
- 134. 77 Evidence for SystolicBlood Pressure Monotherapy Comparisons Metformin Versus SGLT-2 Inhibitors We combined three short RCTs (reported in two articles) directly comparing metformin with a SGLT-2 inhibitor, favoring SGLT-2 inhibitors in systolic blood pressure reduction.88, 89 Each study compared metformin with dapagliflozin,88, 89 resulting in a pooled mean between-group difference in systolic blood pressure of 2.8 mmHg (95% CI, 2.6 mmHg to 3.0 mmHg) favoring SGLT-2 inhibitors over metformin (Figure 33). The metformin arms decreased mean systolic blood pressure by 0.4 mmHg to 1.8 mmHg, and the SGLT-2 inhibitors arms decreased mean systolic blood pressure by 4.0 mmHg to 6.4 mmHg. No single study markedly influenced the results, and there was no substantial heterogeneity. We excluded one study from the meta-analysis given its length.90 This trial was a randomized, open-label, 78-week extension of two shorter trials of empagliflozin. There was a non-significant between-group difference in systolic blood pressure of 3.7 mmHg (95% CI, -1.3 mmHg to 8.7 mmHg).90 The metformin arm increased the mean systolic blood pressure by 2 mmHg, and empagliflozin (10 mg) increased mean systolic blood pressure by 0.1 mmHg, and empagliflozin (25 mg) decreased mean systolic blood pressure by 1.7 mmHg. (SOE: Moderate; SGLT-2 inhibitors favored) Figure 33. Pooled mean between-group difference in systolic blood pressure comparing metformin with SGLT-2 inhibitors CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); mmHg = millimeters mercury; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 135. 78 Metformin Versus GLP-1Receptor Agonists Two RCTs compared metformin with a GLP-1 receptor agonist, with inconclusive results.73, 91 The first study was a 52-week RCT which compared two doses of dulaglutide with metformin; there was a 20 percent loss to followup in all arms. The RCT reported a non-significant between- group difference in systolic blood pressure of 1.7 mmHg (95% CI, -0.7 mmHg to 4.1 mmHg) with dulaglutide (0.75 mg weekly) and 0.9 mmHg (95% CI, -1.5 mmHg to 3.3 mmHg) with dulaglutide (1.5 mg weekly). The metformin arm increased mean systolic blood pressure by 1.0 mmHg, and the dulaglutide arms increased mean systolic blood pressure by 0.1 mmHg to 2.7 mmHg.91 The second study was a 26-week RCT which showed a systolic blood pressure reduction of 1.3 mmHg with exenatide, but did not provide systolic blood pressure results for the metformin arm.73 (SOE: Low; Neither drug favored) Thiazolidinediones Versus GLP-1 Receptor Agonists Two RCTs, ranging in duration from 26 to 48 weeks, compared pioglitazone with exenatide with inconsistent results.73, 105 One 26-week RCT reported a non-significant between-group difference in systolic blood pressure of 0.4 mmHg (95% CI, -2.1 mmHg to 2.9 mmHg).73 One 48-week RCT reported a significant between-group difference in systolic blood pressure of 3.0 mmHg (95% CI, 0.2 to 5.8 mmHg), favoring the GLP-1 arm.105 (SOE: Insufficient) Sulfonylureas Versus GLP-1 Receptor Agonists One 104-week RCT compared glimepiride with liraglutide, and showed a non-significant between-group difference in systolic blood pressure of 0.9 mmHg (95% CI, -1.5 mmHg to 3.2 mmHg) with 1.2 mg of liraglutide and 1.9 mmHg (95% CI, -0.5 mmHg to 4.2 mmHg) with 1.8 mg of liraglutide.113 Participants in all arms had a mean decrease in systolic blood pressure. There was high loss to followup of 50 percent to 60 percent, among all arms. (SOE: Low; Neither drug favored) DPP-4 Inhibitors Versus SGLT-2 Inhibitors One double-blind, 24-week RCT compared sitagliptin (100 mg daily) to empagliflozin (10 mg and 25 mg daily). This RCT reported significant between-group differences in systolic blood pressure of 3.4 mmHg (95% CI, 1.2 mmHg to 5.7 mmHg) with 10mg of empagliflozin and 4.2 mmHg (95% CI, 2.0 mmHg to 6.5 mmHg) with 25mg of empagliflozin, favoring the SGLT-2 inhibitor over the DPP-4 inhibitor.114 The DPP-4 inhibitor increased mean systolic blood pressure by 0.5 mmHg, and the empagliflozin decreased mean systolic blood pressure by 2.9 to 3.7 mmHg. (SOE: Low; SGLT-2 inhibitors favored) DPP-4 Inhibitors Versus GLP-1 Receptor Agonists Two RCTs, ranging in duration from 24 to 26 weeks, compared sitagliptin with a GLP-1 receptor agonist.73, 115 One 26-week RCT, compared sitagliptin (100 mg daily) with exenatide (2 mg weekly), reporting a non-significant between-group difference in systolic blood pressure of 0.5 mmHg (95% CI -2.0 mmHg to 3.0 mmHg).73 The DPP-4 inhibitor decreased mean systolic blood pressure by 1.8 mmHg, and the GLP-1 receptor agonist decreased mean systolic blood pressure by 1.3 mmHg. One 24-week RCT, comparing sitagliptin (50 mg daily) with liraglutide (titrated to 0.9 mg daily), had a small number of participants in each arm and a high withdrawal rate. They reported a non-significant between-group difference in systolic blood pressure of 6.9
- 136. 79 mmHg (95% CI,-13.2 mmHg to 27.0 mmHg). This study reported high loss to followup (16 of 56 participants withdrew).115 (SOE: Low; Neither drug favored) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Seven RCTs (reported in six articles), each lasting less than one year, compared metformin to a combination of metformin and a SGLT-2 inhibitor. All studies favored the combination arm (pooled mean between-group difference in systolic blood pressure of 4.4 mmHg; 95% CI, 2.9 mmHg to 6.0 mmHg) (Figure 34).88, 153, 156, 158, 165, 166, 168 The metformin arms did not have consistent effects on the change in mean systolic blood pressure, which ranged from -2.2 mmHg to 3.3 mmHg. However, the SGLT-2 inhibitor combination arms consistently decreased mean systolic blood pressure by 2.4 mmHg to 8.5 mmHg. No single study markedly influenced the results, and there was no substantial heterogeneity. Two 102-week studies compared metformin with a combination of metformin and a SGLT-2 inhibitor and favored neither arm.169, 170 The first study was a 102-week RCT which compared metformin (at least 1500 mg daily) with a metformin plus two different doses of dapagliflozin. This RCT showed a non-significant between-group difference in systolic blood pressure of 2.6 mmHg (95% CI, -1.6 mmHg to 6.8 mmHg) for the combination with 5mg of dapagliflozin and 1.8 mmHg (95% CI, -2.6 mmHg to 6.2 mmHg) for the combination with 10mg of dapagliflozin.170 The metformin arm increased mean systolic blood pressure by 1.5 mmHg, and the dapagliflozin arms decreased mean systolic blood pressure by 0.3 mmHg to 1.1 mmHg. There was a high loss to followup in this study (47% in the metformin arm and 30% to 40% in the combination arms). The second study was a 102-week RCT with over 20 percent loss to followup in both arms. This second RCT compared metformin at the dosage prior to enrollment with metformin plus dapagliflozin (10 mg daily) and showed a non-significant between-group difference in systolic blood pressure of 2.4 mmHg (95% CI, -1.5 mmHg to 6.3 mmHg).169 The metformin arm increased mean systolic blood pressure by 1.1 mmHg, and the metformin plus SGLT-2 inhibitor arm decreased mean systolic blood pressure by 1.3 mmHg. One 24-week RCT was excluded from the meta-analysis due to differences in medication dosing for the SGLT-2 inhibitor arm. In this study, metformin was compared with metformin plus 5mg of dapagliflozin, showing a between-group difference in systolic blood pressure of 1.1 mmHg (95% CI, -1.4 mmHg to 3.6 mmHg) with both arms decreasing mean systolic blood pressure.88 (SOE: High; Combination of metformin plus a SGLT-2 inhibitor favored for shorter studies; SOE: Low; Neither favored for longer studies)
- 137. 80 Figure 34. Pooledmean between-group difference in systolic blood pressure comparing metformin with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); mmHg = millimeters mercury; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three short RCTs, each lasting less than one year, compared metformin with a combination of metformin plus a GLP-1 receptor agonist. The pooled analysis showed a between-group difference in systolic blood pressure of 3.1 mmHg (95% CI, 1.4 mmHg to 4.9 mmHg), favoring the combination arm over the monotherapy arm (Figure 35).159, 171, 172 These studies used different combinations of metformin plus a GLP-1 receptor agonist (liraglutide, dulaglutide, or exenatide). The metformin arms did not show consistent effects, with the mean change in systolic blood pressure ranging from -3.0 mmHg to 1.1 mmHg; the GLP-1 receptor agonist arms consistently decreased mean systolic blood pressure by 1.7 mmHg to 6.8 mmHg. No single study markedly influenced the results, and there was no substantial heterogeneity. We excluded the 104-week RCT from the meta-analysis due to its long duration.141 Consistent with the meta-analysis results, this study reported a significant between-group difference in systolic blood pressure of 3.2 mmHg (95% CI, 0.03 mmHg to 6.4 mmHg), favoring the combination arm.141 The metformin arm increased mean systolic blood pressure by 2.2 mmHg, and the metformin plus albiglutide arm decreased mean systolic blood pressure by 1 mmHg. There were high losses to followup of 30 percent to 40 percent among both arms in this study. One 30-week RCT was not included in the meta-analysis because it did not provide quantitative blood pressure measurements. The study descriptively reported that no changes in
- 138. 81 systolic blood pressurewere observed between intervention arms, which is inconsistent with the meta-analysis results.174 (SOE: Moderate; Combination of metformin plus a GLP-1 receptor agonist favored) Figure 35. Pooled mean between-group difference in systolic blood pressure comparing metformin with a combination of metformin plus a GLP-1 receptor agonist CI = confidence interval; GLP-1 = glucagon-like peptide-1; ES = effect size (mean between-group difference in the change from baseline); mmHg = millimeters mercury Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist One 26-week RCT compared metformin plus pioglitazone with metformin plus weekly exenatide, reporting a non-significant between-group difference in systolic blood pressure of 2.0 mmHg (95% CI, -0.8 mmHg to 4.8 mmHg).188 The metformin plus pioglitazone combination decreased mean systolic blood pressure by 1.6 mmHg, and the metformin plus exenatide combination decreased mean systolic blood pressure by 3.6 mmHg. (SOE: Low; Neither combination favored) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three 104-week RCTs compared metformin plus a sulfonylurea with metformin plus a SGLT-2 inhibitor (canagliflozin, empagliflozin, or dapagliflozin), favoring SGLT-2 inhibitors in systolic blood pressure reduction.200, 201, 219 The pooled between-group difference in systolic blood pressure was 5.1 mmHg (95% CI, 4.2 mmHg to 6.0 mmHg) (Figure 36), favoring combinations with a SGLT-2 inhibitor for lowering blood pressure. No single study influenced
- 139. 82 the results, andno substantial heterogeneity was found. A 208-week extension study was consistent with the pooled results favoring greater systolic blood pressure reduction with the SGLT-2 inhibitor arm.54, 219 (SOE: High; Combination of metformin plus a SGLT-2 inhibitor favored) Figure 36. Pooled mean between-group difference in systolic blood pressure comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); Met = metformin; mmHg = millimeters mercury; SGLT-2 = sodium-glucose co-transporter-2; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Four RCTs comparing metformin plus a sulfonylurea with metformin plus a GLP-1 receptor agonist reported systolic blood pressure, favoring the combination of metformin plus a GLP-1 receptor agonist for lowering blood pressure.53, 141, 202, 204 These studies were not combined in a meta-analysis due to differences in the duration and absence of sufficient data.53 One 16-week RCT compared metformin plus glimepiride with metformin plus exenatide and showed a non- significant between-group difference in systolic blood pressure of 2.0 mmHg (95% CI, -12.5 mmHg to 16.5 mmHg) with both combinations decreasing mean systolic blood pressure.202 One 16-week RCT compared metformin plus glimepiride with metformin plus liraglutide, showing a significantly greater reduction in systolic blood pressure, by more than 3 mmHg, with the liraglutide combination arms compared with 0.91 mmHg in the glimepiride combination arm (P < 0.05).204 There was a differential loss to followup, with 46 percent to 59 percent, in the higher liraglutide dose arms compared with 16 percent in the metformin plus sulfonylurea arm. One 104-week RCT compared metformin plus glimepiride with metformin plus albiglutide, showing a significant between-group difference in systolic blood pressure favoring the GLP-1
- 140. 83 receptor agonist combinationby 2.5 mmHg (95% CI, 0.3 mmHg to 4.7 mmHg).141 The metformin plus glimepiride increased mean systolic blood pressure by 1.5 mmHg while the metformin plus albiglutide arm decreased mean systolic blood pressure by 1.0 mmHg. (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor We pooled four short-duration RCTs comparing metformin plus a DPP-4 inhibitor versus metformin plus a SGLT-2 inhibitor.153, 156, 158, 209 The pooled results showed a between-group difference in systolic blood pressure of 4.1 mmHg (95% CI, 3.6 mmHg to 4.6 mmHg) (Figure 37), favoring metformin plus SGLT-2 inhibitors. The DPP-4 inhibitor combinations changed mean systolic blood pressure by +0.3 mmHg to -1.8 mmHg, while the SGLT-2 inhibitor combinations consistently decreased mean systolic blood pressure by 3.5 mmHg to 8.5 mmHg. No single study markedly influenced the results, and there was no substantial heterogeneity. We excluded one 104-week RCT from the meta-analysis due to its long duration.90 The trial compared metformin plus sitagliptin with metformin plus empagliflozin. The trial showed a between-group difference in systolic blood pressure of 5.1 mmHg (95% CI, 1.0 mmHg to 9.2 mmHg) with metformin plus 10 mg of empagliflozin and 4.8 mmHg (95% CI, 0.7 mmHg to 8.9 mmHg) with metformin plus 25 mg of empagliflozin, favoring the combination arm with a SGLT-2 inhibitor.90 These results were consistent with the results from the meta-analysis. The metformin plus sitagliptin arm increased mean systolic blood pressure by 1.8 mmHg, and the metformin plus empagliflozin arms decreased mean systolic blood pressure by 3 mmHg to 3.3 mmHg. (SOE: Moderate; Combination of metformin plus a SGLT-2 inhibitor favored)
- 141. 84 Figure 37. Pooledmean between-group difference in systolic blood pressure comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; ES = effect size (mean between-group difference in the change from baseline); Met = metformin; mmHg = millimeters mercury; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Four RCTs comparing metformin plus a DPP-4 inhibitor with metformin plus a GLP-1 receptor agonist showed no clear differences between-groups.141, 159, 188, 210 We did not combine these RCTs in a meta-analysis because of differences in drug type and study duration. One 26-week RCT compared metformin plus sitagliptin with metformin plus exenatide and showed a between-group difference in systolic blood pressure of 4.0 mmHg (95% CI, 1.5 mmHg to 6.5 mmHg), favoring the combination arm with GLP-1 receptor agonists. The metformin plus DPP-4 inhibitor arm increased mean systolic blood pressure by 0.2 mmHg, and the metformin plus GLP-1 receptor agonist arm decreased mean systolic blood pressure by 3.6 mmHg.188 A 26- week RCT compared metformin plus sitagliptin with metformin plus liraglutide, showing a non- significant between-group difference in systolic blood pressure of 0.4 mmHg (95% CI, -2.0 mmHg to 2.7 mmHg) with metformin plus 1.2 mg of liraglutide and 0.2 mmHg (95% CI, -2.1 mmHg to 2.6 mmHg) with metformin plus 1.8 mg of liraglutide.210 All arms decreased mean systolic blood pressure. Among the newest studies, one 52-week RCT compared metformin plus sitagliptin with metformin plus dulaglutide and showed a non-significant between-group difference in systolic blood pressure of 0 mmHg (95% CI, -1.9 mmHg to 1.9 mmHg) with 0.75 mg of dulaglutide weekly and 0.3 mmHg (95% CI, -1.6 mmHg to 2.2 mmHg) with 1.5 mg of dulaglutide weekly.159 All arms decreased mean systolic blood pressure. One 104-week RCT compared metformin plus sitagliptin with metformin plus albiglutide and showed a non-significant
- 142. 85 between-group difference insystolic blood pressure of 1.2 mmHg (95% CI, -1.1 mmHg to 3.5 mmHg).141 The metformin plus DPP-4 inhibitor arm increased mean systolic blood pressure by 0.2 mmHg, and the metformin plus GLP-1 receptor agonist arm decreased mean systolic blood pressure by 1.0 mmHg. (SOE: Low; Neither combination favored) Strength of Evidence for Systolic Blood Pressure The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 13, Table 14, and Table 15 and summarized in the Key Points. All studies were RCTs. Study limitations for all the comparisons were low or medium. In general, we did not find strong differences in outcomes in the lower-quality versus higher-quality studies. We were unable to assess publication bias given the limited number of studies for each comparison for systolic blood pressure. We also did not find any evidence of publication bias or reporting bias in the grey literature review. We considered this outcome direct, since systolic blood pressure is strongly linked with important long-term clinical outcomes.220-222
- 143. 86 Table 13. Strengthof evidence domains for monotherapy comparisons in terms of systolic blood pressure among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. SGLT-2 inhibitors 4 (1651) Medium Consistent Direct Precise Undetected Moderate SGLT-2 inhibitors favored; 2.8 mmHg (2.6 to 3.0 mmHg) Metformin vs. GLP-1 receptor agonists 2 (820) Low Inconsistent Direct Precise Undetected Low Neither drug favored TZD vs. GLP-1 receptor agonists 2 (1048) Low Inconsistent Direct Imprecise Undetected Insufficient Unable to determine SU vs. GLP-1 receptor agonists 1 (746) Medium Unknown Direct Imprecise Undetected Low Neither drug favored DPP-4 inhibitors vs. SGLT-2 inhibitors 1 (899) Low Unknown Direct Imprecise Undetected Low SGLT-2 inhibitors favored DPP-4 inhibitors vs. GLP-1 receptor agonists 2 (860) Low Consistent Direct Precise Undetected Low Neither drug favored DPP-4 = dipeptidyl-peptidase 4; GLP-1 = glucagon-like peptide-1; SGLT-2 = sodium-glucose co-transporter 2; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 144. 87 Table 14. Strengthof evidence domains for metformin versus metformin-based combination comparisons in terms of systolic blood pressure among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + SGLT-2 inhibitors (shorter studies) 7 (3988) Low Consistent Direct Precise Undetected High Metformin + SGLT-2 inhibitors favored; 4.4 mmHg (2.9 to 6.0 mmHg) Metformin vs. metformin + SGLT-2 inhibitors (longer studies) 2 (728) Low Consistent Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + GLP-1 receptor agonists 5 (2688) Medium Consistent Direct Imprecise Undetected Moderate Metformin + GLP-1 receptor agonists favored; 3.1 mmHg (1.4 to 4.9 mmHg) DPP-4 = dipeptidyl-peptidase 4; GLP-1 = glucagon-like peptide-1; SGLT-2 = sodium-glucose co-transporter 2; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 145. 88 Table 15. Strengthof evidence domains for metformin-based combination comparisons in terms of systolic blood pressure among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + TZD vs. metformin + GLP-1 receptor agonists 1 (514) Low Unknown Direct Precise Undetected Low Neither combination favored Metformin + SU vs. metformin + SGLT-2 inhibitors (longer duration studies) 3 (3815) Low Consistent Direct Precise Undetected High Metformin + SGLT-2 inhibitors favored; 5.0 mmHg (4.2 to 6.0 mmHg) Metformin + SU vs. metformin + GLP-1 receptor agonists 4 (3049) Medium Consistent Direct Imprecise Undetected Low Metformin + GLP-1 receptor agonists favored Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors 5 (3423) Medium Consistent Direct Precise Undetected Moderate Metformin + SGLT-2 inhibitors favored; 4.1 mmHg (3.6 to 4.6 mmHg) Metformin + DPP-4 inhibitors vs. metformin + GLP-1 receptor agonists 4 (3322) Low Inconsistent Direct Imprecise Undetected Low Neither combination favored DPP-4 = dipeptidyl-peptidase 4; GLP-1 = glucagon-like peptide-1; SGLT-2 = sodium-glucose co-transporter 2; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 146. 89 Evidence for HeartRate Monotherapy Comparisons Metformin Versus SGLT-2 Inhibitors Two RCTs compared metformin with SGLT-2 inhibitors and showed no differences in heart rate between the arms.89, 90 One 12-week RCT compared metformin with dapagliflozin, showing a non-significant between-group difference in heart rate of 2.1 beats per minute (bpm) (95% CI, - 1.3 bpm to 5.5 bpm) with 5 mg of dapagliflozin and 1.1 bpm (95% CI, -2.4 bpm to 4.7 bpm) with 10 mg of dapagliflozin. The metformin arm increased mean heart rate by 1.1 bpm, and the SGLT-2 inhibitor arms decreased mean heart rate by 0.03 bpm to 1 bpm.89 The other 90-week RCT compared metformin and empagliflozin and reported that “reductions in blood pressure were not associated with increases in heart rate.”90 (SOE: Low; Neither drug favored) Metformin Versus GLP-1 Receptor Agonists Two RCTs compared metformin with a GLP-1 receptor agonist and showed no clear differences in heart rate between the arms.73, 91 One RCT, with 20 percent loss to followup, compared metformin with two doses of dulaglutide, over 52 weeks. There was a non-significant between-group difference in heart rate of 0.5 bpm (95% CI, -1.1 bpm to 2.1 bpm) with 0.75 mg of dulaglutide weekly and 0.7 bpm (95% CI, -0.9 bpm to 2.3 bpm) with 1.5 mg of dulaglutide weekly. All arms had an increase in mean heart rate.91 One 26-week RCT compared metformin with exenatide and showed a non-significant between-group difference in heart rate of 1.2 bpm (95% CI, -0.5 bpm to 2.9 bpm). The metformin arm had a mean heart rate increase of 0.3 bpm, and the GLP-1 receptor agonist arm had an increase of 1.5 bpm.73 (SOE: Moderate; Neither drug favored) Thiazolidinediones Versus GLP-1 Receptor Agonists One double-blind, 26-week RCT compared pioglitazone titrated to 45 mg daily with exenatide (2 mg weekly) and showed a between-group difference in heart rate of 3.2 bpm (95% CI, 1.3 bpm to 5.0 bpm).73 In this study, the pioglitazone arm decreased mean heart rate by 1.7 bpm, and the GLP-1 receptor agonist arm increased mean heart rate by 1.5 bpm. (SOE: Low; Thiazolidinediones favored) Sulfonylureas Versus GLP-1 Receptor Agonists One 104-week RCT compared glimepiride with liraglutide, 1.2 mg and 1.8 mg, and showed a non-significant between-group difference in heart rate of 1.4 bpm (95% CI, -0.2 bpm to 2.9 bpm) for the lower dose and 0.2 bpm (95% CI, -1.3 bpm to 1.8 bpm) for the higher dose, favoring glimepiride.113 The sulfonylurea arm increased the mean heart rate by 0.6 bpm, and the liraglutide arms increased mean heart rate between 0.9 bpm to 2.0 bpm. There was high loss to followup of 50 percent to 60 percent among all arms in this study. (SOE: Low; Neither drug favored) DPP-4 Inhibitors Versus SGLT-2 Inhibitors One 24-week RCT compared 100 mg of sitagliptin to empagliflozin, 10 mg and 25 mg, and showed a non-significant between-group difference in heart rate of 0.2 bpm (95% CI, -1.5 bpm
- 147. 90 to 1.9 bpm)with 10 mg of empagliflozin and 0.5 bpm (95% CI, -2.2 bpm to 1.2 bpm) with 25 mg of empagliflozin, favoring empagliflozin.114 The DPP-4 inhibitor arm increased mean heart rate by 0.2 bpm; lower dose empagliflozin increased mean heart rate by 0.02 bpm, and the higher dose decreased mean heart rate by 0.25 bpm. (SOE: Low; Neither drug favored) DPP-4 Inhibitors Versus GLP-1 Receptor Agonists One 26-week RCT compared sitagliptin with exenatide and showed a non-significant between-group difference in heart rate of 1.0 bpm (95% CI, -0.9 bpm to 2.9 bpm). Both arms increased mean heart rate.73 (SOE: Low; Neither drug favored) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three RCTs comparing metformin to a combination of metformin plus a SGLT-2 inhibitor had inconsistent results.156, 165, 169 We did not combine the studies in a meta-analysis due to differences in study duration. One 12-week RCT compared metformin to metformin plus canagliflozin. Compared to metformin, the between-group difference in heart rate was 1.9 bpm lower (95% CI, 1.5 bpm to 2.3 bpm) with metformin plus 100 mg of canagliflozin, 1.1 bpm (95% CI, 0.7 bpm to 1.5 bpm) lower with metformin plus 200 mg of canagliflozin, and 3.4 bpm (95% CI, 3.0 bpm to 3.8 bpm) lower with metformin plus 300 mg of canagliflozin.156 In this study, the mean heart rate increased by 1.7 bpm in the metformin arm and decreased by 1.7 bpm in the metformin plus 300 mg of canagliflozin arm. One 18-week RCT compared metformin with metformin plus canagliflozin and reported a between-group difference in heart rate of 0.9 bpm with 100 mg of canagliflozin daily and 1.4 bpm with 300mg of canagliflozin daily. Metformin caused no increase in mean heart rate, and metformin plus canagliflozin increased mean heart rate.165 A 102-week RCT, with over 20 percent loss to followup, compared metformin to metformin plus 10 mg of dapagliflozin and showed a non-significant between-group difference in heart rate of 0.1 bpm (95% CI, -2.8 bpm to 3.0 bpm). Both arms increased mean heart rate slightly.169 (SOE: Low; Neither drug favored) Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three RCTs compared metformin to a combination of metformin plus a GLP-1 receptor agonist, with conflicting results.141, 159, 174 The 104-week RCT compared metformin to metformin plus albiglutide and showed a non-significant between-group difference in heart rate of 1.0 bpm (95% CI, -1.2 bpm to 3.2 bpm), favoring metformin.141 A 26-week RCT compared metformin to metformin plus dulaglutide and showed a between-group difference in heart rate of 2.8 bpm (95% CI, 1.1 bpm to 4.5 bpm), favoring metformin. There was high loss to followup of at least 60 percent in all arms in this study.159 One 30-week RCT compared metformin with metformin plus exenatide. The study reported that no changes in heart rate were observed between intervention arms (no quantitative data available).174 (SOE: Insufficient)
- 148. 91 Metformin-Based Combination Comparisons Combinationof Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor We pooled three RCTs, each more than 1 year in duration, that compared metformin plus a sulfonylurea with metformin plus a SGLT-2 inhibitor (Figure 38).198-200 A 104-week extension study compared metformin and glimepiride with metformin and canagliflozin, showing a between-group difference of 0.9 bpm (insufficient data to calculate 95% CI). The glimepiride arm increased heart rate by 0.7 bpm, and the canagliflozin arm decreased heart rate by 0.2 bpm, which is consistent with the pooled results.198, 201 (SOE: Moderate; Combination of metformin plus a SGLT-2 inhibitor favored) Figure 38. Pooled mean between-group difference in heart rate comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor bpm = beats per minute; CI = confidence interval; ES = effect size (mean between-group difference in the change from baseline); Met = metformin; SGLT-2 = sodium-glucose co-transporter-2; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two RCTs compared metformin plus a sulfonylurea with metformin plus a GLP-1 receptor agonist, with conflicting results.53, 141 One RCT, with 75 percent losses to followup, reported that mean heart rate increased by 1.2 bpm (P = 0.024) with metformin plus exenatide but not with metformin plus glimepiride (0.6 bpm; P = 0.28), with no differences between groups at any time.53 One 104-week RCT, with 30 percent loss to followup, compared metformin plus glimepiride with metformin plus albiglutide and showed a between-group difference of 1.8 bpm (95% CI, 0.2 bpm to 3.4 bpm), favoring the metformin with glimepiride treatment.141 The
- 149. 92 sulfonylurea combination decreasedmean heart rate by 0.5 bpm, and the GLP-1 receptor agonist combination increased mean heart rate by 1.3 bpm. (SOE: Insufficient) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Two RCTs compared metformin plus a DPP-4 inhibitor to metformin plus a SGLT-2 inhibitor and showed no clear differences between groups in heart rate.90, 156 One 12-week RCT compared metformin plus sitagliptin with three dose strengths of metformin plus canagliflozin. The RCT showed a non-significant between-group difference in heart rate of 1.5 bpm (95% CI, 1.2 bpm to 1.8 bpm) with 100 mg of canagliflozin, 2.3 bpm (95% CI, 2.0 bpm to 2.6 bpm) with 200 mg of canagliflozin, and 0.0 bpm (95% CI, -0.4 bpm to 0.4 bpm) with 300 mg of canagliflozin.156 The DPP-4 inhibitor with metformin decreased mean heart rate by 1.7 bpm; 100 mg of canagliflozin with metformin decreased mean heart rate by 0.2 bpm, 200 mg increased mean heart rate by 0.6 bpm, and 300 mg decreased mean heart rate by 1.7 bpm. One 90-week RCT compared metformin plus sitagliptin to metformin plus empagliflozin and found no increase in heart rate accompanying blood pressure reduction.90 (SOE: Low; Neither combination favored) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three RCTs compared metformin plus a DPP-4 inhibitor with metformin plus a GLP-1 receptor agonist.141, 159, 210 These studies were not combined in a meta-analysis due to differences in study duration. The two short studies significantly favored the combination of metformin plus a DPP-4 inhibitor. One 26-week RCT compared metformin plus sitagliptin with metformin plus liraglutide and showed a between-group difference in heart rate of 3.0 bpm (95% CI, 1.4 bpm to 4.5 bpm) with 1.2 mg of liraglutide and 4.6 bpm (95% CI, 3.0 bpm to 6.1 bpm) with 1.8 mg of liraglutide. The DPP-4 inhibitor combination decreased mean heart rate by 0.64 bpm, and the GLP-1 combinations increased mean heart rate by 2.3 bpm and 3.9 bpm.210 A 52-week RCT, with over 60 percent loss to followup, compared metformin plus sitagliptin to metformin plus dulaglutide, showing a between-group difference in heart rate of 2.4 bpm (95% CI, 1.0 bpm to 3.8 bpm) with 0.75 mg of dulaglutide weekly and 2.7 bpm (95% CI, 1.3 bpm to 4.1 bpm) with 1.5 mg of dulaglutide weekly. The DPP-4 inhibitor combination decreased mean heart rate by 0.3 bpm, and the GLP-1 combinations increased mean heart rate by 2.1 and 2.4 bpm.159 The 104-week RCT, with over 30% loss to followup, compared metformin plus sitagliptin to metformin plus albiglutide and showed a non-significant between-group difference in heart rate of 0.5 bpm (95% CI, -1.2 bpm to 2.2 bpm). Both arms increased mean heart rate slightly.141 (SOE: Low; Combination of metformin plus a DPP-4 inhibitor favored) Strength of Evidence for Heart Rate The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 16, Table 17, and Table 18 and summarized in the Key Points. All studies were RCTs. Study limitations for all comparisons were low or medium. Where quality influenced the study results, we describe that under the appropriate comparisons. In general, we did not find strong differences in outcomes in the lower-quality versus higher- quality studies. We were unable to assess publication bias given the limited number of studies for
- 150. 93 each comparison forheart rate. We also did not find any evidence of publication bias or reporting bias in the grey literature review. We considered this outcome indirect, since there is limited evidence directly linking heart rate to mortality or other clinical outcomes, including among adults with diabetes.
- 151. 94 Table 16. Strengthof evidence domains for monotherapy comparisons in terms of heart rate among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. SGLT-2 inhibitors 2 (1048) Medium Consistent Indirect Imprecise Undetected Low Neither drug favored Metformin vs. GLP-1 receptor agonists 2 (820) Low Consistent Indirect Precise Undetected Moderate Neither drug favored TZD vs. GLP-1 receptor agonists 1 (820) Low Unknown Indirect Precise Undetected Low TZD favored SU vs. GLP-1 receptor agonists 1 (746) Medium Unknown Indirect Precise Undetected Low Neither drug favored DPP-4 inhibitors vs. SGLT-2 inhibitors 1 (899) Low Unknown Indirect Precise Undetected Low Neither drug favored DPP-4 inhibitors vs. GLP-1 receptor agonists 1 (820) Low Unknown Indirect Precise Undetected Low Neither drug favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 152. 95 Table 17. Strengthof evidence domains for metformin versus metformin-based combination comparisons in terms of heart rate among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + SGLT-2 inhibitors 3 (912) Low Inconsistent Indirect Imprecise Undetected Low Neither drug favored Metformin vs. metformin + GLP-1 receptor agonists 3 (2473) Low Inconsistent Indirect Imprecise Undetected Insufficient Unable to determine DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 153. 96 Table 18. Strengthof evidence domains for metformin-based combination comparisons in terms of heart rate among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + SU vs. metformin +SGLT-2 inhibitors (longer duration studies) 3 (3815) Low Consistent Indirect Precise Undetected Moderate Metformin + SGLT-2 inhibitor favored; mean between-group difference, 1.5 bpm (95% CI, 0.6 to 2.3 bpm) Metformin + SU vs. metformin +GLP-1 receptor agonists 2 (2078) Medium Inconsistent Indirect Imprecise Undetected Insufficient Unable to determine Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors 2 (1110) Medium Consistent Indirect Imprecise Undetected Low Neither combination favored Metformin + DPP-4 inhibitors vs. metformin + GLP-1 receptor agonists 3 (2808) Medium Inconsistent Indirect Precise Undetected Low Metformin + DPP-4 inhibitor favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled mean between-group differences (95 percent confidence intervals). We only include estimates for comparisons with high or moderate strength of evidence.
- 154. 97 Key Questions 2aand 2b: All-Cause Mortality and Macrovascular and Microvascular Outcomes Study Design and Population Characteristics One hundred and eighteen studies (in 141 publications) reported on the comparative effectiveness of oral diabetes medications on long-term outcomes of interest (Appendix D, Tables D5 to D9). Twenty studies occurred in North America, 19 studies occurred in Europe, 13 studies occurred in Asia; all others were multi-continent studies. Ninety-six studies were RCTs, with durations from 12 weeks to 5.5 years. Thirty-four of the RCTs lasted for at least one year. All studies specified intermediate, not long-term, outcomes as their primary outcome (see Key Question 1) but then also reported the incidence of one or more long-term outcomes (e.g., mortality), usually as an adverse event. Two studies used a cross-over design.223, 224 Eighty-two RCTs reported support from a pharmaceutical company. Eighteen of the 62 (29%) RCTs identified in this update did not report on rescue therapy; rescue therapy was allowed in 25 studies (40%) and was not allowed in 15 studies (24%). We also included 21 retrospective cohort studies and one case-control study; duration of followup ranged from 6 months to over 5 years (eight lasted less than 2 years, 12 lasted 2 years or longer, and one lasted at least 12 months but did not specify the mean followup). These studies analyzed data from 12 unique cohorts, including five studies from Danish national databases225-229 and one from the Saskatchewan Health Database.230 Seven of the observational studies were designed to explicitly evaluate cardiovascular outcomes.151, 225-229, 231 Six observational studies reported support from a pharmaceutical company. The mean age of participants ranged from 48 years to 75 years, with the majority of studies reporting a mean age in the upper 50s. About 50 percent of participants were female. Forty-seven studies did not report race or ethnicity. In the studies that reported race, the majority of the participants were Caucasians. Two RCTs reported greater than 25 percent African American participants,118, 232 and two studies reported 70 to 80 percent Hispanic participants.121, 186 Most trials excluded people with coexisting illness, such as renal, cardiovascular, or liver disease. Risk of Bias Ninety-six RCTs were included in this section, all of which were described as randomized. Fifty-eight percent of the trials described their randomization scheme and another 74 percent of the trials were described as being double-blinded. Forty-five percent of all double-blinded RCTs also described the steps taken to ensure blinding. The majority of the trials (87 percent) described the withdrawals and dropouts. Of the 11 RCTs with at least 2 years of followup, ten had over 20 percent losses to followup. Of the 21 observational studies included in this section, 100 percent reported characteristics of subjects and tests of interest, 95 percent reported actual P values, and 85 percent described the measurement of outcomes of interest. All studies described and adjusted for confounding factors and conducted statistical analyses. All of the observational studies described the number of participants who were lost to followup after the start of the period of observation.
- 155. 98 Key Points andEvidence Grades All-Cause Mortality All evidence on all-cause mortality was of low strength or insufficient. Cardiovascular Mortality Sulfonylurea monotherapy was associated with increased cardiovascular mortality compared with metformin monotherapy (relative risk range 1.5 to 1.7 from individual RCTs; range in risk differences, 0.1 to 2.9%; range in duration of follow up, 2.8 to 4.0 years). (SOE: Moderate) To date, there has been uncertainty about the cardiovascular benefits of diabetes medications as evidenced by the FDA labeling stating a lack of known lower macrovascular risk for any diabetes medications; still, all evidence on the comparative effectiveness of the included diabetes medications on cardiovascular mortality was of low strength or insufficient. Cardiovascular and Cerebrovascular Disease Morbidity To date, there has been uncertainty about the cardiovascular benefits of diabetes medications as evidenced by the FDA labeling stating a lack of known lower macrovascular risk for any diabetes medications; still, all evidence on the comparative effectiveness of the included diabetes medications on cardiovascular morbidity was of low strength or insufficient. Retinopathy, Nephropathy, and Neuropathy The evidence was low or insufficient for all comparisons, and almost all RCTs were short-term. Evidence for All-Cause Mortality Monotherapy Comparisons Metformin Versus Thiazolidinediones Randomized Controlled Trials Four RCTs, each lasting 24 to 52 weeks, compared the effects of metformin with pioglitazone and found similar risks of all-cause mortality in the metformin and pioglitazone arms with seven deaths across the studies (pooled OR for metformin versus pioglitazone, 0.91; 95% CI, 0.22 to 3.72) (Figure 39).62, 63, 73, 76 We found no evidence of statistical heterogeneity. Omission of any single study did not change the conclusions. Deaths were not described in the pioglitazone arm of one study, so we imputed “0” events in this arm for the meta-analysis.76 The pooled between-group difference in mortality for metformin versus placebo was 0.0% (95% CI, - 0.6 to 0.6%).
- 156. 99 Figure 39. Pooledodds ratio of short-term all-cause mortality comparing metformin with pioglitazone CI = confidence interval; Group 1 = metformin; Group 2 = thiazolidinediones; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Two RCTs compared the effects of metformin with rosiglitazone on all-cause mortality.50, 59 The A Diabetes Outcome Progression Trial (ADOPT) randomized participants with recently- diagnosed, untreated type 2 diabetes from 488 different centers in the United States, Canada, and Europe to rosiglitazone, metformin, or glyburide and had long-term follow up.50 Mortality was slightly lower in the metformin (31/1454; 2.1%) versus rosiglitazone (34/1456; 2.3%) arm with median followup of 4.0 years.50 The actual number of participants for which ADOPT ascertained mortality is unclear, and withdrawals were high across the arms: 37 percent (rosiglitazone) and 38 percent (metformin).50 The second trial was 32 weeks in duration and reported no deaths in the metformin or rosiglitazone arms.59 Observational Studies Two retrospective cohort studies compared the effects of initiating thiazolidinediones and metformin (Table 19).233, 234 One study found no significant difference in all-cause mortality for metformin and thiazolidinediones.233 The second study found a significantly increased risk of all- cause mortality among women but not among men for rosiglitazone versus metformin.234 [SOE: Low for pioglitazone (short-term mortality); Neither metformin nor pioglitazone favored] (SOE: Low for rosiglitazone; Metformin favored compared with rosiglitazone)
- 157. 100 Table 19. Observationalstudies comparing metformin with thiazolidinediones on all-cause mortality Author, Year Population Mean Followup Results Pantalone, 2009 233 Cleveland Clinic electronic health record system N not reported Not reported HR for rosiglitazone, 1.33, 95% CI, 0.93 to 1.91 HR for pioglitazone, 1.08, 95% CI, 0.78 to 1.51 Reference = metformin Wheeler, 2013 234 Veterans Health Administration 2004-2009 n=132,306 (metformin) n=3753 (rosiglitazone) 1.7 years (metformin) 1.4 years (rosiglitazone) HR for 185,360 men, 1.19; 95% CI, 0.95 to 1.49 HR for 7,812 women, 4.36; 95% CI, 1.34 to 14.20 Reference = metformin P interaction for gender = 0.034 CI = confidence interval; HR = hazard ratio Metformin Versus Sulfonylureas Randomized Controlled Trials Six RCTs compared the effects of metformin and a sulfonylurea on all-cause mortality (Table 20).50, 129, 130, 137, 138, 231 Two of these RCTs had long-term followup and were of medium quality: The smaller of these two RCTs was conducted in China and had 2.8 years of followup. All participants were required to have documented coronary heart disease. The authors reported more than double the risk of death for the glipizide (mean dose 28.3 mg) arm versus metformin (mean dose 1,400 mg) arm.231 Losses to followup were 21 percent in both arms of this study.231 In the other long-term RCT, ADOPT (described above), the absolute difference in mortality was 0.1 percent higher for the sulfonylurea (maximum dose 15 mg; mean dose not reported) arm versus the metformin (maximum dose 2,000 mg; mean dose not reported) arm. As noted, there were high withdrawal rates in this study: 38 percent (metformin) and 44 percent (glyburide). Median followup was less for the sulfonylurea (3.3 years) versus metformin (4.0 years) arm.50 The other four RCTs, judged to be at low risk of bias, lasted less than 30 weeks.129, 130, 137, 138 Three of these studies reported no deaths in either arm;129, 130, 138 the fourth study reported one death in the metformin arm and none in the sulfonylurea arm.137
- 158. 101 Table 20. Randomizedcontrolled trials comparing metformin with sulfonylureas on all-cause mortality Author, Year Mean Followup Number of Deaths (%): Metformin Versus Sulfonylurea Estimate of the Measure of Association (95% CI) (Metformin as Reference Group) Hong, 2013 231 2.8 years 7/156 (4.5) versus 14/148 (9.5) RR, 2.1* (0.9 to 5.1) OR, 2.2* (0.8 to 6.7) RD, 5%* (-0.8% to 10.7%) Kahn, 2006 50 4.1 years (median) 31/1454 (2.1) versus 31/1441 (2.2) RR, 1.0* (0.6 to 1.7) OR, 1.0* (0.6 to 1.7) RD, 0.02%* (-1.0% to 1.1%) Chien, 2007 138 16 weeks 0/17 (0.0) versus 0/17 (0.0) NR Garber, 2003 129 16 weeks 0/164 (0.0) versus 0/151 (0.0) NR Goldstein, 2003 130 18 weeks 0/76 (0.0) versus 0/84 (0.0) NR DeFronzo, 1995 137 29 weeks 1/210 (0.5) versus 0/209 (0.0) NR CI = confidence interval; NR= not reported; OR = odds ratio; RD = absolute risk difference; RR = relative risk * Calculated for this report from values published in the study. Observational Studies We identified eight relevant retrospective cohort studies based on four cohorts (Veterans Health Administration, n=3;234-236 Cleveland Clinic electronic health record, n=2;233, 237 Danish National Patient Health Registry, n=2;225, 229 and the Health Service Database of Lombardy, n=1238 ). All studies reported an increased risk of death for a sulfonylurea versus metformin (Table 21). (SOE: Low; Metformin favored for long-term mortality; Neither favored for short-term mortality)
- 159. 102 Table 21. Observationalstudies comparing metformin with sulfonylureas on all-cause mortality Author, Year Population Followup Number of Deaths (%): Metformin Versus Sulfonylurea Adjusted Results Kahler, 2007 235 Veterans’ Health Administration Diabetes Epidemiology Cohort 3 years 82 / 2988 (2.7%) versus 1005 / 19,053 (5.3%) OR, 0.87; 95% CI, 0.68 to 1.10 Reference = sulfonylurea Wheeler, 2013 234 Veterans’ Health Administration 222,258 p-years (metformin) 47,604 p-years (glipizide) 48,238 p-years (glibenclamide) 2107 / 132,306 versus 1121 / 28,957 (glipizide) and 912 / 28,156 (glibenclamide) HR for glipizide, 1.55; 95% CI, 1.43 to 1.67 HR for glibenclamide, 1.38; 95% CI, 1.27 to 1.5 Reference = metformin Wang, 2014 236 Veterans’ Health Administration – Respondents to Veterans Large Health Survey 1999 5.3 years NR HR 0.69; 95% CI, 0.6 to 0.79 Reference = sulfonylurea Pantalone, 2009 233 Cleveland Clinic EHR 8 years NR HR, 0.54; 95% CI, 0.46 to 0.64 Reference = sulfonylurea Pantalone, 2012 237 * Cleveland Clinic EHR 2.2 years (median) NR / 12,774 (metformin) NR / 4,325 (glipizide) NR / 4,279 (glyburide) NR / 2,537 (glibenclamide) HR for glipizide, 1.64; 95% CI, 1.39 to 1.94 HR for glyburide, 1.59; 95% CI, 1.35 to 1.88 HR for glibenclamide, 1.68; 95% CI, 1.37 to 2.06 Reference = metformin Schramm, 2011 229 National Patient Registry (Denmark) Previous MI 3.3 years (median) 213 / 2906 versus 141 / 660 (glipizide) 737 / 3894 (glimepiride) 265 / 1168 (glibenclamide) HR for glipizide, 1.53; 95% CI, 1.23 to 1.89 HR for glimepiride, 1.3; 95% CI, 1.11 to 1.51 HR for glibenclamide, 1.47; 95% CI, 1.22 to 1.76 Reference = metformin National Patient Registry (Denmark) No previous MI 91.5 weeks 1548 / 43,340 versus 947 / 6965 (glipizide) 4081 / 36,313 (glimepiride) 1546 / 12,495 (glibenclamide) HR for glipizide, 1.27; 95% CI,1.17 to 1.38 HR for glimepiride, 1.32; 95% CI, 1.24 to 1.4 HR for glibenclamide, 1.19; 95% CI, 1.11 to 1.28 Reference = metformin Andersson, 2010 225 National Patient Registry (Denmark) – patients with admission for heart failure (1997-2006) 844 days 239 / 688 (35%) versus 2344 / 3615 (65%) HR, 0.85; 95% CI, 0.75 to 0.98; P = 0.02) Reference = sulfonylurea Corrao, 2011 238 Health Services Database of Lombardy Mean followup 4.8 to 5.1 years NR / 21,810 versus NR / 48,627 HR, 1.37; 95% CI, 1.26 to 1.49 Reference = metformin CI = confidence interval; EHR = electronic health record; HR = hazard ratio; MI = myocardial infarction; NR = not reported; OR = odds ratio * This study population may overlap with Pantalone, 2009.233 Metformin Versus DPP-4 Inhibitors Five RCTs compared the effects of metformin with sitagliptin on all-cause mortality. Meta- analysis of the four RCTs with the most similar durations (24 to 76 weeks) showed no difference in all-cause mortality for DPP-4 inhibitors compared with metformin, based on 10 deaths across
- 160. 103 the studies (pooledOR, 0.53; 95% CI, 0.16 to 1.82) (Figure 40).73, 82, 86, 87 We did not find evidence of statistical heterogeneity. One of these RCTs did not report on deaths in the DPP-4 inhibitor arm, and we imputed “0” for this arm.73 The pooled risk difference for DPP-4 inhibitors versus metformin was -0.1% (95% CI, -0.6 to 0.4%). The fifth RCT compared three metformin arms with sitagliptin 100 mg over 104 weeks. Two deaths were reported in the patients that started on placebo and were switched to metformin at 24 weeks; one death was reported in the metformin arm using 1000 mg as its maximum dose; and no deaths were reported in the metformin arm using 2000 mg as its maximum dose; no deaths occurred in the sitagliptin arm.85 A single retrospective cohort study from the Danish National Patient Registry reported on mortality for metformin (3,024/83,528) and sitagliptin (49/1,228) with mean followup of 0.9 to 1.8 years. The adjusted risk ratio (RR) for metformin versus sitagliptin was 1.25 (95% CI, 0.92 to 1.71; P = 0.15).228 (SOE for short-term mortality: Low; Neither treatment favored) Figure 40. Pooled odds ratio of short-term all-cause mortality comparing metformin with DPP-4 inhibitors CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = dipeptidyl peptidase-4 inhibitors; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus SGLT-2 Inhibitors Four short-term RCTs (12 to 24 weeks in duration, reported in 3 articles) compared the effects of metformin to SGLT-2 inhibitors on all-cause mortality and found no difference for SGLT-2 inhibitors versus metformin (pooled OR, 0.97; 95% CI, 0.10 to 9.36) (Figure 41).88, 89, 239 Only two deaths were reported in the studies (one in a metformin arm88 and one in a SGLT-2 arm88 ). We did not observe significant statistical heterogeneity. Removal of any one study did not change the inference. The pooled risk difference for SGLT-2 inhibitors versus metformin was -0.0% (95% CI, -0.9 to 0.8%) (SOE: Low; Neither favored)
- 161. 104 Figure 41. Pooledodds ratio for short-term all-cause mortality comparing metformin with SGLT-2 inhibitors CI = confidence interval; Group 1 = metformin; Group 2 = sodium-glucose co-transporter-2 inhibitors; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. Metformin Versus GLP-1 Receptor Agonists Two RCTs compared all-cause mortality between metformin and GLP-1 receptor agonists.73, 91 In one, lasting 52 weeks, there were no deaths in the metformin or dulaglutide arms.91 In the other, lasting 36 weeks, one death was reported in the metformin arm (1/246, 0.4%), and deaths were not described in the exenatide once weekly arm.73 (SOE for short-term mortality: Low; Neither favored) Thiazolidinediones Versus Sulfonylureas Randomized Controlled Trials Three RCTs compared thiazolidinediones with sulfonylureas and reported on mortality. The ADOPT trial reported slightly more deaths in the rosiglitazone arm than in the glyburide arm (2.3% versus 2.2%, respectively; risk difference of 0.1% for rosiglitazone compared with metformin) with differential followup time (median 3.3 years for sulfonylurea and 4.0 years for rosiglitazone) and withdrawals (44% for sulfonylurea and 37% for rosiglitazone).50 Two short- term trials reported few deaths in either the thiazolidinedione or sulfonylurea arms: One RCT (N=598) reported no deaths in either the rosiglitazone (4 mg and 8 mg) or sulfonylurea arms at 52 weeks.94 A 56-week trial reported two deaths in the glyburide arm (2/251; 0.8%) and no deaths in the pioglitazone arm (0/251, 0%).95
- 162. 105 Observational Studies Two retrospectivecohort studies compared the effects of thiazolidinediones with sulfonylureas on all-cause mortality.233, 234 In the cohort from the Cleveland Clinic (N=20,450), individuals initiating pioglitazone had a statistically significant lower risk of death compared with those initiating a sulfonylurea (adjusted HR, 0.59; 95% CI, 0.43 to 0.81). Those initiating rosiglitazone did not have a statistically significant lower risk of death compared with those initiating a sulfonylurea (adjusted HR, 0.73; 95% CI, 0.51 to 1.02). Followup time was not specified.233 In the Veterans Health Administration cohort, glipizide and glibenclamide were each compared separately with rosiglitazone.234 Compared with rosiglitazone, the adjusted RR of death for glipizide users was 1.26 (95% CI, 1.00 to 1.58), and the adjusted RR for glibenclamide users was 1.09 (95% CI, 0.87 to 1.38).234 (SOE: Insufficient for comparison of sulfonylurea and pioglitazone) (SOE: Insufficient for comparison of sulfonylurea and rosiglitazone) Thiazolidinediones Versus DPP-4 Inhibitors Two RCTs compared pioglitazone with sitagliptin and reported on mortality.48, 73 The 12- week RCT (N=106) reported no deaths in either arm.48 The 36-week RCT (N=326) did not report on deaths in the pioglitazone or sitagliptin arms, although it did report on deaths in other study arms.73 Of note, Russell-Jones 2012, et al. did not use an intention-to-treat approach and had greater than 13 to 18 percent losses to followup across arms.73 (SOE: Low; Neither favored for short-term mortality) Thiazolidinediones Versus GLP-1 Receptor Agonists A single RCT, with 36 weeks of followup, did not report on deaths in the pioglitazone (n=163) or exenatide once weekly (n=248) arms although it did report on deaths in other study arms.73 (Not graded) Sulfonylureas Versus DPP-4 Inhibitors A single RCT reported seven and three deaths over 58 weeks in the glipizide (7/212, 3.3%) and sitagliptin (3/210, 1.4%) arms, respectively.107 The authors did not use an intention-to-treat approach for mortality, and losses to followup were greater than 19 percent for both arms.107 (SOE: Low; DPP-4 inhibitors favored for short-term mortality) Sulfonylureas Versus GLP-1 Receptor Agonists Two RCTs compared sulfonylureas with liraglutide and reported on all-cause mortality.110, 113 Liraglutide doses varied across the trials, and death rates were low in both trials (Table 22).110, 113 In the longer study (104 weeks), mortality was higher in the sulfonylurea arm compared with the low-dose liraglutide arm (0.4% vs. 0.0%) but similar to that in the high-dose liraglutide arm (0.4%).113 (SOE: Insufficient)
- 163. 106 Table 22. Randomizedcontrolled trials comparing sulfonylureas with GLP-1 receptor agonists on all-cause mortality Author, Year Followup (Weeks) Sulfonylurea (Dose*) Liraglutide (Dose*) Number of Deaths / N (%) in the Sulfonylurea Arm Number of Deaths / N (%) in the Liraglutide Arm Kaku, 2011 110 52 Glibenclamide (fixed at 1.25 to 2.5 mg) Liraglutide (max 0.9 mg) NR/132 1/268 (0.4) Garber, 2011 113 104 Glimepiride (max 8 mg) Liraglutide (max 1.2 mg) 1/248 (0.4) 0/251 (0) Glimepiride (max 8 mg) Liraglutide (max 1.8 mg) 1/248 (0.4) 1/247 (0.4) GLP-1 = glucagon-like peptide-1; max = maximum; mg = milligrams * All doses were titrated, unless otherwise stated. DPP-4 Inhibitors Versus SGLT-2 Inhibitors Two RCTs compared sitagliptin with a SGLT-2 inhibitor (followup 24 to 26 weeks).114, 240 No deaths occurred in one study (N=670),114 and one death was reported in the sitagliptin arm (1/155, <1%) of the other study (no deaths in the SGLT-2 inhibitor arms; N=495).240 Neither study used an intention-to-treat approach for mortality, and losses to followup ranged from 3 to 13 percent across the arms of the trials.114, 240 (SOE: Insufficient) DPP-4 Inhibitors Versus GLP-1 Receptor Agonists A single RCT compared sitagliptin (n=163) with exenatide (n=248) and did not report on deaths in either arm, although it did report on deaths in other study arms.73 (SOE: Insufficient) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a Thiazolidinedione Six articles reported the results from seven RCTs (durations ranging from 24 to 80 weeks) on the effects of metformin versus metformin plus rosiglitazone on all-cause mortality.59, 119, 120, 123, 127, 241 The combined OR comparing metformin plus rosiglitazone with metformin was 2.51 (95% CI, 0.66 to 9.52; I2 = 0.0%) (Figure 42), showing a non-significant increased risk of death with metformin plus rosiglitazone (six deaths) compared with metformin monotherapy (one death). Removal of any one study did not impact substantially the effect size or confidence interval of the combined estimate. The pooled risk difference for the combination of metformin plus rosiglitazone versus metformin monotherapy was 0.3% (95% CI, -0.1 to 0.8%). (SOE for short-term mortality: Low; Metformin monotherapy favored over combination of metformin plus rosiglitazone)
- 164. 107 Figure 42. Pooledodds ratio of short-term all-cause mortality comparing metformin with a combination of metformin plus rosiglitazone CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus rosiglitazone; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. A single 24-week RCT compared the effects of metformin (n=103) to metformin plus pioglitazone (n=110) reported on all-cause mortality; no deaths occurred in either arm.125 (SOE: Insufficient for combination of metformin plus pioglitazone) Metformin Versus a Combination of Metformin Plus a Sulfonylurea Five RCTs, each ranging from 16 to 104 weeks, compared the effects of metformin with the combination of metformin plus a sulfonylurea on all-cause mortality.129, 130, 137, 138, 141 In the one long-term study, all-cause mortality was similar in the metformin (1/101, 1%) and metformin plus sulfonylurea (3/307, 1%) arms at 104 weeks; losses to followup were >30% in these arms.141 For the four short-term studies (16 to 29 weeks of followup), there were only three deaths and no significant difference for metformin plus a sulfonylurea versus metformin (pooled OR, 1.32; 95% CI, 0.09 to 18.56; I2 = 30.1%) (Figure 43). Although removal of one study did change the direction of the combined estimate (pooled OR, 0.33),129 removal of a single study did not substantially change the width of the confidence interval. The pooled risk difference for the combination of metformin plus a sulfonylurea compared with metformin monotherapy was 0.0% (95% CI, -1.0 to 1.0%). (SOE: Low; Neither treatment favored)
- 165. 108 Figure 43. Pooledodds ratio of short-term all-cause mortality comparing metformin with a combination of metformin plus a sulfonylurea CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a sulfonylurea; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Seventeen RCTs (published in 18 articles) comparing metformin plus a DPP-4 inhibitor to metformin monotherapy reported on all-cause mortality.84-87, 141, 142, 145, 146, 148, 149, 151-154, 158, 159, 161, 164 Three RCTs longer than one year (78 to 104 weeks) were not meta-analyzed because of differences in dosing of metformin in the maximally-dosed DPP-4 inhibitor arms.85, 87, 141 Mortality rates were low and did not differ by more than ~0.5 absolute percentage points between metformin and metformin plus DPP-4 inhibitor arms; the dose of medication did not appear to significantly affect results (Table 23).85, 87, 141 Losses to followup ranged from 20 to 48 percent across the arms of these studies. For studies 52 weeks or less, the pooled OR indicated no difference in mortality rates for metformin plus DPP-4 inhibitor versus metformin (pooled OR, 0.89; 95% CI, 0.28 to 2.86) (Figure 44).84, 86, 142, 145, 146, 148, 151-154, 158, 159, 161, 164 We did not find statistical heterogeneity (I2 = 0.0%). Removal of any one study did not change the direction of effect or inference, and there was no evidence of publication bias statistically (P = 0.80) using Harbord’s modified test. Three studies did not report on event rates in the metformin arm, and we imputed “0” events for these studies.146, 158, 159 The pooled risk difference for the combination of metformin plus a DPP-4 inhibitor compared with metformin monotherapy for short-term mortality was -0.0% (95% CI, - 0.3 to 0.3%).
- 166. 109 Figure 44. Pooledodds ratio for short-term all-cause mortality comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. Four RCTs included additional arms, with lower doses than the arms included in the meta- analysis. Results from these arms did not differ from those of the meta-analysis (Table 23).84, 86, 154, 164 We excluded one of the short-term RCTs from the meta-analysis because it underdosed the study drugs substantially; that 12-week study reported no deaths in either arm (Table 23).149 (SOE: Low; Neither favored for short-term mortality)
- 167. 110 Table 23. Randomizedcontrolled trials or arms of randomized controlled trials excluded from the meta-analysis comparing metformin with a combination of metformin plus a DPP-4 inhibitor on all- cause mortality Author, Year Followup (Weeks) Metformin Dose in Monotherapy Arm Metformin Dose in Combination Arm DPP-4 Inhibitor Dose in Combination Arm Number of Deaths / N (%) in Metformin Arm Number of Deaths / N (%) in Metformin + DPP-4 Inhibitor Arm Kadowaki, 2013 149 12 96% of participants on ≤750mg 94% of participants on ≤750mg Sitagliptin 50 mg 0/72 (0) 0/77 (0) Pratley, 2014 84 26 1000 mg 1000 mg Alogliptin 25 mg 0/109 (0) 0/106 (0) 1000 mg 2000 mg Alogliptin 25 mg 0/109 (0) 0/114 (0) 2000 mg 1000 mg Alogliptin 25 mg 0/111 (0) 0/106 (0) 2000 mg* 2000 mg Alogliptin 25 mg 0/111 (0) 0/114 (0) Haak, 2013 164 52 2000 mg 1000 mg Linagliptin 5 mg 1/170 (0.6) 2/225 (0.9) 2000 mg* 2000 mg Linagliptin 5 mg 1/170 (0.6) 1/171 (0.6) Haak, 2012 86 24 1000 mg 1000 mg Linagliptin 5 mg 0/144 (0) 0/143 (0) 1000 mg 2000 mg Linagliptin 5 mg 0/144 (0) 0/143 (0) 2000 mg 1000 mg Linagliptin 5 mg 1/147 (0.7) 0/143 (0) 2000 mg 2000 mg Linagliptin 5 mg 1/147 (0.7) 0/143 (0) Nauck, 2009 154 26 Mean 1868 mg Mean 1837 mg Alogliptin 12.5 mg 0/104 (0) 1/213 (0.5) Mean 1868 mg* Mean 1846 mg Alogliptin 25 mg 0/104 (0) 0/210 (0) Pfutzner, 2011 87 76 2000 mg 2000 mg Saxagliptin 5 mg 5/328 (1.5) 1/320 (0.3) 2000 mg 2000 mg Saxagliptin 10 mg 5/328 (1.5) 2/323 (0.6) Williams- Herman, 2010 85 104 2000 mg 1000 mg Sitagliptin 100 mg 2/176 (1.1) 1/190 (0.6) 1000 mg 1000 mg Sitagliptin 100 mg 1/182 (0.5) 1/190 (0.6) 1000 mg 2000 mg Sitagliptin 100 mg 1/182 (0.5) 1/182 (0.5) 2000 mg 1000 mg Sitagliptin 100 mg 0/182 (0) 1/190 (0.6) 2000 mg 2000 mg Sitagliptin 100 mg 0/182 (0) 1/182 (0.5) Ahren, 2014 141 104 ≥1500 mg ≥1500 mg Sitagliptin 100 mg 1/101 (1) 1/302 (0.3) DPP-4 = dipeptidyl peptidase-4; mg = milligrams *arm included in the meta-analysis Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Nine RCTs (in eight articles) compared metformin monotherapy with metformin plus an SGLT-2 inhibitor and reported on mortality (five deaths).88, 153, 158, 165, 166, 168-170 Two of the RCTs were long-term (102 weeks) and reported low rates of mortality across arms (one death in metformin arm in one study and one death in metformin plus SGLT-2 inhibitor arm in the other).169, 170 Losses to follow up were >20% across the arms of these studies.169, 170 Six of these studies were short (duration range, 12 to 24 weeks), including two trials described by Henry, et al;88 there was no difference in mortality between arms. The combined OR for all-cause mortality for metformin plus SGLT-2 inhibitor versus metformin was 1.14
- 168. 111 (95% CI, 0.18to 7.27) (Figure 45).88, 153, 158, 165, 166, 168 We did not find statistical heterogeneity (I2 = 0.0%). Removal of any one study did not change the overall inference. Two of these RCTs did not report on events in the metformin arm, and we imputed “0” events in these arms.165, 166 The pooled risk difference for the combination of metformin plus a SGLT-2 inhibitor compared with metformin monotherapy for short-term mortality was 0.0% (95% CI, -0.5 to 0.5%). Figure 45. Pooled odds ratio for short-term all-cause mortality comparing metformin with a combination of metformin plus an SGLT-2 inhibitor, stratified by study duration CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. Six of the nine RCTs comparing metformin with metformin plus a SGLT-2 inhibitor had multiple different dosing arms; event rates were low and did not appear to vary by dose.153, 158, 165, 166, 168, 170 (SOE: Low: Neither favored for short-term mortality; SOE: Low: Neither favored for long-term mortality) Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two RCTs compared metformin with the combination of metformin plus a GLP-1 receptor agonist. In the 52-week study, no deaths were reported in the metformin monotherapy arm (0/177, 0%); one death was reported in the metformin plus dulaglutide 1.5 mg weekly arm (1/304, 0.3%); and no deaths were observed in the dulaglutide 0.75 mg weekly arm (0/302, 0%) over 52 weeks.159 In a longer RCT, with 104 weeks of followup, one death in the metformin arm (1/101, 1%), and three deaths (3/302, 1%) in the metformin plus albiglutide arm were reported.141 (SOE: Low; Neither treatment favored)
- 169. 112 Metformin-Based Combination Comparisons Combinationof Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea Two multinational RCTs175, 177 directly compared the effect of the combination of metformin plus rosiglitazone with the combination of metformin plus a sulfonylurea. One study (N=596) reported two deaths in each arm (2/294, 0.7% in the rosiglitazone arm and 2/301, 0.7% in the sulfonylurea arm) over 52 weeks of treatment,175 and the other reported a fatal myocardial infarction in the metformin plus rosiglitazone arm (1/204, 0.5%) and no deaths in the metformin plus sulfonylurea arm (N=514) at 32 weeks.177 A single retrospective observational study of 80,936 patients with both Veterans Health Administration and Medicare coverage between 2000 and 2009 (minimum follow up, 12 month; mean followup, not reported) reported an increased mortality risk for patients taking the combination of metformin plus a sulfonylurea for at least 1 year compared with those on the combination of metformin plus a thiazolidinedione: adjusted HR, 1.5; 95% CI, 1.09 to 2.09; p=0.014.242 (SOE: Low; Neither metformin plus rosiglitazone nor metformin plus a sulfonylurea favored for short-term mortality) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor A single RCT compared the combination of metformin plus pioglitazone with the combination of metformin plus sitagliptin and reported one death in the metformin plus sitagliptin arm (1/172, 0.6%) and did not report on deaths in the metformin plus pioglitazone arm (n=172) at 26 weeks.188 (SOE: Insufficient) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist A single RCT compared the combination of metformin plus pioglitazone (n=172) with the combination of metformin plus weekly exenatide (n=170) at 26 weeks but only provided data on deaths in a third arm (metformin plus sitagliptin).188 Given the reporting in the metformin plus sitagliptin arm, we may infer that there were no deaths in the metformin plus pioglitazone and metformin plus exenatide arms, but this information was not reported.188 (SOE: Insufficient) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor Five RCTs compared the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor and reported on mortality at 104 weeks.194-197 The pooled OR for metformin plus a DPP-4 inhibitor versus metformin plus a sulfonylurea at 2 years was 0.64 (95% CI, 0.27 to 1.51) (Figure 46).141, 194-197 We did not find evidence of substantial statistical heterogeneity (I2 = 21%). The pooled risk difference for the combination of metformin plus a DPP-4 inhibitor compared with the combination of metformin plus a sulfonylurea was - 0.3% (95% CI, -0.8 to 0.2%).
- 170. 113 Figure 46. Pooledodds ratio for long-term all-cause mortality comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = combination of metformin plus a sulfonylurea; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; Met = metformin; OR = odds ratio; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Two additional RCTs evaluated this comparison but were not included in the meta-analysis because of their shorter durations.190, 193 One trial with 52 weeks of followup conducted among persons (predominantly men) greater than 65 years of age reported one death in each arm (1/360 (0.3%) in the metformin plus sulfonylurea arm; 1/360 (0.3%) in the metformin plus saxagliptin arm).193 The other trial had 30 weeks of followup and reported one death in the metformin plus sulfonylurea arm (1/519, 0.2%) and no deaths in the metformin plus sitagliptin arm (0/516, 0%).190 A single retrospective cohort study in the Danish National Registry reported a significantly decreased risk of death among metformin plus DPP-4 inhibitor users (n=11,138) versus metformin plus sulfonylurea users (n=25,092) with median follow up of 2.1 years (adjusted rate ratio, 0.65; 95% CI, 0.54 to 0.8).227 (SOE: Low; Combination of metformin plus a DPP-4 inhibitor favored for long-term mortality; SOE: Insufficient for short-term mortality) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three long-term RCTs (reported in four publications), each with a duration of 104 to 208 weeks, reported on all-cause mortality for this comparison.54, 199-201 Mortality rates were low across the studies. An extension of Nauck 2011, with extremely high losses to followup, reported a higher rate of mortality in the metformin plus sulfonylurea (5/408, 1.2%) versus metformin plus SGLT-2 inhibitor (2/406, 0.5%) arm at 208 weeks.54 Meta-analysis of the data from these trials at 104 weeks suggested that long-term all-cause mortality [which was low (<1%) across studies] was similar for metformin plus SGLT-2
- 171. 114 inhibitors and metforminplus sulfonylurea (pooled OR, 0.86; 95% CI, 0.29 to 2.55) (Figure 47).199-201 We did not find statistical heterogeneity (I2 = 14%), and removal of any one study did not change the inference of no difference between arms. The pooled risk difference for the combination of metformin plus an SGLT-2 inhibitor compared with metformin plus a sulfonylurea was -0.2% (95% CI, -0.8 to 0.5%). One of the RCTs evaluated metformin plus canagliflozin at 100 mg daily (versus 300 mg daily, which was included in the meta-analysis); mortality was the same as in the 300 mg arm (3/483, 0.6%).201 Of note, two200, 201 of three studies did not use an intention-to-treat approach and had large losses to followup across arms. (SOE: Low; Neither favored for long-term mortality) Figure 47. Pooled odds ratio for long-term all-cause mortality comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; Group 1 = combination of metformin plus a sulfonylurea; Group 2 = metformin plus a sodium-glucose co-transporter-2 inhibitor; Met = metformin; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two RCTs with more than two years of followup reported similar mortality rates for the combination of metformin plus a sulfonylurea compared to metformin plus a GLP-1 receptor agonist (Table 24).53, 141 Table 24. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a GLP-1 receptor agonist on all-cause mortality Author, Year Followup Number of Deaths/N (%) in the Metformin + Sulfonylurea Arm Number of Deaths/N (%) in the Metformin + GLP-1 Receptor Agonist Arm Ahren, 2014 141 104 weeks Glimepiride 3.1 mg: 3/307 (1.0) Albiglutide 40.5 mg: 3/302 (1.0) Gallwitz, 2012 53 48 months (assumed) Glimepiride 2.0mg: 5/508 (1.0) Exenatide 17.4 mg: 5/510 (1.0) GLP-1 = glucagon-like peptide-1; mg = milligrams; mean daily dose shown for glimepiride and exenatide; mean weekly dose shown for albiglutide
- 172. 115 A single retrospectivecohort study in the Danish National Registry did not find a significantly decreased risk of death among metformin plus GLP-1 receptor agonist users (n=4,345) versus metformin plus sulfonylurea users (n=25,092) over a median follow up of 2.1 years (adjusted rate ratio, 0.77; 95% CI, 0.51 to 1.17).227 (SOE: Low; Neither favored for long- term mortality) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Basal Insulin A single retrospective cohort study in the Danish National Registry found a significantly increased risk of death among metformin plus basal insulin users (n=6,858) versus metformin plus sulfonylurea users (n=25,092) over median follow up of 2.1 years (adjusted rate ratio, 1.95; 95% CI, 1.7 to 2.25).227 (Not graded) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Premixed Insulin Two multinational RCTs (N=938) compared the effect of the combination of metformin plus a sulfonylurea with the combination of metformin plus a premixed insulin (insulin aspart 70/30 in one study and insulin lispro 75/25 in the other). Each trial reported one death in the metformin plus premixed insulin arms (1/108 (1%) in one study208 and 1/296 (0.3%) in the second study207 ) and no deaths in the metformin plus sulfonylurea arms at 16 weeks.207, 208 (SOE: Low; Combination of metformin plus a sulfonylurea favored for short-term mortality) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Two RCTs compared metformin plus sitagliptin with metformin plus a SGLT-2 inhibitor and reported on mortality.153, 158 A small (N=212) 12-week trial reported no deaths in any arm (metformin plus sitagliptin, metformin plus empagliflozin 10 mg, and metformin plus empagliflozin 25 mg).153 A second trial with 52 weeks of followup reported one death in the metformin plus sitagliptin arm (1/366, 0.3%), one death in the metformin plus canagliflozin 300 mg arm (1/367, 0.3%), and no deaths in the metformin plus canagliflozin 100 mg arm (0/368, 0%).158 (SOE: Low; Neither favored for short-term mortality) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three RCTs compared the combination of metformin plus sitagliptin with metformin plus a GLP-1 receptor agonist (Table 25). In the study with the longest followup (104 weeks), mortality was higher in the metformin plus GLP-1 receptor agonist arm;141 however, the shorter study (52 weeks) reporting on mortality in both arms reported a higher death rate in the metformin plus DPP-4 inhibitor arm.159 (SOE: Insufficient)
- 173. 116 Table 25. Randomizedcontrolled trials comparing a combination of metformin plus sitagliptin with a combination of metformin plus a GLP-1 receptor agonist on all-cause mortality Author, Year Followup Number of Deaths/N (%) in the Metformin + Sitagliptin Arm Number of Deaths/N (%) in the Metformin + GLP-1 Receptor Agonist Arm Ahren, 2014 141 104 weeks 1/302 (0.3) 3/302 (1.0) Nauck, 2014 159 26 weeks 0/315 (0) Dulaglutide 0.75 mg: 0/302 (0) Dulaglutide 1.5 mg: 0/304 (0) 52 weeks 2/315 (0.6) Dulaglutide 0.75 mg: 0/302 (0) Dulaglutide 1.5 mg: 1/304 (0.3) Bergenstal, 2010 188 26 weeks 1/172 (0.6) Exenatide 2 mg weekly: NR/170 GLP-1 = glucagon-like peptide-1; mg = milligrams; NR = not reported Combination of Metformin Plus a GLP-1 Receptor Agonist Versus a Combination of Metformin Plus a Basal Insulin A single RCT (N=321) compared metformin plus exenatide with metformin plus insulin glargine and reported no deaths in either arm at 26 weeks.212 (SOE: Insufficient) Strength of Evidence for All-Cause Mortality We found low or insufficient strength of evidence for all comparisons evaluating all-cause mortality (see Key Points, Table 26, Table 27, and Table 28). Most evidence on this outcome came from RCTs lasting less than 2 years that we found to be at low or medium risk of bias. None of the RCTs were designed to evaluate all-cause mortality. Observational studies had medium risk of bias and tended to support RCT findings. Evidence was more consistent across monotherapy comparisons, with less consistency for combination therapy comparisons, in part because of the smaller number of studies for these comparisons. The RCT evidence on mortality was substantially underpowered and imprecise because of few studies and small sample sizes with few events. As a result, we could not exclude short-term harm for any comparison with moderate strength of evidence. Our evaluation of publication bias was generally limited by the small number of studies. We found unpublished studies that may have affected our grading of the evidence. Published studies suggested a decrease in long-term mortality for metformin plus a DPP-4 inhibitor versus metformin plus a sulfonylurea; one unpublished study was consistent with these conclusions. The single published RCT suggested increased short-term mortality for sulfonylureas versus DPP-4 inhibitors; two unpublished RCTs confirmed those findings. For the comparison of metformin versus metformin plus a sulfonylurea, we only identified one poor-quality, long-term study which showed similar mortality at 104 weeks across arms; however, an unpublished study suggested an increased risk of long-term all-cause mortality for metformin plus a sulfonylurea. For the comparison of metformin versus metformin plus a GLP-1 receptor agonist, we also only identified a single long-term study which suggested similar mortality rates at 104 weeks; however, an unpublished trial found more deaths in the metformin plus GLP-1 receptor agonist arm compared with the metformin monotherapy arm.
- 174. 117 Table 26. Strengthof evidence domains for monotherapy comparisons in terms of all-cause mortality among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. pioglitazone RCTs: 4 (1755) Obs: 1 (NR) Low Medium Consistent Unknown Direct Direct Imprecise N/A Undetected N/A Low Neither treatment favored for short-term mortality Metformin vs. rosiglitazone RCTs: 2 (3224) Obs: 2 (193,172) High Medium Inconsistent Consistent Direct Direct Imprecise Precise Undetected N/A Low Metformin favored Metformin vs. SU (shorter duration studies) RCTs: 4 (928) Low Inconsistent Direct Direct Imprecise Imprecise Undetected N/A Low Neither favored for short-term mortality Metformin vs. SU (longer duration studies) RCTs: 2 (3199) Obs: 7 (398,227) Medium Medium Consistent Consistent Direct Direct Imprecise Precise Undetected N/A Low Metformin favored for long-term mortality Metformin vs. DPP-4 inhibitors RCTs: 5 (4,792) Obs: 1 (84,756) Low Medium Consistent Unknown Direct Direct Imprecise Imprecise Undetected N/A Low Neither treatment favored for short-term mortality Metformin vs. SGLT-2 inhibitors RCTs: 4 (2,041) Medium Consistent Direct Imprecise Undetected Low Neither treatment favored Metformin vs. GLP-1 receptor agonists RCTs: 2 (820) Low Consistent Direct Imprecise Undetected Low Neither treatment favored Incomplete reporting on death Rosiglitazone vs. SU RCTs: 2 (3,484) Obs: 2 (79,681) Medium Medium Inconsistent Consistent Direct Direct Imprecise Precise Undetected N/A Insufficient Unable to determine Pioglitazone vs. SU RCT: 1 (502) Obs: 1 (20,450) Low Medium Unknown Unknown Direct Direct Imprecise Precise Undetected N/A Insufficient Unable to determine
- 175. 118 Table 26. Strengthof evidence domains for monotherapy comparisons in terms of all-cause mortality among adults with type 2 diabetes (continued) Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Pioglitazone vs. DPP-4 inhibitors RCTs: 2 (1031) Medium Consistent Direct Imprecise Undetected Low Neither treatment favored SU vs. DPP-4 inhibitors RCT: 1 (426) Medium Unknown Direct Imprecise Undetected Low DPP-4 inhibitors favored for short-term mortality SU vs. GLP-1 receptor agonists RCTs: 2 (1157) High Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Insufficient reporting of events DPP-4 inhibitors vs. SGLT-2 inhibitors RCTs: 2 (1486) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine DPP-4 inhibitors vs. GLP-1 receptor agonists RCT: 1 (820) Medium Unknown Direct Imprecise Undetected Insufficient Unable to determine DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; Met = metformin; Obs = observational; RCT = randomized controlled trial; RD = risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 176. 119 Table 27. Strengthof evidence domains for metformin versus metformin-based combination comparisons in terms of all-cause mortality among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + pio RCT: 1 (213) Low Unknown Direct Imprecise Undetected Insufficient Unable to determine Metformin vs. metformin + rosiglitazone RCTs: 7 (3242) Medium Consistent Direct Imprecise Undetected Low Metformin monotherapy favored Metformin vs. metformin + SU RCTs: 5 (1983) Medium Consistent Direct Imprecise Suspected ‡ Low Neither treatment favored for short- term mortality Metformin vs. metformin + DPP-4 inhibitors (<2 years) RCTs: 18 (12,446) Low Consistent Direct Imprecise Undetected Low Neither treatment favored for short- term mortality Metformin vs. metformin + DPP-4 inhibitors (long duration studies) RCTs: 2 (2140) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Metformin vs. metformin + SGLT-2 inhibitors (short duration studies) RCTs: 7 (4340) Low Consistent Direct Imprecise Undetected Low Neither treatment favored for short- term mortality Metformin vs. metformin + SGLT-2 inhibitors (long duration studies) RCTs: 2 (728) High Consistent Direct Imprecise Undetected Low Neither treatment favored Metformin vs. metformin + GLP-1 receptor agonists RCT: 2 (2110) Medium Consistent Direct Imprecise Suspected Low Neither treatment favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; pio = pioglitazone; OR = odds ratio; RCT = randomized controlled trial; RD = risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence. ‡ We identified one long-term study showing five deaths in the sulfonylurea arm and one death in the metformin arm at 156 weeks.
- 177. 120 Table 28. Strengthof evidence domains for metformin-based combination comparisons in terms of all-cause mortality among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + rosiglitazone vs. metformin + SU RCTs: 2 (1110) Obs: 1 (80,936) Low Medium Consistent Unknown Direct Direct Imprecise Precise Undetected N/A Low Neither treatment favored for short- term mortality Metformin + pioglitazone vs. metformin + DPP-4 inhibitors RCT: 1 (514) Low Unknown Direct Imprecise Undetected Insufficient Unable to determine Metformin + pioglitazone vs. metformin + GLP-1 receptor agonists RCT: 1 (514) Low Unknown Direct Imprecise Undetected Insufficient Unable to determine Metformin + SU vs. metformin + DPP-4 inhibitors (longer duration studies) RCTs: 5 (6693) Obs: 1 (47,433) Medium Medium Consistent Unknown Direct Direct Imprecise Precise Undetected N/A Low Metformin + DPP-4 inhibitors favored for long-term mortality Metformin + SU vs. metformin + DPP-4 inhibitors (shorter duration studies) RCTs: 2 (1755) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Metformin + SU vs. metformin + SGLT-2 inhibitors (longer duration studies) RCTs: 3 (3815) High Inconsistent Direct Imprecise Undetected Low Neither treatment favored for long- term mortality Metformin + SU vs. metformin + GLP-1 receptor agonists RCT: 2 (1678) Obs: 1 (29,437) High Medium Consistent Unknown Direct Direct Imprecise Precise Undetected N/A Low Neither treatment favored for long- term mortality Metformin + SU vs. metformin + premixed insulin RCTs: 2 (938) High Consistent Direct Imprecise Undetected Low Metformin + SU favored for short- term mortality
- 178. 121 Table 28. Strengthof evidence domains for metformin-based combination comparisons in terms of all-cause mortality among adults with type 2 diabetes (continued) Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors RCTs: 2 (1779) Low Consistent Direct Imprecise Undetected Low Neither favored for short-term mortality Metformin + DPP-4 inhibitors vs. metformin + GLP-1 receptor agonists RCTs: 3 (2216) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Insufficient reporting of events in all arms Metformin + GLP-1 receptor agonists vs. metformin + basal insulin RCT: 1 (321) High Unknown Direct Imprecise Undetected Insufficient Unable to determine DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; Met = metformin; Obs = observational; pio = pioglitazone; RCT = randomized controlled trial; RD = absolute risk difference; rosi = rosiglitazone; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) due to the few longer duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 179. 122 Evidence for CardiovascularMortality Monotherapy Comparisons Metformin Versus Thiazolidinediones Three RCTs compared metformin with thiazolidinediones and did not find differences in cardiovascular mortality.50, 63, 70 Two of the RCTs were small, lasted less than one year, and did not report any cardiovascular deaths (Table 29).63, 70 Studies were not combined because of different lengths of followup and different thiazolidinediones under study. ADOPT was the single long-term RCT (median followup of 4.0 years): the actual number of participants for which ADOPT ascertained CVD mortality is unclear and withdrawals were high across the arms [37% (rosiglitazone) and 38% (metformin)].50 (SOE: Low; Neither favored) Table 29. Randomized controlled trials comparing metformin with thiazolidinediones on cardiovascular mortality Author, Year Followup Number of Events/N (%) in the Metformin Arm Number of Events/N (%) in the Thiazolidinedione Arm Lawrence, 2004 63 24 weeks 0/20 (0%) 0/20 (pioglitazone) (0%) Erem, 2014 70 48 weeks 0/19 (0%) 0/19 (pioglitazone) (0%) Kahn, 2006 50 4 years (median) 2/1454 (0.1%) 1/1456 (0.1%) (rosiglitazone) Metformin Versus Sulfonylureas Two high-quality RCTs compared metformin with sulfonylureas and reported on cardiovascular mortality.50, 231 ADOPT, conducted among patients with newly diagnosed and untreated diabetes (N=2895), reported a slightly higher incidence of fatal MI in the glyburide (3/1441, 0.2%) versus the metformin (2/1454, 0.1%) arm (glyburide vs. metformin: calculated RR, 1.5 (95% CI, 0.3 to 9.0); calculated between-group difference, 0.1%). Median followup was 4.0 years for the metformin (maximum dose 2,000 mg; mean dose not reported) arm and 3.3 years for the glyburide (maximum dose 15 mg; mean dose not reported) arm, and losses to followup were also differential for the metformin (38%) arm vs. the glyburide (44%) arm.50 The smaller RCT was conducted in China among patients with known coronary heart disease (clinical evidence of acute MI or coronary stenosis >50% on angiogram) and also reported a higher risk of cardiovascular mortality in the sulfonylurea (glipizide, mean dose 28.3 mg) arm (11/148, 7.4%) compared with the metformin (mean dose 1,400 mg) arm (7/156, 4.5%) over 2.8 years.231 We calculated the RR of cardiovascular mortality comparing sulfonylurea with metformin to be 1.66 (95% CI, 0.66 to 4.16) and the between-group difference to be 2.9 percent. Losses to followup were 21 percent for each arm of this trial.231 Losses to follow up were the same (20%) across arms in Hong et al.,231 decreasing the risk of non-conservative bias due to losses of follow up across arms. In ADOPT, given differential losses to followup and followup duration across arms, study results were likely biased to the null, lending further support to the inference that metformin was favored over sulfonylurea monotherapy. Three retrospective cohort studies, analyzing two cohorts compared metformin with a sulfonylurea, and all found a higher risk of cardiovascular mortality for sulfonylurea users versus metformin users (Table 30).225, 229, 230 To account for confounding, two of these studies used propensity score matching,225, 229 and one used multivariate regression.230 (SOE: Moderate; Metformin favored for long-term CVD mortality)
- 180. 123 Table 30. Observationalstudies comparing metformin with sulfonylureas on cardiovascular mortality Author, Year Cohort Metformin, N Sulfonylurea, N Median Followu p Adjusted HR (95% CI) for Cardiovascular Mortality Johnson, 2005 230 Saskatchewan Health database 923 2138 4.6 to 5.6 years 0.76 (0.58 to 1.00) Reference: sulfonylurea Schramm, 2011 229 Danish National Patient Register Prior MI 2,906 No prior MI 43,340 Prior MI Glibenclamide: 1,168 Glipizide: 660 Glimepiride: 3,894 No prior MI Glibenclamide: 12,495 Glipizide: 6,965 Glimepiride: 36,313 3.3 years Prior MI Glibenclamide: 1.5 (1.22 to 1.84) Glipizide: 1.63 (1.28 to 2.07) Glimepiride: 1.32 (1.11 to 1.57) No prior MI Glibenclamide: 1.14 (1.03 to 1.25) Glipizide: 1.25 (1.12 to 1.4) Glimepiride: 1.28 (1.18 to 1.38) Reference: metformin Andersson, 2010 225 Danish National Patient Register – Incident admission for heart failure* 688 3615 844 days 0.79 (0.65 to 0.96) Reference: sulfonylurea CI = confidence interval; HR = hazard ratio; MI = myocardial infarction * Unclear if this population was included in Schramm, 2011229 Metformin Versus DPP-4 Inhibitors Three RCTs compared metformin with DPP-4 inhibitors and reported on cardiovascular mortality (Table 31). These studies varied in duration and did not use similar definitions for cardiovascular events. Therefore, we did not combine them in a meta-analysis. Cardiovascular mortality was rare and appeared to be more frequent in the metformin than DPP-4 inhibitor arms, when reported.85-87 The longest study reported no cardiovascular mortality in either arm.85 (SOE: Low; DPP-4 inhibitors favored) Table 31. Randomized controlled trials comparing metformin with DPP-4 inhibitors on cardiovascular mortality Author, Year Outcome Followup (Weeks) Metformin Dose: Number of Events/N (%) DPP-4 Inhibitor Dose: Number of Events/N (%) Haak, 2012 86 Fatal MI 24 1000 mg: 0/142 (0.0%) 2000 mg: 1/147 (0.7%) Linagliptin 5 mg: 0/142 (0.0%) Pfutzner, 2011 87 Sudden death, cardiac arrest, coronary arteriosclerosis, cardiac failure, acute MI 76 2000 mg: 3/328 (0.9%) Saxagliptin 10 mg: 1/335 (0.3%) Williams-Herman, 2010 85 Sudden cardiac death or worsening CHD 104 1000 mg: 0/182 2000 mg: 0/182 Sitagliptin 100 mg: 0/179 CHD = coronary heart disease; DPP-4 = dipeptidyl peptidase-4; mg = milligrams; MI = myocardial infarction
- 181. 124 Thiazolidinediones Versus Sulfonylureas TheADOPT trial compared rosiglitazone with glyburide and reported two fatal myocardial infarctions in the rosiglitazone arm (2/1446, 0.1%) and three fatal myocardial infarctions in the glyburide arm (3/1441, 0.2%) resulting in a calculated risk ratio of 0.66 (95% CI, 0.11 to 3.97) and between-group difference of -0.1% for rosiglitazone versus glyburide. Notably, losses to followup and followup duration were differential across the arms; losses to followup were higher (44% vs. 37%) for the sulfonylurea versus rosiglitazone arm and median followup was shorter (3.3 years vs. 4.0 years) for the sulfonylurea versus rosiglitazone arm.50 (SOE: Low; Rosiglitazone favored for long-term CVD mortality) Sulfonylureas Versus DPP-4 Inhibitors Two small RCTs compared a sulfonylurea with a DPP-4 inhibitor (duration 52 to 58 weeks) and reported mixed results on fatal myocardial infarction.106, 107 The high-quality study reported one fatal myocardial infarction in the linagliptin arm (1/151, 0.7%) and none in the sulfonylurea arm (0/151, 0%),106 and the other study reported two events in the sulfonylurea arm (2/212, 0.9%) and did not report on fatal myocardial infarctions in the sitagliptin arm (n=211).107 The lower-quality study did not use an intention-to-treat analysis for fatal myocardial infarction and had large losses to followup (19.8% and 22.3% in the sulfonylurea and DPP-4 inhibitor arms, respectively).107 (SOE: Insufficient) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a Thiazolidinedione Five RCTs reported in four articles120, 123, 127, 241 compared metformin with metformin plus rosiglitazone and found non-significantly increased odds of short-term cardiovascular mortality for the combination of metformin plus rosiglitazone versus metformin monotherapy (pooled OR, 2.68; 95% CI, 0.42 to 17.08) (Figure 48). Three of the studies reported a single cardiovascular death in the metformin plus rosiglitazone arm, and all studies reported no cardiovascular deaths in the metformin monotherapy arms. The results of the 80-week study127 did not differ from those of the shorter studies.120, 123, 241 We did not find statistical heterogeneity (I2 = 0.0%), and removal of any one study from the meta-analysis did not change the inference. The pooled between-group difference for short-term cardiovascular mortality for metformin plus rosiglitazone versus metformin monotherapy was 0.3% (95% CI, -0.2 to 0.9%).
- 182. 125 Figure 48. Pooledodds ratio for short-term cardiovascular mortality comparing metformin with a combination of metformin plus rosiglitazone CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus rosiglitazone; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. A single 26-week RCT compared metformin monotherapy with metformin plus pioglitazone (15, 30, and 45 mg arms) and reported one sudden cardiac death in the metformin plus pioglitazone 45 mg arm (1/130, 1%). The study did not report on sudden cardiac death for the other arms.126 Of note, two of the RCTs had substantial losses to followup (38% to 45% in one study127 and 12% to 18% in the other126 ). This, along with a lack of reporting on the intention-to-treat population, limits our conclusions. (SOE: Low; Metformin monotherapy favored over combination of metformin plus rosiglitazone for short-term CVD mortality) (SOE: Insufficient for combination of metformin plus pioglitazone) Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Six RCTs comparing metformin with the combination of metformin plus a DPP-4 inhibitor showed a non-significant decreased risk of short-term cardiovascular mortality for the metformin plus DPP-4 inhibitor arms versus metformin, based on 10 deaths across the trials (pooled OR, 0.51; 95% CI, 0.15 to 1.73) (Figure 49). We did not find statistical heterogeneity, and removal of any one study did not change the inference from the meta-analysis.86, 87, 142, 145, 152, 159 Cardiovascular deaths were not described in the metformin plus saxagliptin arm in Pfutzner 2011, et al, and we assumed that no events occurred in that arm for the meta-analysis.87 Three of the RCTs included in the meta-analysis also had additional arms with lower dosages and did not report on events in those arms.86, 87, 126 The pooled between-group difference in short-term cardiovascular mortality was -0.1% (95% CI, -0.4 to 0.3%) for the combination of metformin plus a DPP-4 inhibitor compared with metformin.
- 183. 126 Figure 49. Pooledodds ratio for short-term cardiovascular mortality comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a DPP-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. An additional longer RCT (104 weeks) reported one cardiovascular death in the metformin plus sitagliptin arm and did not report on cardiovascular deaths in the metformin monotherapy arms.85 (SOE: Low; Combination of metformin plus DPP-4 inhibitors favored for short-term CVD mortality) Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor One RCT (N=546) compared metformin with the combination of metformin plus dapagliflozin at different doses (2.5, 5.0, and 10.0 mg) and reported two cardiovascular deaths in the metformin plus 2.5-mg dapagliflozin arm at 102 weeks and did not report on deaths in the other arms.170 (SOE: Insufficient) Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist A single 26-week RCT compared metformin with metformin plus dulaglutide at two different doses (0.5 mg and 1.5 mg per week) and reported one fatal stroke in the metformin plus dulaglutide 1.5 mg/week (1/304, 0.3%) arm and no events in the metformin (0/177, 0%) or metformin plus dulaglutide 0.75 mg/week (0/302, 0%) arm.159 (SOE: Low; Metformin favored for short-term fatal stroke)
- 184. 127 Metformin-Based Combination Comparisons Combinationof Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor A single, five-arm, 26-week RCT (N=1,554) compared metformin plus pioglitazone (arms with doses of 15 mg, 30 mg, and 45 mg) with metformin plus alogliptin (12.5-mg and 25-mg arms) and reported on sudden cardiac death.126 The investigators reported one sudden cardiac death in the metformin plus pioglitazone 45 mg arm (1/129, 0.8%) and did not report on this outcome in the other arms. (SOE: Insufficient) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor Six RCTs addressed cardiovascular/cerebrovascular mortality for this comparison. Results from four RCTs, each with 104 weeks of followup and low event rates (<1%), suggested lower rates of fatal cardiovascular events for metformin plus a DPP-4 inhibitor versus metformin plus a sulfonylurea (pooled OR, 0.57; 95% CI, 0.19 to 1.69) (Figure 50). We did not find statistical heterogeneity (I2 = 0.0%).194-197 Removal of any single study did not change the inference of the meta-analysis. Of note, definitions of cardiovascular mortality varied slightly across the studies included in this meta-analysis (Table 32). Losses to followup were high across these studies. The pooled between-group difference for long-term cardiovascular mortality for metformin plus a DPP-4 inhibitor compared with metformin plus a sulfonylurea was -0.2 (95% CI, -0.5 to 0.1%). One RCT with 52 weeks of followup conducted among persons (predominantly men) older than 65 years of age reported one fatal myocardial infarction in the metformin plus saxagliptin arm and did not report on this outcome explicitly for the metformin plus sulfonylurea arm.193 Two RCTs (N=1,893; durations, 30 weeks and 104 weeks) addressed fatal stroke specifically for this comparison; these reported one event in the metformin plus sulfonylurea arms and did not report on this outcome for the metformin plus DPP-4 inhibitor arms.190, 195 A single retrospective cohort study (N=36,230) from the Danish Patient Register reported a significantly lower risk of cardiovascular mortality, with a median of 2.1 years of followup, for metformin plus DPP-4 inhibitor users versus metformin plus sulfonylurea users (adjusted rate ratio, 0.57; 95% CI, 0.4 to 0.8).227 (SOE: Low; Combination of metformin plus a DPP-4 inhibitor favored for long-term (2-5 years) CVD mortality; SOE: Insufficient for short-term CVD mortality)
- 185. 128 Figure 50. Pooledodds ratio for long-term cardiovascular mortality comparing combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = combination of metformin plus a sulfonylurea; Group 2 = metformin plus a dipeptidyl peptidase-4 inhibitor; Met = metformin; OR = odds ratio; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 186. 129 Table 32. Randomizedcontrolled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor on cardiovascular mortality Author, Year Followup (Weeks) Definition of Fatal Cardiovascular Event Number of Events / N (%) in the Metformin Plus Sulfonylurea Arm Number of Events / N (%) in the Metformin Plus DPP-4 Inhibitor Arm Included in Meta-Analysis Seck, 2010 196 * 104 Sudden cardiac death, fatal MI 3/548 (0.5%) 0/588 (0) Yes Goke, 2010 195 104 Composite CVD mortality outcome (cardiac failure, MI) 1/430 (0.2%) 1/428 (0.2%) Yes Gallwitz, 2012 194 104 Composite CVD mortality outcome (sudden cardiac death, fatal MI, and fatal stroke) 2/775 (0.3%) 2/776 (0.3%) Yes Del Prato, 2014 197 104 Not specified 4/869 (0.5%) 2/873 (0.2%) (alogliptin 12.5 mg) 2/878 (0.2%) (alogliptin 25 mg) Yes Schernthaner, 2015 193 52 Fatal MI NR/360 1/360 (0.3%) No; short-term duration and did not report on events in both arms Goke, 2010 195 104 Fatal stroke 1/430 (0.2%) NR/428 No; did not report on events in both arms Arechavaleta, 2011 190 30 Fatal stroke 1/519 (0.2%) NR/516 No; short-term duration and did not report on events in both arms CVD = cardiovascular disease; DPP-4 = dipeptidyl peptidase-4; MI = myocardial infarction * 104-week followup of Nauck 2007, et al192 Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Two long-term RCTs (104 weeks of followup) each reported one cardiovascular death in the metformin plus sulfonylurea arm (1/408, 0.2%199 and 1/482, 0.2%201 ) and no cardiovascular deaths in the metformin plus SGLT-2 inhibitor arms.199, 201 Therefore, the between-group difference in long-term cardiovascular mortality across trials was 0.2% for the combination of metformin plus a sulfonylurea compared with the combination of metformin plus an SGLT-2 inhibitor. (SOE: Low; Combination of metformin plus a SGLT-2 inhibitor favored for long-term CVD mortality) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist A single retrospective cohort study (N=29,437) from the Danish Patient Register reported a non-significantly lower risk of cardiovascular mortality, with a median of 2.1 years of followup,
- 187. 130 for metformin plusGLP-1 receptor agonist users versus metformin plus sulfonylurea users (adjusted rate ratio, 0.89; 95% CI, 0.47 to 1.68).227 (Not graded) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Basal Insulin A single retrospective cohort study (N=29,437) from the Danish Patient Register reported a significantly increased risk of cardiovascular mortality, with a median of 2.1 years of followup, for metformin plus basal insulin users versus metformin plus sulfonylurea users (adjusted rate ratio, 1.57; 95% CI, 1.23 to 2.01).227 (Not graded) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Premixed Insulin A single 16-week, open-label RCT (N=341) randomized participants with poorly controlled diabetes on metformin alone to the addition of glibenclamide or twice daily insulin aspart 70/30 and reported no deaths in the metformin plus glibenclamide arm and one fatal myocardial infarction in the metformin plus premixed insulin arm.208 (SOE: Low; Combination of metformin plus sulfonylurea favored for short-term CVD mortality) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist A single 26-week, open-label RCT (N=665) randomized participants with poorly controlled diabetes on metformin alone to the addition of oral sitagliptin (100 mg) or one of two doses of daily subcutaneous injections of liraglutide (1.2 mg or 1.8 mg) and reported one fatal cardiac arrest in the metformin plus sitagliptin arm (1/219, 0.5%) and none in the metformin plus liraglutide arms (liraglutide 1.2 mg: 0/221, 0% and liraglutide 1.8 mg: 0/218, 0%).210 A single 26-week RCT (N=921) reported on fatal stroke for metformin plus sitagliptin versus metformin plus dulaglutide at two doses (0.75 mg/week and 1.5 mg/week). The investigators reported one fatal stroke in the metformin plus dulaglutide 1.5-mg/week arm (1/304, 0.3%) and no events in the other arms (dulaglutide 0.75 mg: 0/302, 0% and sitagliptin: 0/315, 0%).159 (SOE: Insufficient) Combination of Metformin Plus a Basal Insulin Versus a Combination of Metformin Plus a Premixed Insulin A single 32-week, open-label, cross-over study (N=597) randomized participants to metformin plus insulin glargine or metformin plus insulin lispro 75/25 twice daily and reported one fatal myocardial infarction in the metformin plus insulin lispro 75/25 arm and no events in the metformin plus glargine arm.224 (SOE: Insufficient) Strength of Evidence for Cardiovascular Mortality Although we identified one comparison for which there was moderate strength of evidence on long-term cardiovascular mortality, evidence was generally of low strength or insufficient for cardiovascular mortality (see Key Points, Table 33, Table 34, and Table 35). Most of the evidence was on short-term cardiovascular mortality, and none of the RCTs were designed to evaluate cardiovascular mortality. We identified observational studies which strengthened the evidence for a few comparisons (metformin versus sulfonylurea and metformin plus a sulfonylurea versus metformin plus DPP-4 inhibitors). Almost all of the evidence on
- 188. 131 cardiovascular mortality wasof medium or high risk of bias. When data were available from more than one study for a given comparison, the evidence tended to be consistent. However, we only identified a single study for many comparisons making consistency indeterminate. Sample size and low event rates in the RCTs limited the power and precision of the evidence on cardiovascular mortality, and the small number of studies limited our ability to assess publication bias. We identified one unpublished study (an extension of an included study with 156 weeks of followup) which addressed several comparisons of interest. This study had few fatal cardiovascular or cerebrovascular events (none in the metformin arm or metformin plus GLP-1 receptor agonist arm; one fatal MI in the metformin plus sitagliptin arm; and one fatal cerebrovascular accident in the metformin plus sulfonylurea arm). This study’s results were slightly contrary to our findings on metformin plus a sulfonylurea versus metformin plus a DPP- 4 inhibitor. While we did not identify any published RCTs comparing cardiovascular mortality for metformin versus metformin plus a sulfonylurea, the unpublished RCT suggested similar long-term fatal cardiovascular mortality for metformin and metformin plus sulfonylurea, but possibly an increased risk of fatal cerebrovascular accident in the metformin plus sulfonylurea arm versus metformin. We also did not identify long-term RCTs of metformin plus a DPP-4 inhibitor versus metformin plus a GLP-1 receptor agonist; in this long-term unpublished study, metformin plus sitagliptin was favored for this comparison.
- 189. 132 Table 33. Strengthof evidence domains for monotherapy comparisons in terms of cardiovascular mortality among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. pioglitazone RCTs: 2 (120) Medium Consistent Direct Imprecise Undetected Low Neither treatment favored Metformin vs. rosiglitazone RCT: 1 (2940) High Unknown Direct Imprecise Undetected Low Neither treatment favored Metformin vs. SU (longer duration studies) RCTs: 2 (4664) Observational: 3 (115,105) Medium Medium Consistent Consistent Direct Direct Imprecise Precise Undetected Undetected Moderate Metformin favored; RR 1.6 to 2.0 and between group differences 0.1% to 2.9% from RCTs for SU vs. metformin Metformin vs. DPP-4 inhibitors RCTs: 3 (3,188) Medium Inconsistent Direct Imprecise Undetected Low DPP-4 inhibitors favored for short-term mortality Rosiglitazone vs. SU (longer- duration studies) RCT: 1 (2,987) High Unknown Direct Imprecise Undetected Low Rosiglitazone favored SU vs. DPP-4 inhibitors (shorter duration studies) RCT: 2 (653) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Events not reported for all arms DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; RCT = randomized controlled trial; SGLT-2 inhibitors = sodium- glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 190. 133 Table 34. Strengthof evidence domains for metformin versus metformin-based combination comparisons in terms of cardiovascular mortality among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + rosiglitazone RCTs: 5 (2,167) High Consistent Direct Imprecise Undetected Low Metformin favored for short- term mortality Metformin vs. metformin + pioglitazone RCT: 1 (1,554) High Unknown Direct Imprecise Undetected Insufficient Unable to determine Metformin vs. metformin+DPP-4 inhibitor RCTs: 7 (6,673) Medium Consistent Direct Imprecise Undetected Low Metformin + DPP-4 inhibitors favored for short-term mortality Metformin vs. metformin + SGLT-2 inhibitor RCT: 1 (546) Low Unknown Direct Imprecise Undetected Insufficient Unable to determine Events not reported on in three arms Metformin vs. metformin + GLP-1 receptor agonist RCT: 1 (1098) Medium Unknown Direct Imprecise Undetected Low Metformin favored for short- term fatal stroke DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; RCT = randomized controlled trial; RD = risk difference; SGLT- 2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) and pooled risk differences (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 191. 134 Table 35. Strengthof evidence domains for metformin-based combination comparisons in terms of cardiovascular mortality among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + pioglitazone vs. metformin + DPP-4 inhibitor RCT: 1 (1554) High Unknown Direct Imprecise Undetected Insufficient Unable to determine Events not reported on in four arms Metformin + SU vs. metformin + DPP-4 inhibitors (104 weeks followup) RCTs: 4 (6184) Observational: 1 (36,620) Medium Medium Inconsistent Unknown Direct Direct Imprecise Precise Undetected N/A Low Metformin + DPP-4 inhibitors favored for long- term CVD mortality Metformin + SU vs. metformin + DPP-4 inhibitors (shorter duration studies) RCTs: 2 (1755) Medium Unknown Direct Imprecise Undetected Insufficient Unable to determine Events not reported on in all arms Metformin + SU vs. metformin + SGLT- 2 inhibitor (longer duration studies) RCT: 2 (2266) Medium Consistent Direct Imprecise Undetected Low Metformin + SGLT-2 inhibitors favored Metformin + SU vs. metformin + premixed insulin RCT: 1 (341) Medium Unknown Direct Imprecise Undetected Low Metformin + SU favored for short-term CVD mortality Metformin + DPP- inhibitor vs. metformin + GLP-1 receptor agonist RCTs: 2 (1,763) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Metformin + basal insulin vs. metformin + premixed insulin RCT: 1 (597) Medium Unknown Direct Imprecise Undetected Insufficient Unable to determine CVD = cardiovascular; DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; RCT = randomized controlled trial; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence due to a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 192. 135 Evidence for Cardiovascularand Cerebrovascular Morbidity Monotherapy Comparisons Metformin Versus Thiazolidinediones Randomized Controlled Trials Two RCTs compared metformin with rosiglitazone and reported on different cardiovascular morbidity outcomes.50, 59 In ADOPT, at 4 years, there was a small increased risk of non-fatal MI for the rosiglitazone vs. metformin arm (between-group difference of 0.3 absolute percentage points) and an increased risk of peripheral vascular disease (between-group difference of 0.6 absolute percentage points). There was also a small increased risk of stroke in the metformin vs. rosiglitazone arm (between-group difference of 0.2 absolute percentage points).50 Total CVD event rates were higher (by 0.7%) in the rosiglitazone (77/1456, 5.3%) vs. metformin (67/1454, 4.6%) arm. The completeness of ascertainment of CVD morbidity was unclear, and losses to followup were 37% and 38% for the rosiglitazone and metformin arms, respectively.50 Event rates in the small 32-week study were the same in both arms (Table 36).59 Table 36. Randomized controlled trials comparing metformin with rosiglitazone on cardiovascular morbidity Author, Year Enrolled N Followup Outcome Number of Events / N (%) for Metformin Vs. Rosiglitazone Kahn*, 2006 50 4360 4 years (median) Non-fatal myocardial infarction Stroke Peripheral vascular disease 21/1454 (1.4) vs. 25/1456 (1.7) 19/1454 (1.3) vs. 16/1456 (1.1) 27/1454 (1.9) vs. 36/1456 (2.5) Rosenstock, 2006 59 468 32 weeks Not defined ischemic heart disease 2/154 (1) vs. 1/159 (1) * ADOPT Study Three small RCTs, each shorter than a year, compared metformin with pioglitazone and reported very few events.63, 70, 71 We did not perform a meta-analysis given the absence of events in two of the three studies and lack of reporting in the third (Table 37). Table 37. Randomized controlled trials comparing metformin with pioglitazone on cardiovascular morbidity Author, Year Enrolled N Followup Outcome Number of Events / N (%) for Metformin Vs. Pioglitazone Erem, 2014 70 57 48 weeks Nonfatal MI 0/19 (0) vs. 0/9 (0) Lawrence, 2004 63 60 24 weeks Nonfatal CVD morbidity/MI 0/20 (0) vs. 0/20 (0) Genovese, 2013 71 58 16 weeks Discontinuation due to myocardial ischemia 1/29 (3.4) vs. NR/29 CVD = cardiovascular disease; MI = myocardial infarction; NR = not reported Observational Studies Three retrospective cohort studies233, 243, 244 compared metformin with rosiglitazone and reported mixed results (Table 38). One study reported no increased risk of ischemic heart disease for rosiglitazone versus metformin,233 and the other two studies suggested an increased risk of cardiovascular morbidity for rosiglitazone versus metformin.243, 244
- 193. 136 (SOE: Low; Metforminfavored for long-term (follow up at least 2 years) CVD morbidity) Table 38. Retrospective cohort studies comparing metformin with rosiglitazone on cardiovascular morbidity Author, Year Population (N) Followup Outcome Adjusted HR (95% CI) Pantalone, 2009 233 Cleveland Clinic electronic health record system (11,515) 8 years Ischemic heart disease 0.96 (0.76 to 1.21) Reference = metformin Hsiao, 2009 243 Taiwan National Health Insurance – newly-diagnosed diabetes (48,537) 6 years Myocardial infarction Angina pectoris Transient ischemic attack Stroke 2.09 (1.36 to 3.24) 1.79 (1.39 to 2.30) 2.57 (1.33 to 4.96) 1.61 (0.72 to 3.62) Reference = metformin Brownstein, 2010 244 United States (34,252) 7 years Hospitalization for acute MI 3.0 (2.4 to 3.7) Reference = metformin CI = confidence interval; HR = hazard ratio; MI = myocardial infarction Two of the retrospective cohort studies compared metformin with pioglitazone233, 243 and found no significant difference in cardiovascular disease risk between groups (Table 39). Of note, participants in the Taiwan National Health Insurance database study prescribed pioglitazone were more likely to have a history of cardiovascular disease than those prescribed metformin.243 (SOE: Moderate; Neither metformin nor pioglitazone favored) Table 39. Retrospective cohort studies comparing metformin with pioglitazone on cardiovascular morbidity Author, Year Population (N) Followup Outcome Adjusted HR (95% CI) Pantalone, 2009 233 Cleveland Clinic electronic health record system (11944) 8 years Ischemic heart disease 1.11 (0.91 to 1.34) Reference = metformin Hsiao, 2009 243 Taiwan National Health Insurance – newly-diagnosed diabetes (46,939) 6 years Myocardial infarction Angina pectoris 1.0 (0.26 to 3.89) 1.15 (0.6 to 2.21) Reference = metformin CI = confidence interval; HR = hazard ratio Metformin Versus Sulfonylureas Randomized Controlled Trials Three RCTs50, 134, 231 compared metformin with sulfonylureas and reported on cardiovascular morbidity (Table 40). Two of these RCTs had long-term followup. In ADOPT, the risk of nonfatal myocardial infarction and stroke were higher in the metformin versus sulfonylurea arm (between-group differences of 0.4% and 0.1% for nonfatal myocardial infarction and stroke, respectively). Notably, losses to followup and duration of followup were differential across these arms with higher losses to followup (44% versus 38%) and shorter median followup (3.3 versus 4.0 years) for sulfonylurea versus metformin.50 Cardiovascular event rates were higher for sulfonylurea versus metformin in the other long-term RCT, which was conducted in a predominantly-male, Chinese population with an established diagnosis of coronary heart
- 194. 137 disease.231 Losses to followupwere the same (21%) for both arms of this trial.231 The third RCT was small and short (6 months) and found higher rates of undefined cardiovascular morbidity in the sulfonylurea than metformin arm.134 Table 40. Randomized controlled trials comparing metformin with sulfonylureas on cardiovascular morbidity Author, Year Followup Outcome Number of Events / N (%) for Metformin Vs. Sulfonylurea Kahn, 2006 50 4 years (median) Non-fatal MI 21/1454 (1.4) vs. 15/1441 (1.0) Stroke Peripheral vascular disease 19/1454 (1.3) vs. 17/1441 (1.2) 27/1454 (1.9) vs. 31/1441 (2.2) Hong, 2013 231 2.8 years Non-fatal MI confirmed by medical records 5/156 (3.2) vs. 6/148 (4.1) Non-fatal stroke confirmed by medical records 10/156 (6.4) vs. 15/148 (10) Arterial revascularization by PTCA or by coronary artery bypass graft confirmed by medical records 21/156 (14) vs. 25/148 (17) CVD morbidity composite outcome* 39/156 (25) vs. 52/148 (35) New critical cardiac arrhythmia confirmed by medical record 30/156 (19) vs. 27/148 (18) New or worsening angina confirmed by medical record 77/156 (49) vs. 71/148 (48) New peripheral vascular disease events confirmed by medical record 1/156 (0.6) vs. 6/148 (4.1) Hermann, 1994 134 6 months Unclear – CVD morbidity/CHD 2/25 (5) vs. 3/21 (9) CHD = coronary heart disease; CVD = cardiovascular disease; MI = myocardial infarction; PTCA = percutaneous transluminal coronary angioplasty * Including nonfatal myocardial infarction, nonfatal stroke, or arterial revascularization by PTCA or by coronary artery bypass graft, death from a cardiovascular cause, and death from any cause, obtained and confirmed by medical record. Observational Studies Five retrospective cohort studies229, 233, 238, 245, 246 and one case-control study226 reported on cardiovascular morbidity for metformin and sulfonylurea use (Table 41 and Table 42). All but one study233 reported a significantly increased risk of incident cardiovascular morbidity among sulfonylurea versus metformin users.226, 229, 233, 238, 245, 246 This risk extended to populations without a history of cardiovascular disease at baseline.229, 238 (SOE: Low; Metformin favored for long-term CVD morbidity)
- 195. 138 Table 41. Retrospectivecohort studies comparing metformin with sulfonylureas on cardiovascular morbidity Author, Year Population (N) Followup Outcome Adjusted HR (95% CI) Pantalone, 2009 233 Cleveland Clinic EHR system (17863) 8 years Incident ischemic heart disease by ICD-9 code 0.94 (0.85 to 1.05) Reference = sulfonylurea Hung, 2013 245 Taiwan National Health Insurance Research Database (N=925) Median 3.1 to 3.8 years Composite cardiovascular outcome based on ICD-9 codes 0.31 (0.24 to 0.4) Reference = sulfonylurea Roumie, 2012 246 Veterans Administration database linked to Medicare files (N= 253,690) 0.61 to 0.78 years Hospitalization for acute MI, stroke, or death 1.21 (1.13 to 1.29)* Reference = metformin Acute MI and stroke 1.15 (1.06 to 1.25)* Reference = metformin Schramm, 2011 229 Danish Patient Register (N=107,806) Median 3.3 years Composite of MI, stroke and cardiovascular death based on ICD-10 codes Prior MI 1.29 (1.09 to 1.52) † 1.46 (1.2 to 1.78) ‡ 1.29 (1.12 to 1.49) § No prior MI 1.12 (1.04 to 1.21) † 1.17 (1.07 to 1.28) ‡ 1.21 (1.14 to 1.29) § Reference = metformin Corrao, 2011 238 Health Service Databases Lombardy (N=70,437) Mean 4.8 to 5.1 years Composite of death from any cause or first hospitalization for MI, cerebrovascular disease, or coronary artery bypass graft based on ICD-9 1.15 (1.08 to 1.21) Reference = metformin CI = confidence interval; EHR = electronic health record; HR = hazard ratio; ICD = International Classification of Diseases; MI = myocardial infarction *33% to 39% data missing on hemoglobin A1c (covariate in model) † glibenclamide ‡ glipizide § glimepiride Table 42. Nested case-control study comparing metformin with sulfonylureas on hospitalization for incidence of myocardial infarction Authour, Year Population Followup Cases Controls Adjusted OR (95% CI) Horsdal, 2011 226 Danish National Patient Registry Median 6 months First-time hospitalization for non-fatal MI (N=10,616) Age- and gender- matched patients with diabetes and no history of MI (N=90,697) 0.86 (0.78 to 0.95) Reference = sulfonylurea CI = confidence interval; MI = myocardial infarction; OR = odds ratio Metformin Versus DPP-4 Inhibitors Two RCTs compared metformin with DPP-4 inhibitor monotherapy and reported on cardiovascular morbidity. One small, 26-week RCT of low quality reported one nonfatal myocardial infarction in the alogliptin arm (1/112, 1%) and did not report on events in the metformin arms (n=109 and n=111 for the 1000-mg and 2000-mg arms, respectively). This study noted that it evaluated nonfatal stroke but did not report these outcomes.84 This study did not use an intention-to-treat approach for cardiovascular morbidity and had greater than 17 percent
- 196. 139 losses to followupin both arms.84 A second larger and longer (76 weeks) RCT of higher quality reported that 2.1 percent of participants in the metformin arm (n=328) experienced an acute cardiovascular adverse event (otherwise unspecified) and did not report on this outcome in the saxagliptin arm (n=335).87 A single retrospective cohort study from the Danish National Patient Registry (N=84,756) reported a non-significant increase in cardiovascular risk (composite outcome: all-cause mortality, acute myocardial infarction, and stroke) for sitagliptin versus metformin over mean followup of 0.9 to 1.8 years (adjusted RR, 1.22; 95% CI, 0.92 to 1.61).228 (SOE: Insufficient) Metformin Versus SGLT-2 Inhibitors A single 12-week RCT reported one episode of Prinzmetal angina in the metformin arm (1/80, 1.3%) and did not report on events in the empagliflozin 10 mg (n=81) or empagliflozin 25 mg (n=82) arms.239 This study did not use an intention-to-treat approach and did not report on withdrawals.239 (SOE: Insufficient) Thiazolidinediones Versus Sulfonylureas Two RCTs50, 217 and two retrospective cohort studies243, 244 compared the effects of sulfonylureas and rosiglitazone on cardiovascular morbidity (Table 43). In the long-term RCT, ADOPT, CVD morbidity was higher in the rosiglitazone versus sulfonylurea arm (between- group difference of 0.7% for in nonfatal MI and 0.1% for stroke); losses to followup (44% versus 37%) were higher in the sulfonylurea arm while followup duration was shorter.50 Results of the shorter RCT were consistent with ADOPT.217 CVD morbidity was non-statistically significantly higher for rosiglitazone versus a sulfonylurea in two of three observational studies.243, 244 A single short-term RCT95 reported a 0.2% increase in coronary heart disease morbidity for glyburide versus pioglitazone at 56 weeks (Table 44). Long-term cohort studies compared the effects of sulfonylureas and pioglitazone and reported mixed findings.233, 243 (SOE: Low; Sulfonylureas favored over rosiglitazone) (SOE: Low; Sulfonylureas favored over rosiglitazone for long-term CVD morbidity; Low; Pioglitazone favored for short-term CVD morbidity)
- 197. 140 Table 43. Studiescomparing rosiglitazone with sulfonylureas on cardiovascular morbidity Author, Year Study Design Population (N) Followup Outcome Results Kahn, 2006 50 RCT ADOPT Study (4360) 4.0 years for rosiglitazone 3.3 years for SU (median) Nonfatal MI Stroke Peripheral vascular disease Rosiglitazone: 1.7%; SU: 1.0% Rosiglitazone: 1.3%; SU: 1.2% Rosiglitazone: 2.5%; SU: 2.2% St John Sutton, 2002 217 RCT N=351 52 weeks Cardiac-related adverse events Rosiglitazone: 15.4%; SU: 12.1% Pantalone, 2009 233 Retrospective cohort Cleveland Clinic electronic health record system (8506) 8 years Incident ischemic heart disease by ICD- 9 code Adjusted HR 0.90; 95% CI, 0.71 to 1.14 Reference = sulfonylurea Hsiao, 2009 243 Retrospective cohort Taiwan National Health Insurance – newly-diagnosed diabetes (99744) 6 years MI Stroke Transient ischemic attack Angina pectoris Adjusted HR 1.49; 95% CI, 0.99 to 2.24 Adjusted HR 1.45; 95% CI, 0.69 to 3.05 Adjusted HR 1.90; 95% CI,1.02 to 3.57 Adjusted HR 1.46; 95% CI, 1.15 to 1.85 Reference = sulfonylurea Brownstein, 2010 244 Retrospective cohort Research Patient Data Registry (34,252) 7 years Hospitalization for MI Adjusted RR, 1.3; 95% CI, 1.0 to 1.6 Reference = sulfonylurea ADOPT = A Diabetes Outcome Progression Trial; CI = confidence interval; HR = hazard ratio; ICD = International Classification of Diseases; MI = myocardial infarction; RCT = randomized controlled trial; RR = rate ratio
- 198. 141 Table 44. Studiescomparing pioglitazone with sulfonylureas on cardiovascular morbidity Author, Year Study Design Population (N) Followup Outcome Results Jain, 2006 95 RCT N=502 56 weeks CHD, MI and chest pain Pioglitazone: 1%; glyburide: 3% Pantalone, 2009 233 Retrospective cohort Cleveland Clinic electronic health record system (8935) 8 years Incident ischemic heart disease by ICD-9 code Adjusted HR 1.04; 95% CI, 0.86 to 1.26 Reference = sulfonylurea Hsiao, 2009 243 Retrospective cohort Taiwan National Health Insurance – newly-diagnosed diabetes (98146) 6 years MI Stroke Transient ischemic attack Angina pectoris Adjusted HR 0.72; 95% CI, 0.19 to 2.77 Adjusted HR 0.59; 95% CI, 0.06 to 6.03 Adjusted HR 1.28; 95% CI, 0.34 to 4.86 Adjusted HR 0.91; 95% CI, 0.47 to 1.74 Reference=sulfonylurea CHD = coronary heart disease; CI = confidence interval; HR = hazard ratio; ICD = International Classification of Diseases; MI = myocardial infarction; RCT = randomized controlled trial Sulfonylureas Versus DPP-4 Inhibitors Two RCTs (each 52 to 58 weeks duration) compared the effects of sulfonylureas with DPP-4 inhibitors on short-term cardiovascular morbidity.106, 107 One study reported one nonfatal myocardial infarction in one participant in the sulfonylurea arm (1/76, 1.3%) and no events in the linagliptin arm (0/151, 0.0%); the between-group risk difference was 1.3% for sulfonylurea versus linagliptin.106 The other study reported 11 vascular events in the sulfonylurea arm (11/212, 5.2%) and eight events in the sitagliptin arm (8/210, 3.8%); the between-group risk difference was 1.4% for the sulfonylurea versus sitagliptin arm.107 The dose of glimepiride was low in one study (4 mg).106 The study by Arjona Ferreira et al did not use an intention-to-treat approach and had greater than 20 percent losses to followup across arms.107 (SOE: Low; DPP-4 inhibitors favored for short-term cardiovascular morbidity) Sulfonylureas Versus GLP-1 Receptor Agonists Two RCTs compared the effects of sulfonylurea with liraglutide on cardiovascular morbidity and reported slightly higher rates of cardiovascular events in the sulfonylurea arms compared with the liraglutide arms. In the longer study (104 weeks), 14 of 248 (6%) participants experienced a cardiac disorder in the sulfonylurea arm, and eight of 251 (3%) participants and 11 of 246 (5%) participants experienced a cardiac disorder in the liraglutide 1.2 mg and 1.8 mg arms, respectively (between-group differences of 1 to 3% for sulfonylurea versus liraglutide).113 The other RCT (N=200) was 52 weeks and reported higher rates of vascular (7.6% versus 6.3%; between-group difference, 1.3%) and cardiac (10.6% versus 6.3%; between-group difference, 4.3%) disorders in the sulfonylurea arm compared with the liraglutide arm.110 Of note, Kaku et al. used low doses of glibenclamide (1.25 to 2.5 mg/day) and liraglutide (0.9 mg/day).110 (SOE: Low; GLP-1 receptor agonists favored)
- 199. 142 Metformin Versus aMetformin-Based Combination Comparison Metformin Versus a Combination of Metformin Plus a Thiazolidinedione Six short-term (18 to 32 weeks) RCTs found a non-significant increase in cardiovascular morbidity for metformin plus rosiglitazone versus metformin (pooled OR, 1.59; 95% CI, 0.60 to 4.25) (Figure 51).59, 118-121, 247 Removal of any one study did not change the inference of the meta-analysis. The pooled risk difference for short-term CVD morbidity for metformin plus rosiglitazone versus metformin was 0.4% (95% CI, -0.2 to 1.1%). Another longer RCT (80 weeks of followup) reported four ischemic events (4/344, 1.7%) and five cerebrovascular events (5/344, 1.5%) in the metformin plus rosiglitazone arm compared with four ischemic events (4/334, 1.2%) and three cerebrovascular events (3/334, 0.9%) in the metformin arm.127 Of note, the text of article contradicts results in the table (reported here); the text reports five ischemic events in the metformin arm.127 (SOE: Low; Metformin favored over combination of metformin plus rosiglitazone for short-term cardiovascular morbidity) Figure 51. Pooled odds ratio of cardiovascular morbidity comparing metformin with a combination of metformin plus rosiglitazone CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus rosiglitazone; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. A single 26-week RCT compared metformin to metformin plus pioglitazone (dosed at 15, 30, and 45 mg in separate arms) and reported one nonfatal stroke in the metformin arm (1/129) and did not report on this outcome for the metformin plus pioglitazone arms (n=388).126 (SOE: Insufficient)
- 200. 143 Metformin Versus aCombination of Metformin Plus a Sulfonylurea One 6-month RCT,134 which was a good although older study, reported rates of 5% and 14% for unspecified cardiovascular events in the metformin versus combination metformin plus sulfonylurea arm, respectively.134 (SOE: Low; Metformin favored for short-term CVD morbidity) Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Six short RCTs compared the combination of metformin plus a DPP-4 inhibitor with metformin monotherapy and found no significant difference in short-term cardiovascular morbidity based on 11 events across studies (pooled OR, 1.90; 95% CI, 0.57 to 6.36) (Figure 52).118, 142, 146, 147, 152, 160 We did not find statistical evidence of heterogeneity, and removal of any one study did not change the inference of this meta-analysis. The pooled risk difference for short-term cardiovascular morbidity for metformin plus a DPP-4 inhibitor versus metformin was 0.3% (95% CI, -0.4% to 1.1%). Figure 52. Pooled odds ratio of short-term cardiovascular morbidity comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. We did not include one RCT (N=651) in the meta-analysis because of its longer duration (76 weeks). The investigators reported a rate of 2.1% for acute cardiovascular adverse events in the metformin arm, 0.3% in the metformin plus saxagliptin 5 mg arm, and did not report this rate in the metformin plus saxagliptin 10 mg arm.87 We did not include three additional RCTs in the meta-analysis because of a lack of reporting of events in at least one arm precluding estimation of an OR (Table 45). Cardiovascular
- 201. 144 morbidity was slightlyhigher in the combination therapy arms based on the limited results reported.84, 151, 164 Table 45. Randomized controlled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on cardiovascular morbidity Author, Year Followup (Weeks) Outcome Number of Events / N (%) for Metformin Vs. Metformin + DPP-4 Inhibitor Raz, 2008 142 30 Acute MI Results presented in Figure 52 Yang, 2011 146 24 Acute myocardial ischemia or MI Results presented in Figure 52 Wang, 2015 160 24 Nonfatal MI Results presented in Figure 52 Scott, 2008 118 18 Acute CV event Results presented in Figure 52 Fonseca, 2012 147 18 Acute CV events (myocardial ischemia or MI Results presented in Figure 52 Ross, 2012 152 12 Acute MI Results presented in Figure 52 Pfutzner, 2011 87 76 Acute CV events NR/328 (2.1) Metformin + saxagliptin 5 mg: NR/320 (0.3) Metformin + saxagliptin 10 mg: NR/323 (NR) Haak, 2013 164 52 Nonfatal MI 1/170 (0.6) Metformin 2000 mg + linagliptin: NR /171 Metformin 1000 mg + linagliptin: 3/225 (1.3) Unstable angina NR/170 (0.6) Metformin 2000 mg + linagliptin: 2/171 (1.2) Metformin 1000 mg + linagliptin: 2/225 (0.9) White, 2014 151 12 Nonfatal MI NR/86 1/74 (1.4) Pratley*, 2014 84 26 Nonfatal MI Metformin 2000 mg: NR/111 Metformin 1000 mg: NR/109 Metformin 2000 mg + alogliptin: NR/114 Metformin 1000 mg + alogliptin: NR/106 CV = cardiovascular; DPP-4 = dipeptidyl peptidase-4; mg = milligrams; MI = myocardial infarction; NR = not reported * included in table even though did not report on outcome for any of the arms because study did provide results for alogliptin monotherapy arm implying that there were likely no nonfatal myocardial infarctions in the other arms Four short-term RCTs comparing metformin with metformin plus a DPP-4 inhibitor reported on nonfatal stroke but did not report on events in all arms (Table 46).84, 126, 160, 164 Nonfatal strokes were uncommon and appeared more common in the metformin monotherapy arms based on limited reporting of results on this outcome. (SOE: Insufficient for short-term cardiovascular morbidity) Table 46. Randomized controlled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on nonfatal stroke Author, Year Followup (Weeks) Number of Events / N (%) for Metformin Vs. Metformin + DPP-4 Inhibitor Haak, 2013 164 52 1/170 (0.6) Metformin 2000 mg + linagliptin: NR/171 Metformin 1000 mg + linagliptin: 1/225 (0.4) DeFronzo, 2012 126 26 1/129 (0.8) Metformin + alogliptin 12.5 mg: NR/128 Metformin + alogliptin 25 mg: NR/129 Pratley, 2014 84 26 Metformin 2000 mg: NR/111 Metformin 1000 mg: NR/109 Metformin 2000 mg + alogliptin: NR/114 Metformin 1000 mg + alogliptin: NR/106 Wang, 2015 160 24 1/100 (1) Metformin + linagliptin 5 mg: 0/205 (0) DPP-4 = dipeptidyl peptidase-4; mg = milligrams; NR = not reported
- 202. 145 Metformin Versus aCombination of Metformin Plus an SGLT-2 Inhibitor A single 24-week RCT (N=182) compared the effects of metformin with metformin plus dapagliflozin on cardiovascular morbidity. This study reported that two participants (2/91, 2.2%) developed angina pectoris in the metformin plus dapagliflozin arm and that there were no events in the metformin arm (0%).169 The investigators also reported one transient ischemic attack in the metformin plus dapagliflozin arm (1/91, 1.1%) and did not report on this outcome in the metformin arm.169 (SOE: Low; Metformin favored for short-term cardiovascular morbidity) Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea Two short-term RCTs compared metformin plus pioglitazone with metformin plus a sulfonylurea and reported mixed results on cardiovascular morbidity.183, 185 One trial (N=288) with submaximally-dosed pioglitazone (30 mg/day) reported three events (coronary heart disease, carotid artery stenosis, and peripheral artery disease) among 142 participants (2%) at 24 weeks in the metformin plus sulfonylurea arm and did not report on events in the metformin plus pioglitazone arm.185 The other RCT (N=250) reported one acute myocardial infarction at 24 weeks in the metformin plus pioglitazone arm (1/103, 1%) versus no events in the metformin plus sulfonylurea arm (0%).183 A single retrospective cohort study from a Veterans Affairs population with Medicare (N=80,936) compared the combination of metformin plus a thiazolidinedione with metformin plus a sulfonylurea and found a non-significant increase in risk of stroke or myocardial infarction (composite outcome) for sulfonylurea-based versus thiazolidinedione-based therapy: adjusted HR, 1.15 (95% CI, 0.8 to 1.66; p=0.46); minimum followup was 12 months, but mean duration of followup was not reported.242 (SOE: Insufficient) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor Two RCTs compared metformin plus pioglitazone with metformin plus a DPP-4 inhibitor at 26 weeks and reported on cerebrovascular events.126, 188 One reported a cerebrovascular accident in each arm (1/165, 1% in the metformin plus pioglitazone arm and 1/166, 1% in the metformin plus sitagliptin arm),188 and the other did not report on cerebrovascular events in the combination therapy arms (reported events for monotherapy as discussed above).126 Bergenstal 2010 et al., also reported three cardiovascular events (unstable angina, n=1; coronary artery occlusion, n=2) in the metformin plus pioglitazone group and no events in the metformin plus sitagliptin (0/166, 0%) group.188 This RCT did not use an intention-to-treat approach and had differential and large losses to followup (13% in the metformin plus DPP-4 inhibitor arm and 21% in the metformin plus pioglitazone arm).188 (SOE: Low; Combination of metformin plus a DPP-4 inhibitor favored over metformin plus pioglitazone for short-term cardiovascular morbidity) Two RCTs compared metformin plus rosiglitazone with metformin plus sitagliptin (duration 16 to 18 weeks) and reported on cardiovascular and cerebrovascular morbidity.118, 186 The trial evaluating cardiovascular events reported none in the metformin plus rosiglitazone arm (0/87, 0%) and two coronary artery disease events in the metformin plus sitagliptin arm (2/94, 2.1%).118 The other trial (N=169) reported a transient ischemic attack in each arm.186 (SOE: Low;
- 203. 146 combination of metforminplus rosiglitazone favored over metformin plus DPP-4 inhibitor for short-term cardiovascular morbidity) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist One RCT (N=325) compared metformin plus pioglitazone with metformin plus exenatide at 26 weeks and reported one cerebrovascular accident (1/165, 1%) and three cardiac events (unstable angina and coronary artery occlusions; 3/165, 2%) in the metformin plus pioglitazone arms and no events in the metformin plus exenatide arm (0/160, 0%).188 (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor Four RCTs reported on long-term cardiovascular morbidity for the comparison of metformin plus sulfonylurea with metformin plus a DPP-4 inhibitor (Tables 47 and 48).192, 194-197 Three of these 104-week studies provided results on non-fatal myocardial infarction and showed a non-significant decrease in fatal myocardial infarction for the combination of metformin plus a DPP-4 inhibitor versus metformin plus a sulfonylurea (pooled OR, 0.68; 95% CI, 0.31 to 1.50) (Figure 53).192, 194, 196, 197 We did not find statistical heterogeneity (I-squared = 0%). The pooled risk difference in non-fatal myocardial infarction for metformin plus a DPP-4 inhibitor versus metformin plus a sulfonylurea was -0.2% (95% CI, -0.5 to 0.2%). Evidence on long-term cerebrovascular morbidity was mixed. One RCT reported higher rates of stroke for metformin plus a sulfonylurea compared with metformin plus a DPP-4 inhibitor;194 another found similar rates across arms,197 and a third reported a single event in the metformin plus sulfonylurea arm but did not report on this outcome in the metformin plus DPP-4 inhibitor arm (Table 48).195 Of note, sulfonylurea doses were submaximal in these RCTs. (SOE: Moderate; Combination of metformin plus a DPP-4 inhibitor favored for long-term non-fatal myocardial infarction)
- 204. 147 Table 47. Randomizedcontrolled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor on cardiovascular morbidity Author, Year Followup Outcome Number of Events / N (%) for Metformin + Sulfonylurea Vs. Metformin + DPP-4 Inhibitor Nauck, 2007 192, 196 104 weeks MI 1/588 (0.2) vs. 0/584 (0) Gallwitz, 2012 194 104 weeks Nonfatal MI 10/775 (1.3) vs. 6/776 (0.8) Unadjusted RR, 0.6; 95% CI, 0.22 to 1.64 Reference = metformin + sulfonylurea Composite: CV death, MI, stroke, or admission to hospital due to unstable angina 26/775 (3.4) vs. 12/776 (1.5) Unadjusted RR, 0.46; 95% CI, 0.23 to 0.91 Reference = metformin + sulfonylurea Admission to hospital due to unstable angina 3/775 (0.4) vs. 3/776 (0.4) Unadjusted RR, 1.0; 95% CI, 0.2 to 4.93 Reference = metformin + sulfonylurea Del Prato, 2014 197 104 weeks Adjudicated non-fatal MI 4/869 (0.5) vs. 1/873 (0.1) (alogliptin 12.5 mg) 4/869 (0.5) vs. 4/878 (0.5) (alogliptin 25 mg) Goke, 2010 195 104 weeks Not defined Qualitative statement: “The incidences of CV AEs…were low and similar between treatment groups.” AE = adverse event; CI = confidence interval; CV = cardiovascular; DPP-4 = dipeptidyl peptidase-4; MI = myocardial infarction; RR = risk ratio Table 48. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor on cerebrovascular morbidity Author, Year Followup Outcome Number of Events / N (%) for Metformin + Sulfonylurea Vs. Metformin + DPP-4 Inhibitor Gallwitz, 2012 194 104 weeks Cerebral infarction 4/775 (0.5) vs. 0/776 (0) Nonfatal stroke* 11/775 (1.4) vs. 3/776 (0.4) Unadjusted RR, 0.27; 95% CI, 0.08 to 0.97 Reference = metformin + sulfonylurea Goke, 2010 195, 248 104 weeks Transient ischemic attack 1/430 (0.2) vs. NR/428 Del Prato, 2014 197 104 weeks Adjudicated non-fatal stroke 3/869 (0.3) vs. 3/873 (0.3) (alogliptin 12.5 mg) 3/869 (0.3) vs. 2/878 (0.2) (alogliptin 25 mg) CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; NR = not reported; RR = risk ratio * This outcome appears to include stroke and transient ischemic attacks.
- 205. 148 Figure 53. Pooledodds ratio of cardiovascular morbidity comparing combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin plus a sulfonylurea; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor A single RCT (N=814) with 208 weeks of follow up reported no CVD events (not otherwise defined) for the combination of metformin plus glipizide or metformin plus dapagliflozin; dose of glipizide or dapagliflozin achieved was not reported.54 (SOE: Low; Neither favored for long- term CVD morbidity) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two short-term (26-week) RCTs (N=991) suggested no difference in cardiovascular morbidity for the combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist.188, 210 One study reported no cardiovascular events in either arm (defined as unstable angina or coronary artery occlusion), and the other study reported “cardiac disorders” in one participant in the metformin plus sitagliptin and the metformin plus liraglutide 1.8 mg arm.188, 210 One of these trials reported on cerebrovascular accidents and reported one event in the metformin plus sitagliptin arm and no events in the metformin plus exenatide arm.188 (SOE: Low; Neither favored for short-term cardiovascular morbidity) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a Basal Insulin A single RCT (N=501) reported that five participants experienced cardiovascular events (carotid artery occlusion, angina pectoris, and unstable angina) in the metformin plus insulin
- 206. 149 glargine arm (5/237,2%) and two participants experienced cardiovascular events (nonfatal acute myocardial infarction and angina pectoris) in the metformin plus sitagliptin arm (2/264, 1%) at 25 weeks.211 Of note, events had to be considered “serious” to be reported.211 This study did not use an intention-to-treat analysis and had moderate, differential losses to followup across arms (5% for sitagliptin and 9% for insulin glargine).211 (SOE: Low; Combination of metformin plus a DPP-4 inhibitor favored for short-term cardiovascular morbidity) Combination of Metformin Plus a Basal Insulin Versus a Combination of Metformin Plus a Premixed Insulin In a 16-week cross-over study, 105 participants with newly-diagnosed type 2 diabetes were randomly assigned to metformin plus insulin glargine or metformin plus insulin lispro 75/25 twice daily.223 During an 8-week lead-in period, participants received neutral protamine Hagedorn (NPH) insulin at night, and the metformin dose was titrated. One participant experienced a myocardial infarction during the lead-in period and one participant experienced chest pain during treatment with premixed insulin; the investigators did not report if this event occurred before or after the crossover.223 (SOE: Insufficient) Strength of Evidence for Cardiovascular and Cerebrovascular Morbidity We did not find any high strength evidence for cardiovascular and cerebrovascular morbidity. Most evidence was low or insufficient because of a paucity of studies reporting on these outcomes (see Key Points, Table 49, Table 50, and Table 51). Notably, none of the RCTs was designed to evaluate cardiovascular outcomes, and the RCTs tended to be short (less than 12 months), and event rates were low. We identified a mixture of RCT and observational study evidence for these outcomes for the monotherapy comparisons, but only RCTs for combination comparisons. Most of the evidence was at medium or high risk of bias. Common study limitations included lack of reporting on randomization and masking procedures and lack of an intention-to-treat approach combined with substantial losses to followup. The consistency of this evidence was limited by the small number of studies and differences in definitions. While observational studies offer the opportunity for precision given their frequently large sizes, most evidence was still imprecise since we did not identify that many high-quality observational studies. Notably, the observational studies did not tend to corroborate RCT findings. We identified unpublished studies for several comparisons. We identified one unpublished study describing the long-term followup (156 weeks) of one of the included studies141 that had not reported on CVD morbidity in its publication. This study compared metformin, metformin plus a sulfonylurea, metformin plus a DPP-4 inhibitor, and metformin plus a GLP-1 receptor agonist. Rescue therapy was highest in the metformin arm (rescue therapy rates: 59, 33, 36, and 26% for the metformin, metformin plus a sulfonylurea, metformin plus a DPP-4 inhibitor, and metformin plus a GLP-1 receptor agonist arms, respectively). Given this pattern of rescue therapy (differential use of medications across arms), these unpublished findings from comparisons of combination therapies with metformin alone were dissimilar to the published results: participants in the metformin arm experienced higher rates of CVD morbidity than those in the metformin plus a sulfonylurea and metformin plus a DPP-4 inhibitor arms. This study also provided a comparison of metformin to metformin plus a GLP-1 receptor agonist, but conclusions are limited by the high and differential use of rescue therapy in this study.
- 207. 150 We identified twoshort-term, unpublished studies comparing sulfonylurea with DPP-4 inhibitors which were not completely consistent with our findings (one suggested similar myocardial infarction rates but increased cerebrovascular events for sulfonylurea, and the other suggested increased risk of unstable angina for sulfonylurea users versus DPP-4 inhibitor users). We identified two unpublished RCTs with long-term follow up comparing the combination of metformin plus a sulfonylurea with metformin plus a DPP-4 inhibitor. Results were consistent with our finding that metformin plus a DPP-4 inhibitor was associated with lower coronary heart disease risk than metformin plus a sulfonylurea. Inclusion of these studies may have raised our strength of evidence for this comparison. We identified one published and one unpublished study of metformin plus a basal insulin versus metformin plus a premixed insulin. The published study did not report on events in both arms whereas the unpublished study suggested an increased risk of cardiovascular morbidity (consistent with the reporting of an event in the metformin plus premixed arm in the published RCT). Therefore, the unpublished study could have led to a higher grade for the strength of this evidence which might have supported a conclusion about this comparison. Finally, we identified unpublished studies for three comparisons for which we did not have published studies. As described above, an unpublished report of the long-term follow up of an included study141 which did not report on CVD morbidity compared metformin plus a sulfonylurea with metformin plus a GLP-1 receptor agonist and suggested that coronary heart disease risk might be higher for metformin plus a GLP-1 receptor agonist versus metformin plus a sulfonylurea. A short-term RCT comparing pioglitazone with a DPP-4 inhibitor suggested increased short-term CVD morbidity with DPP-4 inhibitor monotherapy. Another unpublished study found similar rates of cerebrovascular morbidity (cerebral infarctions) for pioglitazone and exenatide.
- 208. 151 Table 49. Strengthof evidence domains for monotherapy comparisons in terms of cardiovascular and cerebrovascular morbidity among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. rosiglitazone RCTs: 2 (4828) High Inconsistent Direct Imprecise Undetected Low Metformin favored for long-term CVD morbidity Observational: 3 (94,304) Medium Inconsistent Direct Precise N/A Metformin vs. pioglitazone RCTs: 3 (158) Medium Consistent Direct Imprecise Undetected Low Neither treatment favored Observational: 2 (58,883) Medium Consistent Direct Precise N/A Metformin vs. SU RCT: 3 (4808) High Inconsistent Direct Imprecise Undetected Low Metformin favored for long-term CVD morbidity Observational: 6 (545,686) Medium Consistent Direct Precise N/A Metformin vs. DPP-4 inhibitors RCT: 2 (2,090) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Observational: 1 (84,756) Medium Unknown Direct Precise N/A Inadequate reporting of events in all arms Metformin vs. SGLT-2 inhibitors RCT: 1 (408) High Unknown Direct Imprecise N/A Insufficient Unable to determine Rosiglitazone vs. SU RCT: 2 (4711) High Consistent Direct Imprecise Undetected Low SU favored for long- term CVD morbidity Observational: 3 (142,502) Medium Inconsistent Direct Precise N/A Pioglitazone vs. SU RCT: 1 (502) Medium Unknown Direct Imprecise Undetected Low Pioglitazone favored for short-term CVD morbidity Observational: 2 (107,081) Medium Inconsistent Direct Precise N/A SU vs. DPP-4 inhibitors RCTs: 2 (653) Medium Consistent Direct Imprecise Undetected Low DPP-4 inhibitor favored for short-term CVD morbidity SU vs. GLP-1 receptor agonists RCTs: 2 (1157) High Consistent Direct Imprecise Undetected Low GLP-1 receptor agonist favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 209. 152 Table 50. Strengthof evidence domains for metformin versus metformin-based combination comparisons in terms of cardiovascular and cerebrovascular morbidity among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + rosiglitazone (shorter duration studies) RCTs: 7 (3136) Medium Consistent Direct Imprecise Undetected Low Metformin favored for short-term CVD morbidity Metformin vs. metformin + pioglitazone RCTs: 1 (1,554) High Unknown Direct Imprecise Undetected Insufficient Unable to determine Metformin vs. metformin + SU (shorter duration study) RCT: 1 (110) Low Unknown Direct Imprecise Undetected Low Metformin favored for short-term CVD morbidity Metformin vs. metformin + DPP-4 inhibitor (shorter duration studies) RCTs: 11 (4351) Low Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Inadequate reporting of events in all arms Metformin vs. metformin + SGLT-2 inhibitor (shorter duration study) RCT: 1 (182) Low Unknown Direct Imprecise Undetected Low Metformin favored for short-term CVD morbidity DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 210. 153 Table 51. Strengthof evidence domains for combination therapy comparisons in terms of cardiovascular and cerebrovascular morbidity among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Met + TZD vs. Met + SU RCTs: 2 (538) Obs: 1 (80,936) High Medium Inconsistent Unknown Direct Direct Imprecise Imprecise Undetected N/A Insufficient Unable to determine Met + pioglitazone vs. Met + DPP-4 inhibitor (shorter duration studies) RCTs: 2 (2068) High Consistent Direct Imprecise Undetected Low Met + DPP-4 inhibitor favored for short-term cardiovascular morbidity Met + rosiglitazone vs. Met + DPP-4 inhibitor (shorter duration studies) RCTs: 2 (350) High Unknown Direct Imprecise Undetected Low Met + rosiglitazone favored for short-term CVD morbidity Met + pioglitazone vs. Met + GLP-1 receptor agonist (shorter duration study) RCT: 1 (325) High Unknown Direct Imprecise Undetected Low Met + GLP-1 receptor agonist favored for short-term CVD morbidity Met + SU vs. Met + DPP- 4 inhibitor (long-term non- fatal MI) RCTs: 4 (5049) Low Inconsistent Direct Imprecise Undetected Low Met + DPP-4 inhibitor favored for long-term non-fatal MI ‡ Met + SU vs. Met + SGLT-2 inhibitor (long- term) RCT: 1 (814) Medium Unknown Direct Imprecise Undetected Low Neither favored Met + DPP-4 inhibitor vs. Met + GLP-1 receptor agonist (short-term studies) RCTs: 2 (1179) Medium Consistent Direct Imprecise Undetected Low Neither favored Met + DPP-4 inhibitor vs. Met + basal insulin (shorter duration studies) RCTs: 1 (501) High Unknown Direct Imprecise Undetected Low Met + DPP-4 inhibitor favored for short-term CVD morbidity Met + basal insulin vs. Met + premixed insulin RCTs: 1 (105) Medium Unknown Direct Imprecise Suspected Insufficient Unable to determine DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; Obs = observational; OR = odds ratio; RD = absolute risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the clinical outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95% confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence. ‡ The evidence for long-term cerebrovascular morbidity was insufficient.
- 211. 154 Evidence for Retinopathy MonotherapyComparisons Sulfonylureas Versus GLP-1 Receptor Agonists A single RCT (N=400) compared rates of retinopathy at 52 weeks in participants randomized to submaximally dosed glibenclamide or submaximally dosed liraglutide. Nine (9/132, 6.8%) and 16 (16/268, 6%) participants were diagnosed with retinopathy as a “treatment-emergent adverse event” in the sulfonylurea and GLP-1 receptor agonist arms, respectively.110 The study did not report on baseline rates of retinopathy.110 Losses to follow up were greater than 15% in each arm, and the investigators did not use an intention-to-treat approach for this analysis.110 (SOE: Low; Neither favored) Metformin Versus a Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Two RCTs (N=437) compared low-dose metformin with low-dose metformin plus a DPP-4 inhibitor at 12 weeks and reported on retinopathy.149, 157 Each trial reported one case of retinopathy in the metformin monotherapy arm. One of the RCTs reported one case of retinopathy in the metformin plus DPP-4 inhibitor arm (alogliptin 12.5 mg) and no cases in in the metformin plus alogliptin 25 mg arm.157 The other RCT reported no cases in the metformin plus sitagliptin arm.149 (SOE: Low; Neither favored short-term) Strength of Evidence for Retinopathy We identified three RCTs (and no observational studies) evaluating retinopathy, and one study was of poor quality.110 Therefore, evidence is mainly insufficient for this outcome (see Key Points and Table 52). We identified an unpublished study comparing pioglitazone with exenatide which reported similar rates of blurred vision in both arms.
- 212. 155 Table 52. Strengthof evidence domains for comparisons in terms of retinopathy among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † SU vs. GLP-1 receptor agonists Shorter duration study RCT: 1 (411) High Unknown Direct Imprecise Undetected Low Neither arm favored Metformin vs. metformin + DPP-4 inhibitors Shorter duration studies RCTs: 2 (437) Low Inconsistent Direct Imprecise Undetected Insufficient Unable to determine DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of evidence. Unless otherwise specified, conclusions for retinopathy are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 213. 156 Evidence for Nephropathy Weincluded the following nephropathy outcomes: categorical definitions of new or progressive nephropathy, changes in urine albumin, and changes in estimated glomerular filtration rate. Studies of comparisons including a SGLT-2 inhibitor are described in the section on renal insufficiency in this report (see Key Question 3 – renal insufficiency). Monotherapy Comparisons Metformin Versus Thiazolidinediones One RCT and two retrospective cohort studies evaluated this outcome and found mixed results.62, 249, 250 The 12-month RCT (N=1,194) reported a significant decrease in the urine albumin-to-creatinine ratio for participants randomized to the pioglitazone arm compared with the metformin arm (19% versus 1% reduction; P = 0.002).62 The smaller retrospective cohort study of 1,271 patients in the US (Baylor Health Care System, Dallas, TX and Christiana Care Health System, Newark, DE) evaluated nephropathy defined as new albuminuria or new estimated glomerular filtration rate (eGFR) rate below 60 mL/min/1.73 m2 and found no significant difference between metformin users and thiazolidinedione users, for either outcome (adjusted HR, 1.00 and 1.04 for thiazolidinedione versus metformin), with median followup of 2.8 to 3.2 years.250 The large, although short, retrospective cohort study from the Veterans Health Administration (N=93,577) with median followup of 0.7 to 0.9 years also found no significant difference in nephropathy for thiazolidinedione versus metformin users. The adjusted HR for the composite outcome of an eGFR event (persistent decline of 25% or greater from baseline eGFR) or end-stage renal disease (eGFR less than 15 mL/min/1.73 m2 or first inpatient or outpatient code for dialysis or related procedures or renal transplant) was 0.92 (95% CI, 0.71 to 1.18).249 (SOE: Insufficient) Metformin Versus Sulfonylureas One small (N=51) 3-month RCT reported that microalbuminuria decreased significantly with metformin (P = 0.008) and increased non-significantly with glibenclamide (P = 0.09). eGFR was stable in the metformin arm (P = 0.46) and increased in the sulfonylurea arm (P = 0.04).251 The study did not provide statistical comparisons between groups. Three retrospective cohort studies in the United States suggested a decreased risk of nephropathy among metformin versus sulfonylurea users (Table 53).249, 250, 252 Two of the studies were from the Veterans Health Administration and it is not clear if the study populations overlapped.249, 252 (SOE: Low; Metformin favored)
- 214. 157 Table 53. Retrospectivecohort studies comparing metformin with sulfonylureas on nephropathy Author, Year Population Followup Outcome Definition HR (95% CI) Hung, 2013 252 Veterans Health Administration VA Mid-South VISN 9 Data Warehouse (N=13,238) Incident medication users Approximately 1 year Composite of GFR event or ESRD* All: 0.85 (0.72 to 1.01) Urine protein measures at baseline: 0.78 (0.64 to 0.97) Reference = sulfonylurea Hung, 2012 249 Veterans Health Administration N=93,577 Did not appear to exclude incident users 0.7 to 0.9 years (median) Composite of GFR event or ESRD* 1.2 (1.13 to 1.28) Reference = metformin Masica, 2013 250 1,271 participants from an electronic health record database in the US (Baylor Health Care System, Dallas, TX) and Christiana Care Health System, Newark, DE) Median follow up of 2.8 to 3.2 years Microalbuminuria or worse Urine protein measures available: 1.27 (0.93 to 1.74) Reference = metformin eGFR ≥60 ml/min/1.73m 2 at first measurement and an eGFR <60 ml/min/1.73m 2 during follow-up eGFR available: 1.41 (1.05 to 1.91) Reference = metformin CI = confidence interval; eGFR = estimated glomerular filtration rate; ESRD = end-stage renal disease; HR = hazard ratio; ml/min/1.73m2 = milliliters per minute per 1.73 meters squared; VA = Veterans Affairs; VISN = Veterans Integrated Service Network * GFR event= persistent 25% or greater decline from the baseline eGFR; ESRD: eGFR <15 mL/minute/1.73m2 or first code for dialysis or related procedure or renal transplant Thiazolidinediones Versus Sulfonylureas Five small RCTs provided mixed results on the effect of thiazolidinediones and sulfonylureas on nephropathy outcomes, and all studies reported on albuminuria as the outcome.96-98, 102, 253 Four trials found less albuminuria in patients receiving pioglitazone compared with a sulfonylurea;96, 97, 102, 253 only one reported a significant difference.96 One trial compared rosiglitazone and glyburide at 12 months and found no statistically significant difference in the urinary albumin-to-creatinine ratio; progression to microalbuminuria did not differ between groups.98 A retrospective cohort study using a small US database reported a non-significant increased risk of nephropathy (incident albuminuria) among sulfonylurea versus thiazolidinedione users who had a measure of urine protein (adjusted HR, 1.27; 95% CI, 0.93 to 1.74).250 (SOE: Low; Thiazolidinediones favored) Sulfonylureas Versus DPP-4 Inhibitors A single RCT analyzed changes in eGFR and urine albumin-to-creatinine ratio from baseline among 277 participants randomized to glipizide or sitagliptin.107 Over 54 weeks, the eGFR decreased slightly in both arms (-3.3 and -3.9 ml/min/1.73m2 for glipizide and sitagliptin, respectively) and urine albumin-to-creatinine ratio increased slightly in both arms (0.1 and 0.06 for glipizide and sitagliptin, respectively). Of note, approximately 20 percent of participants were
- 215. 158 lost to followupin each arm (423 participants originally randomized), and the investigators did not conduct an intention-to-treat analysis. (SOE: Low; Neither favored) Sulfonylureas Versus GLP-1 Receptor Agonists A single RCT (N=745) reported similar cumulative incidences of 6 percent, 5 percent and 5 percent of “renal and urinary disorders” at 104 weeks in patients randomized to glimepiride 8 mg, liraglutide 1.2 mg, and liraglutide 1.8 mg.113 (SOE: Low; Neither favored) DPP-4 Inhibitors Versus GLP-1 Receptor Agonists A single small RCT with 24 weeks of followup compared sub-maximal sitagliptin (50 mg) to sub-maximal liraglutide (0.9 mg daily) and reported negligible changes in eGFR and urinary albumin excretion in both arms.115 (SOE: Low; Neither favored for short-term nephropathy outcomes) Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea Two RCTs compared metformin plus rosiglitazone with metformin plus a sulfonylurea. Evaluation of the urine albumin-to-creatinine ratio favored the combination of metformin plus rosiglitazone.177, 182 One small RCT (N=34), with 48 weeks of followup, reported a negligible decrease in urine albumin-to-creatinine ratio in the metformin plus thiazolidinedione arm (-0.77 mg/g) and a small increase (12.2 mg/g) in the metformin plus sulfonylurea arm.182 The larger trial (N=389) reported a greater, although non-significant, reduction in the urine albumin-to- creatinine ratio with the combination of metformin plus a thiazolidinedione versus metformin plus a sulfonylurea arm, at 32 weeks.177 The smaller RCT also reported a negligible decrease in eGFR in the metformin plus thiazolidinedione arm (-1.48 ml/min/1.73m2 ) and an increase in eGFR in the metformin plus sulfonylurea group (9.97 ml/min/1.73m2 ); the study did not provide a statistical comparison of the between-group difference.182 This very small trial reported more than 20 percent losses to followup across arms and did not use an intention-to-treat analysis for nephropathy.182 (SOE: Low; Combination of metformin plus a thiazolidinedione favored) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor A single RCT (N=514) with 26 weeks of followup reported on percent change in urine albumin-to-creatinine ratio for the combination of metformin plus pioglitazone and the combination of metformin plus sitagliptin and found similar changes from baseline in both arms: -4% (95% CI, -17% to 12.1%) for metformin plus pioglitazone and -6.9% (95% CI, -20% to unclear but greater than 0%) for metformin plus sitagliptin.188 This study had 13 percent and 21 percent losses to followup in the sitagliptin- and pioglitazone-based arms, respectively, and did not use an intention-to-treat analysis for this outcome.188 (SOE: Low; Neither treatment favored)
- 216. 159 Combination of MetforminPlus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist A single RCT (N=514) with 26 weeks of followup reported on percent change in urine albumin-to-creatinine ratio for the combination of metformin plus pioglitazone and the combination of metformin plus exenatide and found a reduction in urine albumin-to-creatinine from baseline for the metformin plus exenatide arm [-16% (95% CI, -28% to -2%)] and no significant reduction for the metformin plus pioglitazone arm [-4% (95% CI, -17% to 12%)].188 This study had approximately 20 percent losses to followup in both arms and did not use an intention-to-treat analysis for this outcome.188 (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist A small RCT conducted in 42 participants with baseline microalbuminuria compared change in 24-hour urine albumin over 16 weeks for metformin plus sub-maximally dosed glimepiride and metformin plus exenatide. The authors reported a significant reduction in urine albumin in the metformin plus exenatide arm (-42 mg/day; P for change for baseline <0.01) compared to the metformin plus glimepiride arm (5 mg/day; P for change from baseline >0.05; calculated between group difference, -37 mg/day; P for between-group difference <0.001).202 This small study had more than 20 percent losses to followup across arms and did not use an intention-to- treat analysis.202 (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist A single RCT (N=514) with 26 weeks of followup reported on percent change in urine albumin-to-creatinine ratio for the combination of metformin plus sitagliptin and the combination of metformin plus exenatide.188 The trial found a statistically significant reduction in urine albumin-to-creatinine from baseline for the metformin plus exenatide arm (-16%; 95% CI, -28% to -2%) but not for the metformin plus sitagliptin arm (-6.9%; 95% CI, -20% to unclear but greater than 0%).188 This study had approximately 20 percent losses to followup in both arms and did not use an intention-to-treat analysis for this outcome.188 (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored) Strength of Evidence for Nephropathy We found low or insufficient strength of evidence on nephropathy outcomes for all comparisons of interest as described in the Key Points, Table 54, and Table 55. The evidence on nephropathy was limited by the lack of studies. For RCTs, major study limitations included small sample sizes and high rates of withdrawals (>20%), without use of an intention-to-treat approach. We could usually not determine consistency because of a lack of studies, and the evidence on all comparisons was imprecise because of insufficient sample size. We did not detect reporting bias. However, the small number of studies limited our ability to assess publication bias. Many of the studies did not provide measures of dispersion for nephropathy outcomes, but we did not believe that this was actually a source of selective analysis reporting bias as much as a reflection of a lack of a focus on reporting of these outcomes given that they were not primary outcomes.
- 217. 160 Table 54. Strengthof evidence domains for monotherapy comparisons in terms of nephropathy among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. TZD RCT: 1 (1194) Observational: 2 (94,848) Low Medium Unknown Inconsistent Direct Direct Imprecise Precise Undetected N/A Insufficient Unable to determine Metformin vs. SU (shorter duration studies) RCT: 1 (51) Observational: 3 (108,356) High Medium Unknown Consistent Indirect Direct Imprecise Precise Undetected N/A Low Metformin favored TZD vs. SU (mainly shorter duration studies) RCTs: 5 (308) Observational: 2 (1271) High Medium Consistent Unknown Direct Direct Imprecise Imprecise Undetected N/A Low TZD favored for short-term nephropathy outcomes SU vs. DPP-4 inhibitors (shorter duration study) RCT: 1 (423) High Unknown Indirect Imprecise Undetected Low Neither treatment favored SU vs. GLP-1 receptor agonists (longer duration study) RCT: 1 (746) Medium Unknown Indirect Imprecise Undetected Low Neither treatment favored DPP-4 inhibitors vs. GLP-1 receptor agonists (shorter duration study) RCT: 1 (56) High Unknown Direct Imprecise Undetected Low Neither treatment favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; N/A = not applicable; RCT = randomized controlled trial; SGLT- 2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Studies of comparisons including an SGLT-2 inhibitor are graded in the section on renal insufficiency in this report. Unless otherwise specified, conclusions for nephropathy are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 218. 161 Table 55. Strengthof evidence domains for metformin-based combination comparisons in terms of nephropathy among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + TZD vs. metformin + SU (shorter duration studies) RCT: 2 (433) High Consistent Direct Imprecise Undetected Low Metformin + TZD favored Metformin + TZD vs. metformin + DPP-4 (shorter duration study) RCT: 1 (514) High Unknown Direct Imprecise Undetected Low Neither treatment favored Metformin + TZD vs. metformin + GLP-1 receptor agonist (shorter duration study) RCT: 1 (514) High Unknown Direct Imprecise Undetected Low Metformin + GLP-1 favored Metformin + SU vs. metformin + GLP-1 receptor agonist (shorter duration study) RCT: 1 (42) High Unknown Direct Imprecise Undetected Low Metformin + GLP-1 favored Metformin + DPP-4 inhibitor vs. metformin + GLP-1 receptor agonist RCT: 1 (514) High Unknown Direct Imprecise Undetected Low Metformin + GLP-1 favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; RCT = randomized controlled trial; SGLT-2 inhibitors = sodium- glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Studies of comparisons including an SGLT-2 inhibitor are graded in the section on renal insufficiency in this report. Unless otherwise specified, conclusions for nephropathy are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 219. 162 Evidence for Neuropathy Forthe neuropathy analyses, we included studies where newly developed neuropathy was reported for each treatment group. Four short trials reported on neuropathy as an adverse outcome.121, 142, 154, 183 Monotherapy Comparisons Metformin Versus DPP-4 Inhibitors A single RCT compared the effects of metformin with alogliptin on undefined neuropathy. At 26 weeks of followup, one participant developed unspecified neuropathy in each of the alogliptin arms [12.5 mg (n=213) and 25 mg (n=210)]); neuropathy was not reported on in the metformin arm.154 (SOE: Insufficient) Metformin Versus a Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a Thiazolidinedione In a single RCT of 26 weeks duration, one withdrawal owing to undefined neuropathy occurred in the metformin arm (n=34); no events were reported on in the two metformin plus rosiglitazone arms (n=35 and n=36).121 (SOE: Insufficient) Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor One RCT reported the occurrence of undefined diabetic neuropathy as 2.1 percent among participants in the metformin arm (n=94) and 4.2 percent in the metformin plus sitagliptin arm (n=96) at 30 weeks.142 (SOE: Low; Metformin favored) Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea In a 6-month trial, neuropathy was described but was not a pre-specified outcome.183 One of 103 participants in the metformin plus thiazolidinedione arm developed neuropathy, and none of the 80 participants in the metformin plus sulfonylurea arm developed neuropathy.183 (SOE: Low; Neither favored) Strength of Evidence for Neuropathy The evidence grading for neuropathy is summarized in Table 56.
- 220. 163 Table 56. Strengthof evidence domains for comparisons in terms of neuropathy among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. DPP-4 inhibitors RCT: 1 (527) Low Unknown Direct Imprecise Undetected Insufficient Unable to determine Results not reported for all arms Metformin vs. metformin + TZD RCT: 1 (105) Medium Unknown Direct Imprecise Undetected Insufficient Unable to determine Results not reported for all arms Metformin vs. metformin + DPP-4 inhibitor (shorter duration study) RCT: 1 (190) High Unknown Direct Imprecise Undetected Low Metformin favored Metformin + TZD vs. metformin + SU (shorter duration study) RCT: 1 (183) High Unknown Indirect Imprecise Undetected Low Neither treatment favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for neuropathy are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 221. 164 Key Questions 3aand 3b: Safety Study Design and Population Characteristics We included 145 studies describing adverse effects for the comparisons of interest (Appendix D, Tables D10 to D13). We included 64 articles from our prior 2011 review16 and identified an additional 81 studies for this update. Six of the newly-included studies were updates of earlier studies.85, 87, 110, 113, 196, 210 The majority were RCTs (62 from the previous report and an additional 75 for the update). Most of the RCTs (109 out of 137, 80%) lasted 1 year or less. Sixteen RCTs (12%) had at least 2 years of followup. Few studies were designed explicitly to evaluate adverse events. Of the 22 studies designed to evaluate the adverse events specified in Key Question 3, most focused on cancer254, 255 and renal toxicity.249, 252 Thirty-one of 81 RCTs (38%) did not report on the use of rescue therapy; rescue therapy was allowed in 29 studies (36%) and was disallowed in 21 studies (26%). The mean age of participants ranged from approximately 40 years to 81 years, with the majority of studies reporting a mean age in the upper 50s. About 50 percent of the participants were female. Sixty-two studies did not report race or ethnicity. In the studies that reported race, the majority of the participants were Caucasians. No study included more than 25 percent African American participants; two studies included more than 70 percent Hispanic participants121, 186 and seven studies included more than 70 percent Asian participants.74, 92, 109, 155, 160, 254, 256 Risk of Bias We included 137 trials in this section. All of the trials were described as randomized. Fifty- one percent of the trials described their randomization scheme, and another 65 percent of the trials were described as being double-blinded. Thirty-six percent of all double-blinded RCTs also described the steps taken to ensure blinding. The majority of trials (86 percent) described the withdrawals and dropouts. Of the 16 RCTs with at least 2 years of followup, 12 had over 20% loss to followup. Of the eight observational studies included for this Key Question, 88 percent reported actual probability values and 63 percent described their measurement of the outcomes of interest. All studies described and adjusted for confounding factors and conducted statistical analyses. Seventy-five percent of studies described the number of participants who were lost to followup after the start of the period of observation. Key Points and Evidence Grades Hypoglycemia Mild, Moderate, or Total Hypoglycemia Unless otherwise noted, results on hypoglycemia refer to the number of participants experiencing hypoglycemia and not to the number of events. Metformin monotherapy was favored over the following: o Sulfonylurea monotherapy for mild-moderate hypoglycemia (pooled OR for sulfonylurea versus metformin, 4.00; 95% CI, 1.75 to 9.83) (SOE: High)
- 222. 165 o The combinationof metformin plus a thiazolidinedione (pooled OR for metformin plus a thiazolidinedione versus metformin monotherapy for total hypoglycemia, 1.56; 95% CI, 0.99 to 2.44) (SOE: Moderate) o The combination of metformin plus a sulfonylurea for mild, moderate, or total hypoglycemia (range in ORs, 2.15 to 28.6) (SOE: Moderate) o The combination of metformin plus an SGLT-2 inhibitor for mild or moderate hypoglycemia (pooled OR, 1.74; 95% CI, 0.83 to 3.66) (SOE: Moderate) The risks of mild-moderate and total hypoglycemia were similar for metformin monotherapy and the combination of metformin plus a DPP-4 inhibitor. (SOE: High) o Pooled OR for metformin plus a DPP-4 inhibitor versus metformin monotherapy: ▪ Mild-moderate hypoglycemia: 0.97; 95% CI, 0.63 to 1.51 ▪ Total hypoglycemia: 0.96; 95% CI, 0.55 to 1.67 Sulfonylurea monotherapy increased the risk of total hypoglycemia compared with thiazolidinedione monotherapy (pooled OR 6.31; 95% CI, 4.08 to 9.76). (SOE: High) SGLT-2 inhibitor monotherapy was associated with a lower risk of hypoglycemia compared with metformin monotherapy (pooled OR, 0.46; 95% CI, 0.16 to 1.30). (SOE: Moderate) DPP-4 inhibitor monotherapy was favored over sulfonylurea monotherapy (range of OR, 0.08 to 0.26 from individual studies for sulfonylurea versus DPP-4 inhibitor monotherapy). (SOE: Moderate) Mild-moderate hypoglycemia was more common with sulfonylurea monotherapy than with GLP-1 receptor agonist monotherapy (range in OR, 3.2 to 5.3). (SOE: Moderate) When compared with metformin plus a sulfonylurea, metformin plus an SGLT-2 inhibitor had less risk of mild or moderate hypoglycemia (range in OR, 0.03 to 0.13). (SOE: High) When compared with metformin plus sulfonylurea, several combinations had less risk of mild, moderate, or total hypoglycemia: metformin plus a thiazolidinedione, metformin plus a DPP-4 inhibitor, and metformin plus a GLP-1 receptor agonist (range in OR, 0.07 to 0.19). (SOE: High or Moderate for all comparisons) When compared with metformin plus basal insulin or premixed insulin, metformin plus a GLP-1 receptor agonist had less risk of mild or moderate hypoglycemia (range in OR, 0.18 to 0.35). (SOE: Moderate) Moderate grade evidence showed a lower risk of hypoglycemia when metformin is combined with a basal insulin rather than a premixed insulin (range in OR, 0.23 to 0.89). Severe Hypoglycemia Only the sulfonylurea comparisons convincingly demonstrated an increased risk of severe hypoglycemia in the sulfonylurea arms compared with nonsulfonylurea medications: o Sulfonylurea versus metformin (range in OR, 1.41 to 2.04) (SOE: Moderate) o Sulfonylurea versus thiazolidinediones (OR, 8.1) (SOE: Moderate) o Metformin plus sulfonylurea versus metformin plus SGLT-2 inhibitors, and metformin plus sulfonylurea versus metformin plus DPP-4 inhibitors. (SOE: Moderate or High) Gastrointestinal (GI) Side Effects GI adverse events are more common with:
- 223. 166 o Metformin thanwith DPP-4 inhibitors (pooled OR 2.6 and 2.7 for diarrhea and nausea, favoring DPP-4 inhibitors), (SOE: High); o Metformin than thiazolidinediones (between 1.7 to 4.2 fold higher odds), (SOE: Moderate); o Metformin than sulfonylureas (between 2.2 to 2.4 fold higher odds), (SOE: Moderate); o GLP-1 receptor agonists than metformin for nausea and vomiting (between 1.3 to 1.7 fold increased odds with GLP-1 receptor agonists). (SOE: Moderate) o GLP-1 receptor agonists than sulfonylureas (between 1.5 to 2.4 fold higher odds of diarrhea), (SOE: Moderate) o Metformin plus a GLP-1 receptor agonist than metformin plus a DPP-4 inhibitor (between 1.0 to 7.8 fold higher odds with metformin plus GLP-1 receptor agonists), (SOE: Moderate); o Metformin plus a GLP-1 receptor agonist than metformin plus a thiazolidinedione (between 2.9 to 6.3 fold higher odds with metformin plus a GLP-1 receptor agonist). (SOE: Moderate) GI adverse events are equally common with: o Thiazolidinediones and sulfonylureas (SOE: High) o Metformin monotherapy and metformin plus a DPP-4 inhibitor (SOE: Moderate for any GI adverse event, nausea, and vomiting for shorter duration studies) o Metformin plus a sulfonylurea and metformin plus a DPP-4 inhibitor in longer studies (SOE: High) o Metformin monotherapy and combination therapy with metformin plus a SGLT-2 inhibitor (for diarrhea) (SOE: Moderate) o Metformin plus a thiazolidinedione and metformin plus a sulfonylurea. (SOE: Moderate) Cancer Type of cancer was not designated a priori in most of the studies reporting on cancer; thus, the following conclusions apply to any cancer, unless specified. Even though the FDA has issued warnings regarding increased risk of bladder cancer risk with pioglitazone, we found low or insufficient strength of evidence on TZD-based comparisons and cancer outcomes. Despite FDA warnings of a possible increased risk of thyroid cancer with GLP-1 receptor agonists, we found low-strength or insufficient evidence on GLP-1 receptor agonist- based comparisons and cancer outcomes. Congestive Heart Failure Thiazolidinediones alone increase the risk of heart failure when compared with sulfonylureas (pooled OR in four RCTs of 1.6; 95% CI, 0.96 to 2.8) or metformin (two RCTs lasting less than a year with no events, one 4-year RCT with an absolute risk difference of 3% and range in HR of 1.2 to 1.5 in two observational studies with 6 to 8 years of followup). (SOE: Low) Despite recent concerns of congestive heart failure with DPP-4 inhibitors, we found low or insufficient strength of evidence on the comparative safety of this drug class for this outcome.
- 224. 167 Pancreatitis Despite FDAwarnings regarding an increased risk of pancreatitis with GLP-1 receptor agonists and DPP-4 inhibitors, we found low or insufficient evidence on the comparative safety of these drug classes for this outcome. Genital Mycotic Infections (for Comparisons That Include SGLT-2 Inhibitors) Compared with metformin monotherapy, genital infection rates were higher for SGLT-2 inhibitor monotherapy and for the combination of metformin plus an SGLT-2 inhibitor. Rates of genital infections for combination therapy with metformin plus an SGLT-2 inhibitor were higher compared to the following: o Metformin monotherapy: pooled OR for women, 3.0; 95% CI, 1.2 to 7.2 and pooled OR for men, 2.7; 95% CI, 0.8 to 9.0 (SOE: High) o Metformin plus a sulfonylurea: pooled OR for women, 5.2; 95% CI, 3.4 to 7.8; pooled OR for men, 7.6; 95% CI, 4.0 to 14.4 (SOE: High) o Metformin plus a DPP-4 inhibitor Other Serious Adverse Events There was no moderate or high strength of evidence for the following adverse events: liver injury, lactic acidosis, severe allergic reactions, macular edema/decreased vision, urinary tract infections (for SGLT-2 inhibitors) impaired renal function (for SGLT-2 inhibitors), fractures (for SGLT-2 inhibitors), and volume depletion (for SGLT-2 inhibitors). Therefore, we were unable to draw any firm conclusions regarding the diabetes medication comparisons and these safety outcomes. Evidence for Hypoglycemia Monotherapy Comparisons Metformin Versus Thiazolidinediones Five RCTs addressed hypoglycemia for metformin versus thiazolidinediones, finding no consistent differences in hypoglycemia by arm (Table 57).50, 70, 71, 73, 74 We were unable to conduct a meta-analysis because of differences in the definitions of hypoglycemia and lengths of followup. (SOE: Low; Metformin favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia)
- 225. 168 Table 57. Randomizedcontrolled trials comparing metformin with thiazolidinediones on hypoglycemia Author, Year Followup Metformin (Dose*) TZD (Dose*) Definition of Hypoglycemia Results † (Metformin Vs TZD) Kahn, 2006 50 4 years Metformin (max 2000 mg) Rosiglitazone (max 8 mg) Total (self- reported) 168/1454 (11.6%) vs 142/1456 (9.8%) Yoon, 2011 74 48 weeks Metformin (max 2000 mg; mean 1234.2 mg) Rosiglitazone (max 8 mg; mean 5.9 mg) Total (signs or symptoms) 4/114 (3.5%) vs 8/117 (6.8%) Erem, 2014 70 48 weeks Metformin (max 2000 mg) Pioglitazone (max 45 mg) Total (not specified) 0/19 (0%) vs 0/19 (0%) Russell-Jones, 2012 73 26 weeks Metformin (max 2500 mg) Pioglitazone (max 45 mg) Severe ‡ Total (signs or symptoms) 0/246 (0%) vs 0/163 (0%) 10/246 (4.1%) vs 6/163 (3.7%) Genovese, 2013 71 16 weeks Metformin (max 2550 mg) Pioglitazone (max 45 mg) Total (not specified) 0 episodes among 26 patients vs 4 episodes among 24 patients Max = maximum; mg = milligram; TZD = thiazolidinedione * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. ‡ Severe hypoglycemia was defined as symptoms resulting in loss of consciousness or seizure that showed prompt recovery after glucose administration, or documented blood glucose less than 3.0 mmol/L that required the assistance of another person because of severe impairment in consciousness. Metformin Versus Sulfonylureas Fifteen studies addressed this comparison (14 RCTs and one observational study).50, 60, 74, 129, 131-134, 136-138, 231, 257-259 Meta-analysis of five short-term RCTs deemed sufficiently homogeneous for quantitative synthesis suggested an increased risk of mild to moderate hypoglycemia for sulfonylureas versus metformin (pooled OR, 2.59; 95% CI, 0.98 to 8.86) (Figure 54). Exclusion of any one study did not change this inference. We did not include several studies in the meta-analysis because of differences in hypoglycemia definitions,131, 133, 134, 231 in study duration,50, 258 in how hypoglycemic events were reported,258 dosing of SU,129 and in study design.259 We did not include two studies in the meta- analysis of relative odds because neither had any events in either arm.138, 257 Results of the studies not included in the meta-analyses were consistent with the findings of the meta-analysis showing an increased risk of hypoglycemia for sulfonylurea versus metformin monotherapy (Table 58). Based on limited data to evaluate this, rates of hypoglycemia did not appear to be higher for the glyburide compared to other sulfonylurea arms across the studies. Three studies (two RCTs and one observational study) reported on severe hypoglycemia (range in OR 0.49 to 0.71; range in RD -1% to -23%), all favoring metformin (Table 58).134, 231, 259 (SOE: High; Metformin favored for mild, moderate, or total hypoglycemia) (SOE: Moderate; Metformin favored for severe hypoglycemia)
- 226. 169 Figure 54. Pooledodds ratio of mild or moderate hypoglycemia comparing metformin with sulfonylureas CI = confidence interval; Group 1 = metformin; Group 2 = sulfonylureas; OR = odds ratio; pl = profile likelihood estimate Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The line at the bottom of the graph indicates the 95 percent confidence interval for the profile likelihood pooled estimate.
- 227. 170 Table 58. Studiescomparing metformin with sulfonylureas for hypoglycemia Author, Year Study Design Followup Metformin (Dose in mg*) SU (Dose in mg*) Definition of Hypoglycemia Results † (Metformin Vs SU) Hermann, 1994 134 RCT 24 weeks Metformin (max 3000) Glyburide (max 14) Severe (based on clinical findings or available BG) 8/38 (21.1%) vs 12/34 (35.3%); P = 0.18 Hong, 2013 231 RCT 36 months Metformin (max 1500; mean 1400) Glipizide (max 30; mean 28.3) Severe (required assistance and/or PG < 56 mg/dL [3.1 mmol/L]) 3/156 (1.9%) vs 4/148 (2.7%) Chien, 2007 138 RCT 16 weeks Metformin (max 2000; mean 1910) Glyburide (max 20; mean 19) Mild-moderate (symptomatic or BG < 60 mg/dL) 0/25 (0%) vs 0/23 (0%) Blonde, 2002 131 RCT 16 weeks Metformin (max 2000) Glyburide (fixed at 10) Symptomatic and FSG <= 60 mg/dL 1/153 (1%) vs. 3/164 (2%) Garber, 2003 129 RCT 16 weeks Metformin (max 2000) Glyburide (max 10) Symptoms suggesting hypoglycemia 29/164 (18%) vs. 98/151 (65%) Marre‡ , 2002 132 RCT 16 weeks Metformin (max 2000) Glibenclamid e (max 20) Symptoms or labs 0/104 (0%) vs. 7/103 (7%) Garber, 2002 133 RCT 20 weeks Metformin (max 2000) Glyburide (max 10) Symptomatic and BG < 60 mg/dL 0/159 (0%) vs. 10/160 (6%) DeFronzo ‡ , 1995 137 RCT 29 weeks Metformin (max 2500) Glyburide (max 20) Not reported 4/210 (2%) vs. 6/209 (3%) Kahn, 2006 50 RCT 4 years Metformin (max 2000) Glyburide (max 15) Total (self-reported) 168/1454 (11.6%) vs 557/1441 (38.7%) Wright, 2006 258 RCT 6 years Metformin (max 2550) Glyburide (max 20) Mild to severe (not just transient symptoms) Mean annual percentage 0.3% among 290 patients vs 1.2% among 1418 patients Yamanouchi‡ , 2005 60 RCT 12 weeks Metformin (fixed at 750) Glimepiride (max 2) Not reported 0/39 (0%) vs. 1/37 (3%) Charpentier‡ , 2001 136 RCT 20 weeks Metformin (fixed at 2550) Glimepiride (max 6) Symptomatic 8/75 (11%) vs. 17/150 (11%) Derosa, 2004 257 RCT 48 weeks Metformin (max 3000) Glimepiride (max 4) Mild-moderate (not specified) 0/75 (0%) vs 0/73 (0%) Yoon‡ , 2011 74 RCT 48 weeks Metformin (mean 1234.2; max 2000) Glimepiride (mean 4.5; max: 8) Symptomatic 4/114 (4%) vs. 23/118 (19%) Weir, 2011 259 Retrospective cohort 3 months Metformin (NR) Glyburide (NR) Total (presented to an emergency room or hospital with an admission diagnosis of hypoglycemia) Among patients with normal renal function 27/572 (4.7%) vs 53/193 (27.5%); aOR = 9.0 (95% CI, 4.9 to 16.4) Among patients with impaired renal function 29/580 (5.0%) vs 109/444 (24.5%); aOR = 6.0 (95% CI, 3.8 to 9.5) aOR = adjusted odds ratio; BG = blood glucose; CI = confidence interval; max = maximum; mg = milligrams; mg/dL = milligrams per deciliter; mmol/L = millimoles per liter; PG = plasma glucose; RCT = randomized controlled trial; SU = sulfonylurea * All doses were titrated, unless otherwise stated.
- 228. 171 † Results arepresented as n/N (%) unless otherwise stated. ‡ Results were included in the meta-analysis. Metformin Versus DPP-4 Inhibitors Six RCTs (reported in seven publications) compared metformin with DPP-4 inhibitors and reported on hypoglycemia.73, 82-87 Meta-analysis of the short-term, sufficiently-homogeneous RCTs favored DPP-4 inhibitors over metformin for symptomatic hypoglycemia (pooled OR, 0.52; 95% CI, 0.30 to 0.90) (Figure 55).73, 82, 83 Consistent with these findings, longer-term followup from two studies also revealed less hypoglycemia in the DPP-4 inhibitor arms compared with the metformin arms.85, 87 Of note, differences in hypoglycemia rates across the arms were not as clear when the definition of hypoglycemia required biochemical confirmation.73, 84 Rates of severe hypoglycemia were low in the studies reporting on this outcome. Of four short RCTs (24 to 26 weeks), two reported no severe hypoglycemia in either arm.73, 84 One study reported a single event in the metformin arm and none in the DPP-4 inhibitor arm,86 and the other study reported two events in the DPP-4 inhibitor arm and did not report on severe hypoglycemia in the metformin arm.82 Of two RCTs with long-term followup (76 to 104 weeks), one reported no severe hypoglycemia events,87 and the other reported three events of severe hypoglycemia in the metformin monotherapy arms (n=2 for metformin 1000 mg and n=1 for metformin 2000 mg daily) and none in the DPP-4 inhibitor arm.85 Three of the six RCTs did not use an intention-to-treat approach, and withdrawals were high in all three of these studies,83-85 with two excluding data from persons initiating rescue therapy.83, 84 (SOE: Low; DPP-4 inhibitors favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia) Figure 55. Pooled odds ratio of symptomatic hypoglycemia comparing metformin with DPP-4 inhibitors CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = dipeptidyl peptidase-4; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 229. 172 Metformin Versus SGLT-2Inhibitors Four RCTs (reported in three articles) compared metformin with SGLT-2 inhibitors and reported on total hypoglycemia.88, 89, 239 The meta-analysis favored SGLT-2 inhibitors versus metformin for any hypoglycemia, although the combined result was not statistically significant (Figure 56). In a 2013 RCT, Ferrannini et al.,90 an extension of one of the included studies239 with 78 weeks of followup found slightly higher rates of hypoglycemia in the metformin arm (3.6%) versus empagliflozin arms (10 mg, 0.9%; 25 mg, 1.8%); we did not include this study in the meta-analysis because of its longer duration. Two studies reported no events of severe hypoglycemia.88 (SOE: Moderate; SGLT-2 inhibitors favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia) Figure 56. Pooled odds ratio of any hypoglycemia comparing metformin with SGLT-2 inhibitors CI = confidence interval; Group 1 = metformin; Group 2 = sodium-glucose co-transporter-2; OR = odds ratio; SGLT-2 = sodium- glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. Metformin Versus GLP-1 Receptor Agonists Two of three RCTs compared metformin with GLP-1 receptor agonists (duration 26 to 52 weeks) and found a slightly higher risk of mild or moderate hypoglycemia for GLP-1 receptor agonists compared to metformin (Table 59).73, 91, 92 The third RCT reported similar risks across arms for this outcome. No study reported severe hypoglycemia events in either arm. (SOE: Low; Metformin favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia)
- 230. 173 Table 59. Randomizedcontrolled trials comparing metformin with GLP-1 receptor agonists on hypoglycemia Author, Year Followup (Weeks) Metformin (Dose*) GLP-1 Receptor Agonist (Dose*) Definition of Hypoglycemia Results † (Metformin Vs GLP-1 Receptor Agonist) Russell-Jones, 2012 73 26 Metformin (max 2500 mg) Exenatide (fixed at 2.0 mg weekly) Mild-moderate (signs or symptoms associated with BG < 3.0 mmol/L (either self-treated or resolved independently)) Severe ‡ Total (signs or symptoms) 0/246 (0%) vs 5/248 (2.0%) 0/246 (0%) vs 0/248 (0%) 10/246 (4.1%) vs 13/248 (5.2%) Umpierrez, 2014 91 52 Metformin (max 2000 mg or ≥ 1500 mg depending on tolerability) Dulaglutide (fixed at 0.75 mg weekly) Total (signs or symptoms and/or PG ≤ 70 mg/dL [3.9 mmol/L]) Severe (required third party assistance) 34/268 (12.7%) vs 30/270 (11.1%) 0/268 (0%) vs 0/270 (0%) Umpierrez, 2014 91 52 Metformin (max 2000 mg or ≥ 1500 mg depending on tolerability) Dulaglutide (fixed at 1.5 mg weekly) Total (signs or symptoms and/or PG ≤ 70 mg/dL [3.9 mmol/L]) Severe (required third party assistance) 34/268 (12.7%) vs 33/269 (12.3%) 0/268 (0%) vs 0/269 (0%) Yuan, 2012 92 26 Metformin (max 2000 mg) Exenatide (max 2.0 mg) Mild (not specified) Severe (required third party assistance or hospital treatment) 1/26 (3.8%) vs 4/33 (12.1%) 0/26 (0%) vs 0/33 (0%) BG = blood glucose; GLP-1 = glucagon-like peptide-1; max = maximum; mg = milligrams; mg/dL = milligram per deciliter; mmol/L = millimole per liter; PG = plasma glucose * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. ‡ Severe hypoglycemia was defined as symptoms resulting in loss of consciousness or seizure that showed prompt recovery after glucose administration, or documented blood glucose less than 3.0 mmol/L that required the assistance of another person because of severe impairment in consciousness. Thiazolidinediones Versus Sulfonylureas Nine RCTs compared thiazolidinedione with sulfonylurea monotherapy and reported on hypoglycemia (Table 60).50, 60, 74, 94, 95, 100, 103, 217, 253 Results from the meta-analysis of five sufficiently-homogeneous, short-term RCTs showed that the risk of total hypoglycemia was higher for sulfonylurea compared with thiazolidinedione monotherapy (pooled OR for sulfonylurea compared with thiazolidinedione monotherapy, 6.31; 95% CI, 4.08 to 9.76) (Figure 57). We did not find evidence of significant statistical heterogeneity, and removal of any one study did not change the overall inference.
- 231. 174 We did notinclude one short-term (16 weeks) study in the meta-analysis, because it reported on the number of events and not the number of participants experiencing events; this study reported two events of hypoglycemia in the thiazolidinedione arm and three events in the sulfonylurea arm.253 We excluded another short-term (24 weeks) RCT from the meta-analysis, because its mean daily dose of glimepiride (1.5 mg/day) was much lower than the dosing of sulfonylureas in the other studies included in the meta-analysis. This study still found higher rates of hypoglycemia (blood glucose < 60 mg/dL) in the sulfonylurea (7/95, 7.4%) than thiazolidinedione (5/96, 5.2%) arm.103 The longer study, ADOPT, also found higher rates of total hypoglycemia for the sulfonylurea arm (557/1441, 38.7%) compared with the thiazolidinedione arm (142/1456, 9.8%). This study also found more severe hypoglycemia in the sulfonylurea arm (8/1441, 0.6%) compared with the thiazolidinedione arm (1/1456, 0.1%). One of the short-term studies reported that two participants experienced severe hypoglycemia in the sulfonylurea arm but did not report on this outcome for the thiazolidinedione arm.94 (SOE: High; Thiazolidinediones favored for mild, moderate, or total hypoglycemia) (SOE: Moderate; Thiazolidinediones favored for severe hypoglycemia) Figure 57. Pooled odds ratio of any hypoglycemia comparing thiazolidinediones with sulfonylureas CI = confidence interval; Group 1 = thiazolidinediones; Group 2 = sulfonylureas; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 232. 175 Table 60. Randomizedcontrolled trials comparing thiazolidinediones with sulfonylureas on mild to moderate hypoglycemia Author, Year Study Design Followup (Weeks) TZD (Dose*) SU (Dose*) Definition of Hypoglycemia Results † (TZD Vs SU) Jain‡ , 2006 95 24 Pioglitazone (max 45 mg) Glyburide (max 15 mg) 2 or more symptoms or BG < 60 mg/dL 11/251 (4%) vs 61/251 (24%) St. John Sutton ‡ , 2002 217 52 Rosiglitazone (max 8 mg) Glyburide (max 20 mg) Symptomatic 0/104 (0%) vs 7/99 (7%) Tan‡ , 2004 100 52 Pioglitazone (max 45 mg) Glibenclamide (max 10.5 mg) Symptoms or BG < 50 mg/dL 4/91 (4%) vs 32/109 (29%) Hanefeld‡ , 2007 94 52 Rosiglitazone (max 8 mg) Glibenclamide (max 15 mg) Not reported 3/200 (2%) vs 25/207 (12%) Kahn, 2006 50 312 Rosiglitazone (max 8 mg) Glyburide (max 15 mg) Self-reported 142/1456 (10%) vs 557/1441 (39%) Yamanouchi, 2005 60 12 Pioglitazone (30 mg for women and 45 mg for men) Glimepiride (max 2 mg) Not reported 0/38 (0%) vs 1/37 (3%) Agarwal, 2005 253 16 Pioglitazone Glipizide Not reported 2 events among 22 patients vs 3 events among 22 patients Shihara, 2011 103 24 Pioglitazone (mean 23.24 mg; max 30 mg for women and 45 mg for men) Glimepiride (mean 1.51 mg; max 6 mg) BG < 60 mg/dL 5/96 (5%) vs 6/95 (6%) Yoon‡ , 2011 74 48 Rosiglitazone (mean 5.9 mg; max 8 mg) Glimepiride (mean 4.5 mg; max 8 mg) Symptomatic 8/117 (7%) vs 23/118 (19%) BG = blood glucose; mg = milligrams; mg/dL = milligrams per deciliter; SU = sulfonylurea; TZD = thiazolidinedione * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. ‡ Included in meta-analysis. Thiazolidinediones Versus DPP-4 Inhibitors Three short-term studies evaluated hypoglycemia for the comparison of thiazolidinediones to DPP-4 inhibitors (Table 61).48, 73, 104 Of two studies reporting on total hypoglycemia, rates were similar in one study and higher in the pioglitazone versus the sitagliptin arm in the other.48, 73 Two studies reported on severe hypoglycemia and observed no events.73, 104 (SOE: Insufficient for total hypoglycemia; Low: Neither favored for severe hypoglycemia)
- 233. 176 Table 61. Randomizedcontrolled trials comparing thiazolidinediones with DPP-4 inhibitors on hypoglycemia Author, Year Followup (Weeks) TZD (Dose*) DPP-4 Inhibitor (Dose*) Definition of Hypoglycemia Results † (TZD Vs DPP-4 Inhibitor) Alba, 2013 48 12 Pioglitazone (fixed at 30 mg) Sitagliptin (fixed at 100 mg) Total (all reports of hypoglycemia; no glucose measurement required) 2/54 (3.7%) vs 0/52 (0%) Rosenstock, 2010 104 26 Pioglitazone (fixed at 30mg) Alogliptin (fixed at 25 mg) Severe (required third party assistance) 0/163 (0%) vs 0/164 (0%) Russell-Jones, 2012 73 26 Pioglitazone (max 45 mg) Sitagliptin (fixed at 100 mg) Total (signs or symptoms) Severe ‡ 6/163 (3.7%) vs 5/163 (3.1%) 0/163 (0%) vs 0/163 (0%) DPP-4 = dipeptidyl peptidase-4; mg = milligrams; TZD = thiazolidinedione * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. ‡ Severe hypoglycemia was defined as symptoms resulting in loss of consciousness or seizure that showed prompt recovery after glucose administration, or documented blood glucose less than 3.0 mmol/L that required the assistance of another person because of severe impairment in consciousness. Thiazolidinediones Versus GLP-1 Receptor Agonists Two RCTs (26 and 48 weeks in duration) found higher rates of non-severe hypoglycemia for exenatide compared to pioglitazone but no difference in severe hypoglycemia (no events) across arms.73, 105 In the 26-week trial 0/163 (0%) experienced mild hypoglycemia in the pioglitazone arm versus 5/248 (2%) in the exenatide arm.73 In the 48-week trial, corresponding event (symptoms with blood glucose <3.9 mmol/l [<70 mg/dl]) rates were 13/142 (9.2%) and 5/136 (3.7%) in the exenatide and pioglitazone arms, respectively.105 (SOE: Low; Thiazolidinediones favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia) Sulfonylureas Versus DPP-4 Inhibitors Comparisons for both mild and severe hypoglycemia favored the DPP-4 inhibitor arms over sulfonylureas (for mild, moderate, or total hypoglycemia for SU vs. DPP-4 inhibitors: range in OR 3.8 to 12.4; range in RD 6% to 15%). Four RCTs examined hypoglycemia with this comparison; differences in followup length and definitions of hypoglycemia precluded a meta- analysis (Table 62). (SOE: Moderate; DPP-4 inhibitors favored for mild, moderate, or total hypoglycemia) (SOE: Low; DPP-4 inhibitors favored for severe hypoglycemia)
- 234. 177 Table 62. Randomizedcontrolled trials comparing sulfonylureas with DPP-4 inhibitors on hypoglycemia Author, Year Followup (Weeks) SU (Dose*) DPP-4 Inhibitor (Dose*) Definition of Hypoglycemia Results † (SU Vs DPP-4 Inhibitor) Arjona Ferreira, 2013 107 58 Glipizide (max 20 mg; mean 7.7 mg) Sitagliptin (fixed at 50 mg for those with moderate renal insufficiency and 25 mg for those with severe renal insufficiency) Severe (required third party assistance or medical intervention or exhibited markedly depressed level of consciousness, loss of consciousness, or seizure) Total (signs and/or symptoms) 6/212 (2.8%) vs 3/210 (1.4%) 36/212 (17%) vs 13/210 (6.2%); P < 0.001 Gupta, 2013 260 24 Glimepiride (max 4 mg) Sitagliptin (max 200 mg) Total (not specified) 11 episodes among 71 patients vs 3 episodes among 77 patients Barnett, 2012 106 34 Glimepiride (max 4 mg) Linagliptin (fixed at 5 mg) Mild-moderate (Symptoms and/or PG ≤ 70 mg/dL [3.9 mmol/]) Severe (required third party assistance) 5/64 (7.8%) vs 3/137 (2.2%) 0/64 (0%) vs 0/137 (0%) Scott, 2007 108 12 Glipizide (max 20 mg) Sitagliptin (fixed at 25 mg) Total (self-report and glucose measurements) 21/123 (17.1%) vs 5/123 (4.1%) Scott, 2007 108 12 Glipizide (max 20 mg) Sitagliptin (fixed at 50 mg) Total (self-report and glucose measurements) 21/123 (17.1%) vs 5/123 (4.1%) Scott, 2007 108 12 Glipizide (max 20 mg) Sitagliptin (fixed at 100 mg) Total (self-report and glucose measurements) 21/123 (17.1%) vs 2/122 (1.6%) DPP-4 = dipeptidyl peptidase-4; max = maximum; mg = milligrams; mg/dL = milligrams per deciliter; mmol/L = millimole per liter; PG = plasma glucose; SU = sulfonylurea * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. Sulfonylureas Versus GLP-1 Receptor Agonists Sulfonylureas had greater risk of mild to moderate hypoglycemia compared with GLP-1 receptor agonists (range in OR 3.1 to 5.3; range in RD 12% to 21%) (Table 63). Five studies assessed this outcome and could not be pooled because of heterogeneity in outcome definitions and followup length. No study reported any events of severe hypoglycemia. (SOE: Moderate; GLP-1 receptor agonists favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia)
- 235. 178 Table 63. Randomizedcontrolled trials comparing sulfonylureas with GLP-1 receptor agonists on hypoglycemia Author, Year Followup (Weeks) SU (Dose*) GLP-1 Receptor Agonist (Dose*) Definition of Hypoglycemia Results † (SU Vs GLP-1 Receptor Agonist) Garber, 2011 113 104 Glimepiride (max 8mg) Liraglutide (max 1.2 mg) Mild-moderate (did not require assistance, BG < 56 mg/dL [3.1 mmol/L]) Severe (required third- party assistance) 64/248 (25.8%) vs 30/251 (12%) 0/248 (0%) vs 0/251 (0%) Garber, 2011 113 104 Glimepiride (max 8 mg) Liraglutide (max 1.8 mg) Mild-moderate (did not require assistance, BG < 56 mg/dL [3.1 mmol/L]) Severe (required third- party assistance) 64/248 (25.8%) vs 25/247 (10.1%) 0/248 (0%) vs 1/247 (0.4%) ‡ Kaku, 2011 110 52 Glibenclamid e (fixed at 1.25 -2.5 mg) Liraglutide (max 0.9 mg) Mild-moderate (self- treated) Severe (required third party assistance) 1.10 events per patient- year vs 0.19 events per patient-year 0/132 (0%) vs 0/268 (0%) Madsbad, 2004 111 12 Glimepiride (max 4 mg) Liraglutide (Fixed (0.60 mg)) Mild-moderate (BG < 2.8 mmol/L) 4/26 (15.4%) vs 1/30 (3.3%) Madsbad, 2004 111 12 Glimepiride (max 4 mg) Liraglutide (Fixed (0.75 mg)) Mild-moderate (BG < 2.8 mmol/L) 4/26 (15.4%) vs 0/28 (0%) Seino, 2010 109 24 Glibenclamid e (max 2.5 mg) Liraglutide (max 0.9 mg) Mild-moderate (symptoms) Mild-moderate (symptoms and BG < 3.1 mmol/L) Severe (required third party assistance) 45/132 (34.1%) vs 36/268 (13.4%) 29/132 (22%) vs 22/268 (8.2%) 0/132 (0%) vs 0/268 (0%) BG = blood glucose; GLP-1 = glucagon-like peptide-1; max = maximum; mg = milligrams; mg/dL = milligrams per deciliter; mmol/L = millimole per liter; SU = sulfonylurea * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. ‡ Event in the context of insulin infusion as part of a “sub-study” procedure. DPP-4 Inhibitors Versus SGLT-2 Inhibitors One study assessed hypoglycemia for the comparison of DPP-4 inhibitors versus SGLT-2 inhibitors.114 The study compared sitagliptin with empagliflozin at 24 weeks, with one of 223 patients (<1%) in the sitagliptin arm with any hypoglycemia, one of 224 patients (<1%) in the 10 mg empagliflozin arm, and one of 223 patients (<1%) in the 25 mg empagliflozin arm. No
- 236. 179 patients experienced severehypoglycemia. (SOE: Low; Neither favored for mild, moderate, or total hypoglycemia) DPP-4 Inhibitors Versus GLP-1 Receptor Agonists One 26-week RCT assessed hypoglycemia for the comparison of DPP-4 inhibitors versus GLP-1 receptor agonists.73 Investigators compared sitagliptin with exenatide at 26 weeks. Total hypoglycemia was slightly higher for the GLP-1 receptor agonist arm: 13/248 (5.2%) patients in the exenatide arm vs. 5/163 (3.1%) in the sitagliptin arm. Mild hypoglycemia was also higher in the exenatide (5/248, 2.0%) than the sitagliptin arm (0/163, 0%). No patients had severe hypoglycemia in either arm. (SOE: Low; DPP-4 inhibitors favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a Thiazolidinedione More patients experienced hypoglycemia in the combination arm than in the metformin- alone arm. We combined eight sufficiently-homogeneous, short-term RCTs and found an increased odds of mild or moderate hypoglycemia for metformin plus thiazolidinedione versus metformin alone (pooled OR, 1.56; 95% CI, 0.99 to 2.44) (Figure 58).59, 117-120, 122, 123, 247 We did not find statistical heterogeneity. We excluded one RCT from this meta-analysis because its longer followup. This study compared metformin (titrated to a maximum of 2000 mg daily) with metformin plus rosiglitazone (titrated to a maximum of 8 mg/2000 mg daily) at 80 weeks and reported 10 total hypoglycemia events in the metformin-alone arm (3% of patients), and 20 hypoglycemia events in the metformin-rosiglitazone arm (6% of patients).127 These results were consistent with our findings reported above. Another study was excluded from the meta-analysis because of its low- dose combination arm. The study compared metformin (fixed at 1700 mg daily) with a lower- dose combination arm [metformin (fixed at 500 mg daily) and rosiglitazone (fixed at 4 mg daily)] and reported no hypoglycemia in either arm at 24 weeks.124 (SOE: High, Metformin favored for mild, moderate, or total hypoglycemia)
- 237. 180 Figure 58. Pooledodds ratio of any hypoglycemia comparing metformin with combination of metformin plus a thiazolidinedione CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus thiazolidinedione; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus a Combination of Metformin Plus a Sulfonylurea Ten RCTs compared metformin with the combination of metformin plus a sulfonylurea and found more mild, moderate, and total hypoglycemia in the combination arms compared with monotherapy arms.128, 129, 131, 132, 136-141 We did not pool these studies because of differences in definitions of hypoglycemia and dosing of medications; individual study characteristics are shown in Figure 59 and Table 64. Rates of hypoglycemia did not appear higher with glyburide than with any other sulfonylurea. Only two studies reported on severe hypoglycemia and did not report any events.140, 141 (SOE: Moderate; Metformin favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia)
- 238. 181 Figure 59. Oddsratios for studies evaluating mild or moderate hypoglycemia comparing metformin with combination of metformin plus a sulfonylurea CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus sulfonylurea; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study. The width of the horizontal lines represents the 95 percent confidence intervals for each study.
- 239. 182 Table 64. Additionalrandomized controlled trials comparing metformin with a combination of metformin plus a sulfonylurea on hypoglycemia Author, Year Followup (Weeks) Metformin (Dose in mg*) Metformin + SU (Dose*) Definition of Hypoglycemia Results † (Metformin Vs SU) Ahren, 2014 141 104 Metformin (fixed at ≥ 1500) Metformin (fixed at ≥ 1500) + glimepiride (max 4) Mild-moderate (Asymptomatic, but BG ≤ 3.9 mmol/L) Mild-moderate (Symptomatic and BG ≤ 3.9 mmol/L) Severe (required third party assistance) 1/101 (1.0%) vs 3/307 (1.0%) 4/101 (4.0%) vs 55/307 (17.9%) 0/101 (0%) vs 0/307 (0%) Kim, 2014 140 26 Metformin (max 2500) Metformin (max 2000) + glimepiride (fixed at 1-8) Total (symptomatic) Severe (not specified) 4/108 (3.7%) vs 39/100 (0.4%) 0/108 (0%) vs 0/100 (0%) Forst, 2010 139 12 Metformin (fixed) Metformin (fixed) + glimepiride (max 3) Total (not specified) 0/71 (0%) vs 3/65 (4.6%) DeFronzo, 1995 137 29 Metformin (max 2500) Metformin (max 2500) + glyburide (max 20) Not reported 4/210 (2%) vs 38/213 (18%) Charpentier, 2001 136 20 Metformin (fixed at 2550) Metformin (fixed at 2550) + glimepiride (max 6) Symptomatic 8/75 (11%) vs 30/147 (20%) Marre, 2002 132 16 Metformin (max 2000) Metformin (max 2000) + glibenclamide (max 10) Symptoms or labs 0/104 (0%) vs 12/103 (12%) Blonde, 2002 131 16 Metformin (max 2000) Metformin (max 2000) + glyburide (max 20) FSG<=60mg/dl + symptomatic 1/153 (1%) vs 22/162 (14%) Garber, 2003 129 16 Metformin (max 2000) Metformin (max 1000 mg) + glyburide (max 5) Symptoms suggesting hypoglycemia 29/164 (18%) vs 59/171 (35%) Feinglos, 2005 128 16 Metformin (fixed at ≥ 1000) Metformin (fixed at ≥ 1000 mg) + glipizide (fixed at 2.5) FSG <60 mg/dl with symptoms or FSG <50 mg/dl without symptoms or FBG<55 mg/dl without symptoms 2/56 (4%) vs 9/56 (16%) Chien, 2007 138 16 Metformin (max 2000) Metformin (max 2000 mg) + glyburide (max 20) Symptomatic or BG < 60 mg/dL 0/25 (0%) vs. 0/26 (0%) BG = blood glucose; FBG = fasting blood glucose; FSG = fingerstick glucose; max = maximum; mg = milligrams; mmol/L = millimole per liter; SU = sulfonylurea * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated.
- 240. 183 Metformin Versus aCombination of Metformin Plus a DPP-4 Inhibitor We included 27 studies (31 publications) for the comparison of metformin and combination of metformin plus a DPP-4 inhibitor for hypoglycemia.51, 80, 81, 83-87, 118, 139, 141-156, 159-162, 164, 261 Six RCTs were from the 2010 report,80, 81, 83, 142-144, 261 with two RCTs being published extensions of those prior studies.85, 87 We identified 21 new studies for this report. Overall, mild, moderate, total, or severe hypoglycemia were similar for metformin versus the combination of metformin plus a DPP-4 inhibitor. Mild or Moderate Hypoglycemia One long-term RCT reported mild hypoglycemia in 1% of patients in the metformin arm compared with 1.3% of patients in the metformin plus sitagliptin arm at 104 weeks.141 Fourteen short-term studies reported on mild or moderate hypoglycemia (Figure 60).80, 83, 84, 86, 118, 142-144, 147, 148, 151, 152, 155, 261 The pooled odds ratio for mild or moderate hypoglycemia across these studies suggested similar risk of hypoglycemia for the combination of metformin plus a DPP-4 inhibitor compared to metformin (pooled OR, 0.97; 95% CI, 0.63 to 1.51). We did not identify significant statistical heterogeneity, and removal of any single study did not affect the overall inference of no difference in hypoglycemia risk across treatments. Another short-term RCT reported mild hypoglycemia events per person year and found a higher rate of events (4.8 events per person-year) in the metformin arm compared with the metformin plus sitagliptin arm (0.1 events per person-year) at 26 weeks.159 Figure 60. Pooled odds ratio of mild or moderate hypoglycemia comparing metformin with combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 241. 184 Severe Hypoglycemia Thirteen RCTsreported on severe hypoglycemia for this comparison.51, 86, 144, 145, 148, 151, 152, 155, 156, 160, 261 Few events of severe hypoglycemia were reported with 10 of 13 studies reporting no events (Figure 61). Figure 61. Pooled odds ratio of severe hypoglycemia comparing metformin with combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Total Hypoglycemia Eleven studies reported on total hypoglycemia events with short term (<52 week) followup.139, 145, 147-150, 153, 156, 160-162, 164 The pooled odds ratio for metformin vs. metformin plus a DPP-4 inhibitor was 0.96 (95% CI, 0.55, 1.67), suggesting similar risk of total hypoglycemia for metformin and the combination of metformin plus a DPP-4 inhibitor (Figure 62). We did not find evidence of significant statistical heterogeneity, and the exclusion of a single study did not change the inference. Three studies shown in Figure 62 did not contribute to the pooled OR because no events occurred in either arm.139, 149, 150 One study had longer follow up (76 weeks) and reported 20 events in the metformin arm (6.1%) compared with 15 in the metformin plus saxagliptin arm (4.7%) at 76 weeks.87 (SOE: High; Neither favored for mild, moderate, or total hypoglycemia) (SOE: Moderate; Neither favored for severe hypoglycemia)
- 242. 185 Figure 62. Pooledodds ratio of any hypoglycemia comparing metformin with combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Seven short-term studies (published in six articles) reported on total hypoglycemia.88, 153, 156, 165, 166, 168 Meta-analysis of these studies demonstrated a weighted pooled odds ratio for metformin plus an SGLT-2 inhibitor vs. metformin of 1.74 (95% CI, 0.83, 3.66), suggesting a possible increased risk of total hypoglycemia for the combination treatment (Figure 63). Another study reported on total hypoglycemia with 78 weeks of followup and reported higher rates of total hypoglycemia in the metformin compared to metformin plus SGLT-2 inhibitor arm: two patients of 56 in the metformin arm (3.6%), three patients of 166 (1.8%) in the arm receiving 10 mg of empagliflozin, and four patients of 166 (2.4%) in the arm receiving 25 mg of empagliflozin.90 Mean metformin dose was not reported in this study, and the article states that participants were on their pre-enrollment dose of metformin (1500 mg or greater or maximum tolerated dose) during the study.90 One long-term RCT (102 weeks) study reported on mild hypoglycemia events and reported four events in both the metformin arm (4.4%) and the metformin plus dapagliflozin combination arm (4.4%).169 Six short-term studies (published in five articles) reported on severe hypoglycemia events with followup of less than 1 year (range 12 to 24 weeks), and none of these studies reported any severe events.88, 156, 165, 166, 168 One study with more 102 weeks of followup reported no severe hypoglycemic events.170 (SOE: Moderate; Metformin favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia)
- 243. 186 Figure 63. Pooledodds ratio of any hypoglycemia comparing metformin with combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Four studies compared metformin with a combination of metformin plus a GLP-1 receptor agonist,141, 159, 171, 174 and reported on hypoglycemia (Table 65). The single long-term (104-week) study showed no severe hypoglycemia in either arm and did not find a clinically-significant difference in mild-moderate hypoglycemia between arms; this study had large losses to followup and did not use an intention-to-treat approach to analysis.141 Two of three shorter studies (26 to 48 weeks) showed no difference in non-severe hypoglycemia for the metformin and combination arms, but definitions of hypoglycemia varied across these three studies.159, 171, 174 The study suggesting an increased risk of hypoglycemia in the metformin plus GLP-1 receptor agonist arms versus metformin had large losses to followup across its arms. No severe hypoglycemia events were reported in this study, which was the only short-term study reporting on this outcome.159 (SOE: Low; Neither favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia)
- 244. 187 Table 65. Randomizedcontrolled trials comparing metformin with a combination of metformin plus a GLP-1 receptor agonist on hypoglycemia Author, Year Followup (Weeks) Metformin (Dose*) Metformin + GLP-1 Receptor Agonist (Dose*) Definition of Hypoglycemia Results † (Metformin Vs Metformin + GLP-1 Receptor Agonist) Ahren, 2014 141 104 Metformin (fixed at ≥ 1500 mg) Metformin (fixed at ≥ 1500 mg) + albiglutide (max 50 mg weekly) Mild-moderate (Asymptomatic, but PG ≤ 3.9 mmol/L) Mild-moderate (Symptomatic and PG ≤ 3.9 mmol/L) Severe (required third party assistance) 1/101 (1%) vs 4/302 (1.3%) 4/101 (4.0%) vs 9/302 (3.0%) 0/101 (0%) vs 0/302 (0%) Nauck, 2014 159 26 Metformin (fixed at ≥ 1500 mg) Metformin (fixed at ≥ 1500 mg) + dulaglutide (fixed at 0.75 mg/week) Mild-moderate (Signs and symptoms and/or PG ≤ 70 mg/dL [3.9 mmol/L]) Severe (required third party assistance) 2/177 (1.1%) ‡ vs 16/302 (5.3%) 0/177 (0%) ‡ vs 0/302 (0%) Nauck, 2014 159 26 Metformin (fixed at ≥ 1500 mg) Metformin (fixed at ≥ 1500 mg) + dulaglutide (fixed at 1.5 mg/week) Mild-moderate (Signs and symptoms and/or PG ≤ 70 mg/dL [3.9 mmol/L]) Severe (required third party assistance) 2/177 (1.1%) ‡ vs 31/304 (10.2%) 0/177 (0%) ‡ vs 0/304 (0%) Derosa, 2013 171 48 Metformin (mean 2500 mg) Metformin (mean 2500 mg) + exenatide (max 20 mcg) Total (FPG < 60 mg/dL) 0/85 (0%) vs 0/86 (0%) DeFronzo, 2005 174 30 Metformin (fixed at ≥ 1500 mg) Metformin (fixed at ≥ 1500 mg) + exenatide 10 mcg Symptoms (with or without PG <60 mg/dl [3.3 mmol/l]) 6/113 (5.3%) vs. 5/110 (4.5%) DeFronzo, 2005 174 30 Metformin (fixed at ≥ 1500 mg) Metformin (fixed at ≥ 1500 mg) + exenatide 20 mcg Symptoms (with or without PG <3.3 mmol/l) 6/113 (5.3%) vs. 6/110 (5.3%) FPG = fasting plasma glucose; GLP-1 = glucagon-like peptide-1; mcg = micrograms; mg = milligrams; mg/dL = milligrams per deciliter; mmol/L = millimole per liter; PG = plasma glucose; vs = versus * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. ‡ Although study had 52 weeks of followup, incidence of hypoglycemia was only available at 26 weeks. In the trial, patients were switched from a combination of metformin and placebo to a combination of metformin and sitagliptin.
- 245. 188 Metformin-Based Combination Comparisons Combinationof Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea Six short-duration RCTs reporting on mild or moderate hypoglycemia compared the combination of metformin plus a thiazolidinedione with metformin plus a sulfonylurea, showing higher risk of hypoglycemia in the metformin plus sulfonylurea arm (pooled OR, 7.5; 95% CI, 4.0 to 13.8) (Figure 64).175, 177, 178, 180, 183, 185 The trial by Hamann et al. was designed so that patients were withdrawn from the study if they did not reach an efficacy target after 8 weeks of treatment.175 The rates of hypoglycemia were high as medications were titrated up to efficacy, although the relative odds of hypoglycemia in the two arms were comparable to the other studies. No single study strongly influenced the results of the meta-analysis, and no substantial heterogeneity was identified. One study reported on severe hypoglycemia, showing results consistent with the mild to moderate hypoglycemia outcome.180 In Garber et al., seven of 159 patients had severe hypoglycemic events in the metformin plus sulfonylurea arm, and none did in the metformin plus thiazolidinedione arm.180 This study included patients with high baseline HbA1c and had a higher proportion of Asian patients than most studies (12% Asian). (SOE: High; Combination of metformin plus a thiazolidinedione favored for mild, moderate, or total hypoglycemia) (SOE: Low; Combination of metformin plus a thiazolidinedione favored for severe hypoglycemia) Figure 64. Pooled odds ratio of any hypoglycemia comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a sulfonylurea CI = confidence interval; Group 1 = combination of metformin plus thiazolidinedione; Group 2 = combination of metformin plus a sulfonylurea; Met = metformin; OR = odds ratio; SU = sulfonylurea; TZD = thiazolidinedione Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 246. 189 Combination of MetforminPlus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor Three studies (two RCTs and one observational study) compared the combination of metformin plus a thiazolidinedione with metformin plus a DPP-4 inhibitor, showing no clear differences between-groups in hypoglycemia risk.186, 188, 262 One low-quality study randomized 56 patients to metformin and rosiglitazone and 56 patients to metformin and sitagliptin. One patient in the rosiglitazone group withdrew for hypoglycemia, but it is not clearly reported how many in each group experienced hypoglycemia.186 One study compared mild hypoglycemia in a metformin plus pioglitazone arm with a metformin plus sitagliptin arm at 26 weeks.188 There was one patient with an event in the metformin plus pioglitazone arm and five patients with events in the metformin plus sitagliptin arm. This study also evaluated severe hypoglycemia, finding no events in either arm. A prospective cohort study also assessed severe hypoglycemia; no patients with these events were recorded.262 (SOE: Low; Neither favored for mild, moderate, or total hypoglycemia) (SOE: Moderate; Neither favored for severe hypoglycemia) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two short RCTs compared metformin plus a thiazolidinedione with metformin plus a GLP-1 receptor agonist, showing few differences between-group in hypoglycemia risk. The first 20- week study randomized 45 patients to metformin and rosiglitazone and 45 patients to metformin and exenatide, at comparable doses.189 No patients receiving metformin plus rosiglitazone reported hypoglycemia, and two patients receiving metformin plus exenatide reported hypoglycemia, although this difference was not statistically significant.189 There were no severe hypoglycemic events in this study. The second 26-week study randomized 325 patients to either metformin and pioglitazone or metformin and exenatide, at comparable doses.188 The study reported one patient with mild hypoglycemia in the metformin plus pioglitazone arm and two patients with mild hypoglycemia in the metformin plus exenatide arm. The study also reported that no patients had severe hypoglycemia in either arm. (SOE: Low; Neither favored for mild, moderate, total, or severe hypoglycemia) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor Eleven studies of the comparison of a combination of metformin plus a sulfonylurea versus a combination of metformin plus a DPP-4 inhibitor found more patients with severe and non- severe hypoglycemia in the metformin plus sulfonylurea arms compared with the metformin plus DPP-4 inhibitor arms (Figure 65).139, 141, 190-197, 263 Five studies, each lasting 2 years, reported on the outcome of severe hypoglycemia, favoring metformin plus a DPP-4 inhibitor over metformin plus a sulfonylurea (pooled OR, 0.09; 95% CI, 0.03 to 0.26).141, 194-197 Similarly, three studies, each lasting less than 1 year, reported on the outcome of severe hypoglycemia, favoring metformin plus a DPP-4 inhibitor over metformin plus a sulfonylurea (pooled OR, 0.16; 95% CI, 0.05 to 0.56).190-192 The range in RD for severe hypoglycemia in the shorter and longer studies was 0% to 3%. Four additional studies were pooled for a meta-analysis of hypoglycemia (defined as mild, moderate, or total) with followup of 12 to 52 weeks. The pooled odds ratio was 0.19 (95% CI, 0.10 to 0.36), favoring metformin plus a DPP-4 inhibitor.139, 190, 193, 263 Four longer studies, each with followup of 2 years, reported on mild, moderate, or total hypoglycemia, with all four studies
- 247. 190 reporting significantly lesshypoglycemia in the metformin plus DPP-4 inhibitor arms compared with the metformin plus sulfonylurea arms (pooled OR, 0.07; 95% CI, 0.04 to 0.14).141, 194, 195, 197 For all meta-analyses, no single study markedly influenced the results. Only one meta-analysis had substantial heterogeneity, yet the point estimates were fairly similar among these studies. (SOE: High; Combination of metformin plus a DPP-4 inhibitor favored for mild, moderate, total, or severe hypoglycemia in shorter studies (<1 year) and longer studies (lasting 2 years) Figure 65. Pooled odds ratio of hypoglycemia comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor, stratified by study duration and severity of hypoglycemia CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = combination of metformin plus a sulfonylurea; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; Met = metformin; OR = odds ratio; SU - sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. * The profile likelihood estimate provided a similar result. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor For the combined outcome of mild or total hypoglycemia, and for severe hypoglycemia, the comparison of a combination of metformin plus a sulfonylurea versus a combination of metformin plus a SGLT-2 inhibitor favored the metformin plus SGLT-2 inhibitor combinations over the combination of metformin plus a sulfonylurea. Three 2-year studies were pooled, assessing mild or total hypoglycemia for this comparison.199-201 The weighted odds ratio was 0.08 (95% CI, 0.03 to 0.17), favoring metformin plus SGLT-2 inhibitor combinations. There was substantial heterogeneity for this meta-analysis (I-squared, 83%); however, point estimates were fairly similar among the trials (Figure 66). Two
- 248. 191 of the threestudies used equipotent drug dosing between the treatment arms.199, 201 One study mildly underdosed the metformin plus sulfonylurea arm (mean glimepiride dose of 2.7 mg) in comparison with the SGLT-2 inhibitor arm.200 No single study strongly influenced the meta- analysis results. Two of these studies reported hypoglycemia in shorter (52-week) and longer (208-week) studies with similar findings favoring the metformin plus SGLT-2 inhibitor comparison.54, 198 Two of these trials also assessed severe hypoglycemia for this comparison (range in RD 1% to 3%).198, 199 As above, the combination of metformin plus a SGLT-2 inhibitor was favored in both studies. The first RCT assessed severe hypoglycemia at 52 weeks in 965 randomized patients.198 There were 15 patients with a severe hypoglycemic event in the metformin plus sulfonylurea arm and two patients in the metformin plus SGLT-2 inhibitor arm. The second RCT assessed severe hypoglycemia at 104 weeks in 814 randomized patients.199 Severe hypoglycemic events were reported in 7 percent of patients in the metformin plus sulfonylurea arm and 1.7 percent of patients in the metformin plus SGLT-2 inhibitor arm. In two extension studies, lasting 2 and 4 years, the findings were similar, favoring the combination of metformin plus a SGLT-2 inhibitor.54, 201 (SOE: High; Combination of metformin plus a SGLT-2 inhibitor favored for mild, moderate, or total hypoglycemia) (SOE: Moderate; Combination of metformin plus a SGLT-2 inhibitor favored for severe hypoglycemia) Figure 66. Pooled odds ratio of mild or moderate hypoglycemia comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor* CI = confidence interval; Group 1 = combination of metformin plus a sulfonylurea; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; Met = metformin; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. * The profile likelihood estimate provided a similar result.
- 249. 192 Combination of MetforminPlus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three RCTs compared the combination of metformin plus a sulfonylurea with the combination of metformin plus a GLP-1 receptor agonist, showing a lower risk of total/mild/moderate hypoglycemia with the combination of metformin plus a GLP-1 receptor agonist (range in RD -15% to -30%) and no clear between-group differences in severe hypoglycemia risk (Figure 67).53, 141, 204 No meta-analysis could be conducted for this comparison because of differences in study duration and hypoglycemia definitions. In all studies, glimepiride was the sulfonylurea given in combination with metformin. (SOE: Moderate; Combination of metformin plus a GLP-1 receptor agonist favored for mild, moderate, or total hypoglycemia) (SOE: Low, Neither favored for severe hypoglycemia) Figure 67. Odds ratio of hypoglycemia comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a GLP-1 receptor agonist, stratified by study duration and severity of hypoglycemia CI = confidence interval; GLP-1 = glucagon-like peptide-1; Group 1 = combination of metformin plus a sulfonylurea; Group 2 = combination of metformin plus a glucagon-like peptide-1 agonist; Met = metformin; OR = odds ratio; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Basal Insulin One study addressed the comparison of a combination of metformin plus a sulfonylurea versus a combination of metformin plus a basal insulin, favoring metformin and insulin glargine. This 48-week RCT compared metformin and glimepiride with metformin and insulin glargine.206 While patients continued on the same pre-study dose of metformin of around 1500 mg, both the sulfonylurea and the insulin glargine were titrated to reach blood sugar targets. Nineteen patients
- 250. 193 of 30 (63%)in the metformin plus glimepiride arm had mild hypoglycemia compared with 10 of 34 (29%) in the metformin plus insulin glargine arm. No severe hypoglycemia events occurred in either treatment arm. (SOE: Low; Combination of metformin plus basal insulin favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither arm favored for severe hypoglycemia) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a Premixed Insulin Two RCTs compared metformin plus sulfonylurea with metformin plus a premixed insulin, showing no clear between-group differences in hypoglycemia risk (Table 66).207, 208 (SOE: Low; Neither favored for mild, moderate, or total hypoglycemia) (SOE: Insufficient for severe hypoglycemia) Table 66. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a premixed insulin on hypoglycemia Author, Year Followup Comparison Outcome Results Malone, 2003 207 16 weeks Metformin + glargine versus metformin + lispro 75/25 Nocturnal (N = 597 in trial) Greater number of participants with nocturnal hypoglycemia (p < 0.01) with metformin plus sulfonylurea than metformin plus insulin Severe Comparable number with severe hypoglycemia (p=0.10) Kvapil, 2006 208 16 weeks Metformin + glibenclamide versus metformin + aspart 70/30 Mild or moderate 9/114 versus 13/108; RR = 1.5 (95% CI, 0.7 to 3.4) CI = confidence interval; RR = relative risk Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Five studies considered hypoglycemia for the comparison of a combination of metformin plus a DPP-4 inhibitor versus a combination of metformin plus a SGLT-2 inhibitor, showing no clear between-group differences in hypoglycemia risk (Figure 68). Four RCTs, with equipotent doses in each arm, showed no significant between-group differences in severe hypoglycemia risk.156, 158, 209 One 52-week RCT158 randomized 714 patients to metformin plus sitagliptin or metformin plus canagliflozin, finding 4.1 percent of patients in the metformin plus sitagliptin arm with any hypoglycemic event compared with 6.8 percent of patients in the metformin plus canagliflozin arm. A 78-week, lower-quality RCT90 found 3.6 percent of patients with any hypoglycemic event in the metformin plus sitagliptin arm, 1.8 percent of patients with such events in the low-dose metformin plus empagliflozin arm, and 2.4 percent of patients in the high-dose metformin plus empagliflozin arm. Two shorter RCTs assessed total hypoglycemia with followup at 12 weeks, showing a non- significant greater risk of total hypoglycemia in the metformin plus DPP-4 inhibitor arms.153, 156 (SOE: Low; Neither arm favored for mild, moderate, total, and severe hypoglycemia)
- 251. 194 Figure 68. Oddsratio of hypoglycemia comparing a combination of metformin plus an SGLT-2 inhibitor with a combination of metformin plus a DPP-4 inhibitor, stratified by severity of hypoglycemia CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; Met = metformin; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three studies reported on hypoglycemia for this comparison, showing no clear between- group differences in hypoglycemia risk (Table 67).141, 159, 188 All three studies compared metformin plus sitagliptin, but each of the studies used a different GLP-1 receptor agonist in the metformin plus GLP-1 receptor agonist comparator arm (albiglutide, exenatide, and dulaglutide). None of the three studies had any severe hypoglycemia in either arm. For mild to moderate hypoglycemia, there were conflicting results with two of the three studies favoring the metformin plus DPP-4 inhibitor arms. This may be due to different types of GLP-1 receptor agonists, although there may be unidentified sources of heterogeneity, too. (SOE: Low; Neither arm favored for mild, moderate, total, and severe hypoglycemia)
- 252. 195 Table 67. Randomizedcontrolled trials comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist on hypoglycemia Author, Year Followup (Weeks) Metformin + DPP-4 Inhibitor (Dose*) Metformin + GLP-1 Receptor Agonist (Dose*) Definition of Hypoglycemia Results † (Metformin + DPP-4 Inhibitor Vs Metformin + GLP-1 Receptor Agonist) Ahren, 2014 141 104 Metformin (fixed at ≥ 1500 mg) + sitagliptin (fixed at 100 mg) Metformin (fixed at ≥ 1500 mg) + albiglutide (max 50 mg weekly) Mild-moderate (asymptomatic, but BG ≤ 3.9 mmol/L) Mild-moderate (symptomatic and BG ≤ 3.9 mmol/L) Severe (required third party assistance) 4/302 (1.3%) vs 4/302 (1.3%) 5/302 (1.7%) vs 9/302 (3%) 0/302 (0%) vs 0/302 (0%) Bergenstal, 2010 188 26 Metformin (fixed, mean 1583 mg) + sitagliptin (fixed at 100 mg) Metformin (fixed, mean 1504 mg) + exenatide (fixed at 2 mg weekly) Mild-moderate (symptomatic and BG < 3 mmol/L) Severe ‡ 5/166 (3%) vs 2/160 (1.3%) 0/166 (0%) vs 0/160 (0%) Nauck, 2014 159 52 Metformin (fixed at ≥ 1500 mg) + sitagliptin (fixed at 100 mg) Metformin (fixed at ≥ 1500 mg) + dulaglutide (fixed 0.75 mg weekly) Mild-moderate (signs, symptoms and/or BG ≤ 70 mg/dL) Severe (required third party assistance) 15/315 (4.8%) vs 16/302 (5.3%) 0/315 (0%) vs 0/302 (0%) Metformin (fixed at ≥ 1500 mg) + sitagliptin (fixed at 100 mg) Metformin (fixed at ≥ 1500 mg) + dulaglutide (fixed 1.5 mg weekly) Mild-moderate (signs, symptoms and/or BG ≤ 70 mg/dL) Severe (required third party assistance) 15/315 (4.8%) vs 31/304 (10.2%) 0/315 (0%) vs 0/304 (0%) BG = blood glucose; DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; max = maximum; mg = milligrams; mg/dL = milligrams per deciliter; mmol/L = millimole per liter * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. ‡ Severe hypoglycemia was defined as a loss of consciousness, seizure, or coma that resolved after treatment with glucagon or glucose, or severe impairment that required third-party assistance to resolve the episode and a blood glucose concentration of lower than 3 mmol/L. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a Basal Insulin One RCT assessed hypoglycemia for the comparison of a combination of metformin plus a DPP-4 inhibitor versus a combination of metformin plus a basal insulin, finding more hypoglycemic events in the metformin plus insulin arm (Table 68).211 (SOE: Low; Combination of metformin plus a DPP-4 inhibitor favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia)
- 253. 196 Table 68. Randomizedcontrolled trials comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a basal insulin on hypoglycemia Author, Year Followup (Weeks) Metformin + DPP-4 Inhibitor (Dose*) Metformin + Basal Insulin (Dose*) Definition of Hypoglycemia Results † (Metformin + DPP-4 Inhibitor Vs Metformin + Basal Insulin) Aschner, 2012 211 24 Metformin (baseline dose 1835 mg) + sitagliptin (fixed at 100 mg) Metformin (baseline dose 1852 mg) + insulin glargine (max 0.5 U/kg) Severe (severe symptomatic) Total (symptomatic and BG ≤ 3.9 mmol/L) 1/264 (0.4%) vs 3/237 (1.3%) 28/264 (10.6%) vs 86/237 (36.3%) BG = blood glucose; DPP-4 = dipeptidyl peptidase-4; max = maximum; mg = milligrams; mmol/L = millimole per liter; U/kg = units per kilogram * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. Combination of Metformin Plus a GLP-1 Receptor Agonist Versus a Combination of Metformin Plus a Basal Insulin Two RCTs compared metformin plus basal insulin with the combination of metformin plus exenatide, with lower risk of mild or moderate hypoglycemia in the metformin plus exenatide arms in both studies (range in RD of -3% to -25%) (Table 69).212, 264 Both RCTs reported no between-group differences in severe hypoglycemia.212, 264 (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia) Table 69. Randomized controlled trials comparing a combination of metformin plus a GLP-1 receptor agonist with a combination of metformin plus a basal insulin on hypoglycemia Author, Year Followup (Weeks) Metformin + GLP-1 (Dose*) Metformin + Basal Insulin (Dose*) Definition of Hypoglycemia Results † (Metformin + GLP-1 Vs Metformin + Basal Insulin) Davies, 2013 264 26 Metformin (fixed at ≥ 1000 mg) + exenatide (fixed at 2 mg weekly) Metformin (fixed at ≥ 1000 mg) + insulin detemir (mean initial dose 0.21 IU/kg; mean end dose 20.8 IU, end dose 0.51 IU/kg) Mild-moderate (symptoms that were self-treated or resolved on their own, with documented BG < 3.0 mmol/L) Severe hypoglycemia ‡ 0/33 (0%) vs 1/29 (3.4%) 0/33 (0%) vs 0/29 (0%) Diamant, 2010 212 84 Metformin (continued stable dose) + exenatide (fixed at 2 mg weekly) Metformin (continued stable dose) + insulin glargine (started at 10 IU then titrate to glycemic goal of 4- 5.5 mmol/L) Mild-moderate (symptoms and BG < 3.0 mmol/L and was either self-treated or resolved independently) Severe hypoglycemia‡ 13/164 (7.9%) vs 51/157 (32.5%) 1/164 (0.6%) vs 1/157 (0.6%) BG = blood glucose; GLP-1 = glucagon-like peptide-1 receptor agonist; IU = international units; IU/kg = international units per kilogram; mg = milligrams; mmol/L = millimole per liter; * All doses were titrated, unless otherwise stated. † Results are presented as n/N (%) unless otherwise stated. ‡ Any hypoglycemic episode with symptoms consistent with hypoglycemia that led to loss of consciousness or seizure, with prompt recovery in response to glucagon or glucose administration, or documented hypoglycemia [blood glucose <3.0mmol] necessitating assistance of another person
- 254. 197 Combination of MetforminPlus a GLP-1 Receptor Agonist Versus a Combination of Metformin Plus a Premixed Insulin One study assessed the comparison of a combination of metformin plus a GLP-1 receptor agonist versus a combination of metformin plus a premixed insulin, showing less hypoglycemia in the metformin plus GLP-1 receptor agonist arm compared with the metformin plus premixed insulin arm. This 26-week RCT found an incidence of first hypoglycemic episodes of 8.0 percent in the metformin plus exenatide group versus 20.5 percent in the metformin plus insulin aspart 70/30 group (RD, -12.5%).213 No severe hypoglycemia was reported. (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither favored for severe hypoglycemia) Combination of Metformin Plus a Basal Insulin Versus a Combination of Metformin Plus a Premixed Insulin Five trials examined the comparison of metformin plus basal insulin to metformin plus a mix of long- and short-acting insulin, consistently favoring the former (range in RD, -5% to -28%) (Table 70).214-216, 223, 224 Due to the heterogeneity of these trials (I-squared, 78.8%), they were not pooled in a meta-analysis. The heterogeneity may be owing to the difference in followup times, insulin preparations, and insulin dosing. (SOE: Moderate; Combination of metformin plus a basal insulin favored for mild, moderate, or total hypoglycemia) (SOE: Low; Neither arm favored for severe hypoglycemia) Table 70. Randomized controlled trials comparing a combination of metformin plus a basal insulin with a combination of metformin plus a premixed insulin on hypoglycemia Author, Year Comparison Outcome Results RR and Comments (Combination Metformin and Another Insulin as Reference Group) Malone, 2004 223 Metformin + glargine versus metformin + lispro 75/25 Mild or moderate at 32 weeks 40/101 versus 57/100 (87 versus 181 events) RR = 0.69 (95% CI 0.5 to 0.9), both arms of cross-over pooled Severe at 32 weeks None NA Malone, 2005 224 Metformin + glargine versus metformin + lispro 75/25 Mild or moderate at 32 weeks 0.44 versus 0.61 events/patient/30 days P = 0.47; more daytime hypoglycemia with lispro 75/25 but less nocturnal hypoglycemia Severe at 32 weeks None NA Raskin, 2007 215 Metformin + glargine versus metformin + aspart 70/30 Mild or moderate at 28 weeks 11/78 versus 33/79 (23 versus 121 events) RR = 0.34 (95% CI 0.2 to 0.6) Robbins, 2007 214 Metformin + glargine versus metformin + lispro 50/50 Mild or moderate at 24 weeks 75/158 versus 79/157 RR = 0.94 (95% CI 0.8 to 1) Severe at 24 weeks 2/158 versus 3/157 RR = 0.66 (95% CI 0.1 to 4) Davies, 2007 216 Metformin + NPH versus metformin + NPH/regular 70/30 Mild or moderate 7/29 versus 8/27 RR = 0.81 (95% CI 0.34 to 1.9); a poorly conducted trial CI = confidence interval; NA = not available; NPH = neutral protamine Hagedorn; RR = relative risk
- 255. 198 Strength of Evidencefor Hypoglycemia As noted in the Key Points, Table 71, Table 72, and Table 73, we found moderate or high strength of evidence for many of the monotherapy comparisons evaluating hypoglycemia and also found a number of combination comparisons with high or moderate strength of evidence. We found several comparisons of interest for which there was no or minimal evidence, especially among the combination comparisons. Study limitations for most comparisons were low or medium, with only two comparisons having high study limitations owing to lack of blinding, lack of description of withdrawals and dropouts, or high losses to followup. In general, we did not find strong differences in outcomes in the lower- versus higher-quality studies. When we found low strength of evidence for hypoglycemia, this tended to occur in the setting of fair to poor study quality and inconsistency for monotherapy comparisons and was related to insufficient data in the combination comparisons. We generally found consistency among studies if there were more than three studies for a given comparison. Most evidence on hypoglycemia was precise for monotherapy comparisons; there was less precision when there were fewer studies, as in the combination comparisons. We did not find any evidence of publication bias using the Begg’s and Egger’s test for the comparisons with greater than ten studies. We identified unpublished studies that could have influenced our rating of the evidence. A single unpublished study found more hypoglycemia in a DPP-4 inhibitor arm than in the comparator metformin arm; this was consistent with our findings and could have strengthened the evidence. Also, we identified two additional studies of the comparison of sulfonylurea to DPP-4 inhibitor monotherapy which were consistent with the published studies; the addition of this evidence may have allowed us to rate the strength of evidence as high for this comparison.
- 256. 199 Table 71. Strengthof evidence domains for monotherapy comparisons in terms of hypoglycemia among adults with type 2 diabetes Comparison* Outcome Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. TZD Mild, moderate, total 5 (4,197) Medium Inconsistent Direct Precise Undetected Low Metformin favored Severe 1 (409) Low Unknown Direct Imprecise Undetected Low Neither favored Metformin vs. SU Mild, moderate, total RCTs: 14 (7,332) Medium Consistent Direct Precise Undetected High Metformin favored; 4.0 (1.8 to 9.8) Observational: 1 (1789) Medium Unknown Direct Precise n/a Severe 2 (376) Low Consistent Direct Imprecise Undetected Moderate Metformin favored; range in OR, 0.49 to 0.71; range in RD, -1% to -23% Metformin vs. DPP-4 inhibitors Mild, moderate, total 6 (6,710) High Consistent for symptomatic hypoglycemia Direct Precise Undetected Low DPP-4 inhibitor favored Severe 6 (6,710) High Inconsistent Direct Imprecise Suspected Low Neither favored Metformin vs. SGLT-2 inhibitors Mild, moderate, total 5 (2,700) Medium Consistent Direct Precise Undetected Moderate SGLT-2 inhibitors favored; 0.5 (0.2 to 1.3) Severe 2 (831) Medium Consistent Direct Imprecise Undetected Low Neither favored Metformin vs. GLP-1 receptor agonists Mild, moderate, total 3 (1360) Low Inconsistent Direct Imprecise Undetected Low Metformin favored Severe 3 (1360) Low Consistent Direct Imprecise Undetected Low Neither favored
- 257. 200 Table 71. Strengthof evidence domains for monotherapy comparisons in terms of hypoglycemia among adults with type 2 diabetes (continued) Comparison* Outcome Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † TZD vs. SU Mild, moderate, total 8 (6,212) Low Consistent Direct Precise Undetected High TZD favored; 6.3 (4.1 to 9.8) Severe 2 (3,304) Low Consistent Direct Precise Undetected Moderate TZD favored; OR, 8.1; RD, 0.4% TZD vs. DPP- 4 inhibitors Mild, moderate, total 3 (1,686) Low Inconsistent Direct Precise Undetected Insufficient Unable to determine Severe 2 (653) Low Consistent Direct Imprecise Undetected Low Neither favored TZD vs. GLP- 1 receptor agonists Mild, moderate, total 2 (689) Medium Consistent Direct Imprecise Undetected Low TZD favored Severe 1 (411) Low Unknown Direct Imprecise Undetected Low Neither favored SU vs. DPP-4 inhibitors Mild, moderate, total 4 (1,065) Medium Consistent Direct Precise Undetected Moderate DPP-4 favored; range in OR, 3.8 to 12.4; range in RD, 6% to 15% Severe 2 (623) Low Inconsistent Direct Imprecise Undetected Low DPP-4 favored SU vs. GLP-1 receptor agonists Mild, moderate, total 5 (2,467) Medium Consistent Direct Precise Undetected Moderate GLP-1 favored for mild- moderate hypoglycemia; range in OR, 3.1 to 5.3; range in RD, 12% to 21% Severe 3 (1546) High Consistent Direct Imprecise Undetected Low Neither favored DPP-4 inhibitors vs. SGLT-2 inhibitors Mild, moderate, total 1 (670) Low Unknown Direct Imprecise Undetected Low Neither favored Severe 1 (670) Low Unknown Direct Imprecise Undetected Low Neither favored
- 258. 201 Table 71. Strengthof evidence domains for monotherapy comparisons in terms of hypoglycemia among adults with type 2 diabetes (continued) Comparison* Outcome Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † DPP-4 inhibitors vs. GLP-1 receptor agonists Mild, moderate, total 1 (411) Low Unknown Direct Imprecise Undetected Low DPP-4 favored Severe 1 (411) Low Unknown Direct Imprecise Undetected Low Neither favored CI = confidence interval; DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; OR = odds ratio; RD = risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) due to the few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 259. 202 Table 72. Strengthof evidence domains for metformin versus metformin-based combination comparisons in terms of hypoglycemia among adults with type 2 diabetes Comparison* Outcome Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + TZD Mild, moderate, total 10 (3,906) Medium Consistent Direct Imprecise Undetected Moderate Metformin favored; 1.6 (0.99 to 2.4) Metformin vs. metformin + SU Mild, moderate, total 12 (3,732) Medium Consistent Direct Precise Undetected Moderate Metformin favored, range in OR, 0.99 to 28.55; range in RD, 0% to 35% Severe 2 (544) Medium Consistent Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + DPP-4 inhibitors Mild, moderate, total 27 (17,946) Low Consistent Direct Precise Undetected High Neither favored; pooled OR for mild- moderate, 0.97 (0.6 to 1.5) Pooled OR for total, 1.0 (0.6 to 1.7 Severe 12 (5,674) Medium Consistent Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + SGLT-2 inhibitors (<2 years) Mild, moderate, total 10 (6,178) Low Consistent Direct Imprecise Undetected Moderate Metformin favored; 1.7; 95% CI, 0.8 to 3.7 Severe 7 (2,934) Low Consistent Direct Imprecise Undetected Moderate Neither favored; no events Metformin vs. metformin + GLP-1 receptor agonists Mild, moderate, total 4 (2,654) High Inconsistent Direct Imprecise Undetected Low Neither favored Severe 2 (1,186) High Consistent Direct Imprecise Undetected Low Neither favored CI = confidence interval; DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; OR = odds ratio; RD = risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 260. 203 Table 73. Strengthof evidence domains for metformin-based combination comparisons in terms of hypoglycemia among adults with type 2 diabetes Comparison* Outcome Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + TZD vs. metformin + SU Mild, moderate, total 7 (975) Medium Consistent Direct Imprecise Undetected High Metformin + TZD favored; 7.5 (4.0 to 13.8) Severe 1 (314) Medium Unknown Direct Imprecise Undetected Low Metformin + TZD favored Metformin + TZD vs. metformin + DPP-4 inhibitors Mild, moderate, total 2 (603) Medium Consistent Direct Imprecise Undetected Low Neither drug combination favored Severe RCT 1 (491) Obs 1 (83) Medium Consistent Direct Imprecise Undetected Low Neither favored Metformin + TZD vs. metformin + GLP-1 receptor agonists Mild, moderate, total 2 (415) Medium Consistent Direct Imprecise Undetected Low Neither drug combination favored Severe 2 (415) Medium Consistent Direct Imprecise Undetected Low Neither drug combination favored Metformin + SU vs. metformin + DPP-4 inhibitors Mild, moderate, total 10 (6,757) Medium Consistent Direct Imprecise Undetected High Metformin + DPP4- inhibitors favored Pooled OR for studies ≤52 weeks: 0.2 (0.1 to 0.4) Pooled OR for studies >52 weeks: 0.1 (0.04 to 0.14) Severe 6 (4,717) Medium Consistent Direct Imprecise Undetected High Metformin + DPP-4 inhibitors favored Pooled OR for studies <52 weeks: 0.2 (0.1 to 0.6) Pooled OR for studies ≥52 weeks: 0.1 (0.03 to 0.3)
- 261. 204 Table 73. Strengthof evidence domains for metformin-based combination comparisons in terms of hypoglycemia among adults with type 2 diabetes (continued) Comparison* Outcome Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + SU vs. metformin + SGLT- 2 inhibitors (< 2 years) Mild, moderate, total 3 (3,815) Low Consistent Direct Precise Undetected High Metformin + SGLT-2 inhibitors favored; 0.1 (0.03 to 0.2) Severe 2 (1,779) Low Consistent Direct Imprecise Undetected Moderate Metformin + SGLT-2 inhibitors favored in studies lasting 1-2 years; range in OR, 0.13 to 0.23; range in RD, -3% to -1% Metformin + SU vs. metformin + GLP-1 receptor agonists Mild, moderate, total 3 (2,557) Medium Consistent Direct Precise Undetected Moderate Metformin + GLP-1 inhibitor favored in studies lasting 16 to 238 weeks; range in OR, 0.07 to 0.29; range in RD, -30% to -15% Severe 3 (2,557) Medium Inconsistent Direct Imprecise Undetected Low Neither favored Metformin + SU vs. metformin + basal insulin Mild, moderate, total 1 (75) Medium Unknown Direct Imprecise Undetected Low Metformin + insulin favored Severe 1 (75) Medium Unknown Direct Imprecise Undetected Low Neither arm favored Metformin + SU vs. metformin + premixed insulin Mild, moderate, total 2 (827) Medium Consistent Direct Imprecise Suspected Low Neither arm favored Severe 1 (597) High Unknown Direct Imprecise Suspected Insufficient Unable to draw a conclusion Metformin + DPP-4 inhibitors vs. metformin + SGLT- 2 inhibitors Mild, moderate, total 4 (2,889) Medium Inconsistent Direct Imprecise Undetected Low Neither arm favored in studies lasting 12 to 78 weeks Severe 2 (1,359) Low Consistent Direct Imprecise Undetected Low Neither arm favored
- 262. 205 Table 73. Strengthof evidence domains for metformin-based combination comparisons in terms of hypoglycemia among adults with type 2 diabetes (continued) Comparison* Outcome Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + DPP-4 inhibitors vs. metformin + GLP-1 receptor agonists Mild, moderate, total 3 (1,851) Medium Inconsistent Direct Precise Undetected Low Neither arm favored in studies lasting 26 to 104 weeks Severe 3 (1,851) Medium Consistent Direct Imprecise Undetected Low Neither arm favored in studies lasting 26 to 104 weeks Metformin + DPP-4 inhibitors vs. metformin + basal insulin Mild, moderate, total 1 (515) Medium Unknown Direct Precise Undetected Low Metformin + DPP-4 inhibitor favored Severe 1 (515) Medium Unknown Direct Imprecise Undetected Low Neither arm favored Metformin + GLP-1 receptor agonists vs. metformin + basal insulin Mild, moderate, total 2 (397) Medium Consistent Direct Imprecise Undetected Low ‡ Metformin + GLP-1 receptor agonist favored in studies lasting 26 to 84 weeks Severe 2 (383) Medium Consistent Direct Imprecise Undetected Low Neither arm favored Metformin + GLP-1 receptor agonists vs. metformin + premixed insulin Mild, moderate, total 1 (363) High N/A Direct Imprecise Undetected Low ‡ Metformin + GLP-1 receptor agonist favored Severe 1 (363) High Unknown Direct Imprecise Undetected Low Neither arm favored Metformin + basal insulin vs. metformin + premixed insulin Mild, moderate, total 5 (530) Medium Consistent Direct Imprecise Undetected Moderate Metformin + basal insulin favored; range in OR, 0.3 to 0.9; range in RD, -28% to -5% Severe 3 (613) Medium Consistent Direct Imprecise Undetected Low Neither arm favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; OR = odds ratio; RD = risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack
- 263. 206 of available evidence.Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence. ‡ If we compare the metformin plus GLP-1 receptor agonists versus metformin plus premixed or basal insulin, then metformin plus GLP-1 receptor agonists have less hypoglycemia over metformin plus either premixed or basal insulin with moderate strength of evidence
- 264. 207 Evidence for GastrointestinalSide Effects Monotherapy Comparisons Metformin Versus Thiazolidinediones Eight RCTs compared GI adverse events between metformin and either pioglitazone or rosiglitazone.50, 59, 62, 64, 67, 70, 73, 74 GI adverse events were more common with metformin compared with thiazolidinediones in the majority of RCTs, except for dyspepsia, where the number of events were comparable for both treatments. More people had diarrhea and nausea with metformin than thiazolidinediones (Figure 69). There were no overly influential studies in the meta-analysis for diarrhea, and there was little heterogeneity between these studies. (SOE: Moderate; Thiazolidinediones favored for diarrhea and nausea) Figure 69. Pooled odds ratio of gastrointestinal adverse events comparing metformin with thiazolidinediones CI = confidence interval; Group 1 = metformin; Group 2 = thiazolidinediones; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus Sulfonylureas Twelve RCTs compared GI adverse events between metformin and a sulfonylurea.50, 74, 129- 131, 133, 134, 136-138, 251, 257 GI adverse events tended to be more common with metformin than with sulfonylurea (Table 74). Based on meta-analyses, there were fewer GI adverse events with sulfonylureas than with metformin for the outcomes of diarrhea (OR, 0.42; 95% CI, 0.24 to 0.72; I-squared, 48.4%; six studies), abdominal pain (OR, 0.44; 95% CI, 0.29 to 0.67; I-squared, 0%; three studies), nausea and vomiting (OR, 0.45; 95% CI, 0.31 to 0.65; I-squared, 0%; three
- 265. 208 studies) and anyGI adverse event (OR, 0.45; 95% CI, 0.28 to 0.72; I-squared, 22.2%; four studies). (SOE: Moderate; Sulfonylureas favored) Table 74. Studies comparing metformin with sulfonylureas on gastrointestinal adverse events Author, Year Outcome Event Rates (Metformin Versus Sulfonylureas) Hermann, 1994 134 Any GI outcome Abdominal pain Diarrhea Nausea Withdrawal for GI symptoms 63% (24/38) versus 32% (11/34) 18% (7/38) versus 6% (2/34) 50% (19/38) versus 0 (0/34) 24% (9/38) versus 9% (3/34) 14% versus 0% DeFronzo, 1995 137 Nausea and diarrhea 1.4% (3/210) versus 1.0% (2/209) Amador-Licona, 2000 251 Diarrhea and abdominal pain 14.3% (4/28) for metformin; event rates are not reported for sulfonylurea Charpentier, 2001 136 Diarrhea 7% (5/75) versus 1% (1/150) Blonde, 2002 131 Nausea and vomiting Dyspepsia/heartburn Flatulence 12.4% (19/153) versus 5.5% (9/164) 4.6% (7/153) versus 3% (5/164) 2% (3/153) versus 0% (0/164) Garber, 2002 133 Any GI outcome Diarrhea Nausea/vomiting Abdominal pain Dyspepsia Metformin (n = 159); glyburide (n = 160) 43% versus 24% 15.1% versus 4.4% 6.3% versus 0.6% 5% versus 3.1% 5% versus 2.5% Garber, 2003 129 Nausea/vomiting Abdominal pain Diarrhea 10.4% (17/164) versus 6.6% (10/151) 6.1% (10/164) versus 4% (6/151) 18% (30/164) versus 5.3% (18/151) Goldstein, 2003 130 Diarrhea 17.3% (13/75) versus 13.1% (11/84) Derosa, 2004 257 Nausea and diarrhea 2.4% (2/75) versus 0% (0/73) Kahn, 2006 50 Combined GI events Nausea Vomiting Diarrhea Abdominal discomfort 38% (557/1454) versus 22% (316/1441) 11.7% (170/1454) versus 6.9% (99/1441) 5.8% (84/1454) versus 3.1% (45/1441) 23.7% (345/1454) versus 9.9% (142/1441) 15.4% (224/1454) versus 11.3% (163/1441) Chien, 2007 138 Combined GI events 32% (8/25) versus 13% (3/23) Yoon, 2011 74 Diarrhea Discomfort, pain, nausea or vomiting 8.8% (10/114) versus 3.4% (4/118) 8.8% (10/114) versus 8.5% (10/118) GI = gastrointestinal Metformin Versus DPP-4 Inhibitors Six RCTs compared metformin with a DPP-4 inhibitor and reported on GI adverse events.73, 82, 84-87 Metformin had more GI adverse events compared with each of the DPP-4 inhibitors (Figure 70). One trial identified solely in ClinicalTrials.gov had results consistent with the published studies (NCT01076088). We combined the three studies with similar study durations and dosages for nausea and diarrhea. We did not combine “any” GI adverse event outcomes, because study durations were not sufficiently similar. Based on meta-analyses, there were fewer
- 266. 209 nausea outcomes forDPP-4 inhibitors compared with metformin (pooled OR, 0.37; 95% CI, 0.15 to 0.91; I-squared, 4%; three studies) and fewer diarrhea outcomes for the same comparison (pooled OR, 0.38, 95% CI, 0.18 to 0.83; I-squared, 25%; three studies). The excluded longer studies were consistent with the findings, favoring DPP-4 inhibitors over metformin (Figure 70). (SOE: High; DPP-4 inhibitors favored) Figure 70. Odds ratio of gastrointestinal adverse events comparing metformin with DPP-4 inhibitors CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = dipeptidyl peptidase-4 inhibitors; OR = odds ratio Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. Metformin Versus SGLT-2 Inhibitors Three trials (published in two articles) compared metformin with dapagliflozin.88, 89 One trial compared metformin with empagliflozin,239 Diarrhea and nausea tended to be more common with metformin than with the SGLT-2 inhibitors (Figure 71). We did not pool the trials owing to dosage differences. (SOE: Low; SGLT-2 inhibitors favored for diarrhea and nausea)
- 267. 210 Figure 71. Oddsratio of gastrointestinal adverse events comparing metformin with SGLT-2 inhibitors CI = confidence interval; Dapa = dapagliflozin; Empa = empagliflozin, Group 1 = metformin; Group 2 = sodium-glucose co- transporter-2 inhibitors; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. Metformin Versus GLP-1 Receptor Agonists Two trials compared metformin with exenatide.73, 92 One trial compared metformin with once-weekly subcutaneously injected 0.75 mg or 1.5 mg of dulaglutide.91 Diarrhea non- significantly differed between metformin and GLP-1 receptor agonists (OR, 0.78; 95% CI, 0.54 to 1.14; I-squared, 0%; three studies). However, nausea (OR, 1.28 and 1.71), vomiting (pooled OR, 1.73; 95% CI, 1.01 to 2.95; I-squared, 0%; three studies) and dyspepsia (OR, 2.33) were more common with the GLP-1 receptor agonists (Figure 72). (SOE: Low; GLP-1 receptor agonists favored for diarrhea; SOE: Moderate; Metformin favored for nausea/vomiting)
- 268. 211 Figure 72. Oddsratio of gastrointestinal adverse events comparing metformin with GLP-1 receptor agonists CI = confidence interval; GLP-1 = glucagon-like peptide-1; Group 1 = metformin; Group 2 = glucagon-like peptide-1 agonists; OR = odds ratio Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. Thiazolidinediones Versus Sulfonylureas Five RCTs compared diarrhea occurrence with pioglitazone or rosiglitazone to either glyburide, glibenclamide or glimepiride and showed no differences between treatments (Figure 73).50, 52, 74, 94, 95 The range in percentages of subjects with any type of GI adverse events in the thiazolidinediones (1% to 13%) was similar to the range with any type of GI adverse events in the sulfonylurea arms (1% to 18%), with a range in RD of -1.2% to 1.7%. Studies of diarrhea were not combined in a meta-analysis due to differences in study duration. (SOE: High; Neither favored)
- 269. 212 Figure 73. Oddsratio of gastrointestinal adverse events comparing thiazolidinediones with sulfonylureas CI = confidence interval; Group 1 = thiazolidinediones; Group 2 = sulfonylureas; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study. The width of the horizontal lines represents the 95 percent confidence intervals for each study. Thiazolidinediones Versus DPP-4 Inhibitors Two trials compared pioglitazone with sitagliptin with no meaningful difference between treatments for GI adverse events.48, 73 One trial identified solely in ClinicalTrials.gov (NCT01183013) reported no cases of diarrhea or vomiting in any of the 134 participants receiving pioglitazone (45 mg daily) or the 130 participants receiving linagliptin (5 mg daily). (SOE: Low; Neither favored) Thiazolidinediones Versus GLP-1 Receptor Agonists Two trials compared pioglitazone with exenatide for outcomes of constipation, diarrhea, dyspepsia, nausea, and vomiting.73, 105 Exenatide-treated participants tended to have more gastrointestinal side effects than those receiving pioglitazone, in both trials (range in RD 0.1% to 7% depending on the GI adverse event reported). A trial identified in ClinicalTrials.gov (NCT01147627) reported significantly more GI adverse events with exenatide compared with pioglitazone (37/142 versus 1/136 for nausea; 15/142 versus 1/136 for vomiting; 6/142 versus 4/136 for diarrhea). (SOE: Low; Thiazolidinediones favored) Sulfonylureas Versus DPP-4 Inhibitors One RCT compared glipizide (n=212) with sitagliptin (n=210).107 The authors stated that there were no significant differences between treatments for outcomes of abdominal pain, diarrhea, and vomiting at 58 weeks, but that sitagliptin-treated participants had statistically
- 270. 213 significantly less nausea(P = 0.025); the number or percent of individuals with nausea was not reported. One RCT compared glimepiride with linagliptin.106 By 34 weeks, 9 percent of individuals on each drug had an unspecified GI adverse event. One trial was identified only in ClinicalTrials.gov (NCT01006603). A comparable number of GI adverse events occurred with glimepiride and with saxagliptin (19/359 versus 15/359 for diarrhea; 8/359 versus 4/359 for nausea). (SOE: Low; Neither favored) Sulfonylureas Versus GLP-1 Receptor Agonists Three RCTs compared GI adverse events between glibenclamide and liraglutide.109, 110, 113 GI adverse events were more common with GLP-1 receptor agonists (range in RD 3% to 9% for studies lasting 52 weeks or less) (Figure 74). We did not pool any of these outcomes due to insufficient studies per outcome or differences in study duration for diarrhea. (SOE: Moderate; Sulfonylureas favored for diarrhea) (SOE: Low; Sulfonylureas favored for other GI adverse events) Figure 74. Pooled odds ratio of gastrointestinal adverse events comparing sulfonylureas with GLP-1 receptor agonists CI = confidence interval; GLP-1 = glucagon-like peptide-1; Group 1 = sulfonylureas; Group 2 = glucagon-like peptide-1 agonists; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. DPP-4 Inhibitors Versus GLP-1 Receptor Agonists One RCT compared sitagliptin to exenatide.73 At 26 weeks, there tended to be more diarrhea (5.5% vs 10.9%), dyspepsia (1.8% vs 7.3%), nausea (3.7% vs 11.3%), and vomiting (1.8% vs 4.8%) with exenatide than sitagliptin. (SOE: Low; DPP-4 inhibitors favored)
- 271. 214 Metformin Versus Metformin-BasedCombination Comparisons Metformin Versus a Combination of Metformin Plus a Thiazolidinedione Ten RCTs compared metformin with a combination of metformin and a thiazolidinedione for the rates of GI adverse events. Diarrhea was more common among the metformin monotherapy group (OR, 0.59; 95% CI, 0.45 to 0.76; I-squared, 16.4%; five studies), with no consistent differences in other GI adverse events (Figure 75).59, 67, 117-121, 125, 127, 247 Dosages of metformin were generally similar in both arms within trials, and differences in dosages between trials did not appear to impact the GI adverse events. (SOE: Low; Combination of metformin plus a thiazolidinedione favored for diarrhea; Neither favored for other GI-related outcomes) Figure 75. Odds ratio of gastrointestinal adverse events comparing metformin with a combination of metformin plus a thiazolidinedione CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a thiazolidinedione; mg = milligrams; OR = odds ratio Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. Metformin Versus a Combination of Metformin Plus a Sulfonylurea Twelve RCTs examined GI adverse events comparing metformin with metformin plus a sulfonylurea.47, 129-131, 133, 134, 136-141 No clear differences in GI adverse events were identified between arms (Figure 76). For the outcome with at least three studies, there was a pooled OR of 0.66 (95% CI, 0.34 to 1.28; I-squared, 0%; three studies) for diarrhea. We did not combine the four studies reporting on any GI adverse event owing to differences in study duration and dosing differences. Two of the studies used lower doses of metformin with combination therapy compared with monotherapy.47, 129 Studies that reported on combinations of adverse events that did not conform with the definition of “any” adverse event (diarrhea, nausea, vomiting or pain) are not included in the summary figure. (SOE: Low; Neither arm favored)
- 272. 215 Figure 76. Oddsratio of gastrointestinal adverse events comparing metformin with a combination of metformin plus a sulfonylurea* CI = confidence interval; GI = gastrointestinal; Group 1 = metformin; Group 2 = combination of metformin plus a sulfonylurea; mg = milligrams; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study. The width of the horizontal lines represents the 95 percent confidence intervals for each study. *Studies with more than one dosing arm under the same gastrointestinal outcome are reported twice in the figure to demonstrate effects of different dosing arms. Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Three RCTs compared metformin with metformin plus alogliptin.84, 154, 157 Five RCTs compared metformin with metformin plus linagliptin.86, 139, 152, 155, 162 Six RCTs compared metformin and metformin plus saxagliptin.87, 144, 146, 147, 151, 161 Twelve RCTs compared metformin and metformin plus sitagliptin.85, 118, 141-143, 145, 148, 149, 153, 156, 159, 256 There were inconsistent findings for GI adverse events depending on the outcome examined. There were no between-group differences for abdominal pain, nausea, and any GI adverse event (Figure 77 and Figure 78). Diarrhea was more common in the metformin monotherapy arm in the shorter studies but may have been more common in the metformin plus DPP-4 inhibitor arms in the longer studies (Figure 79). Vomiting occurred less often with metformin monotherapy in the longer studies, and showed no differences between groups in studies lasting less than a year (Figure 80). All the RCTs except for two147, 162 had comparable dosing of metformin in the monotherapy and combination arms. The first study with dosing differences was included in the meta-analyses where appropriate since dosing differences were small between arms (1500 to 2000 mg metformin in the monotherapy and 1500 mg metformin in the combination arm) and unlikely to impact the findings.147 Sensitivity analyses excluding this study confirmed that the study did not influence the results.147 The second study162 was excluded from the meta-analyses since the metformin monotherapy arm was at 2000 mg daily and the metformin dosing in the combination arm was only at 1000 mg daily, which could bias the study findings to favor the combination
- 273. 216 arm. In thisexcluded study,162 diarrhea occurred 16 percent of the time in the metformin monotherapy arm and 12 percent in the combination arm. These findings and other GI adverse events reported in this study162 were consistent with the meta-analysis results for the comparably-dosed studies. Two trials were identified solely in ClinicalTrials.gov (NCT00960076; NCT01076088). The former found less diarrhea with metformin monotherapy compared with metformin combined with saxagliptin (3.5% versus 5.8%). The other trial reported similar numbers of individuals with GI adverse events in the metformin monotherapy group (nine people reported diarrhea; 4 people reported nausea) and the metformin plus sitagliptin group (4 people reported diarrhea; 8 people reported nausea). (SOE for abdominal pain: Low; Neither favored for shorter and longer studies) (SOE for nausea: Moderate for shorter studies and low for longer studies; Neither favored) (SOE for any GI adverse event: Moderate for shorter studies and Low for longer studies; Neither favored) (SOE for diarrhea: Low; combination favored for shorter studies and metformin favored for longer studies) (SOE for vomiting: Low for longer studies; Metformin monotherapy favored and Moderate for shorter studies; neither favored) Figure 77. Pooled odds ratio of abdominal pain or nausea comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events.
- 274. 217 Figure 78. Pooledodds ratio of any gastrointestinal adverse event comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 275. 218 Figure 79. Pooledodds ratio of diarrhea comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 276. 219 Figure 80. Pooledodds ratio of vomiting comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor One RCT compared metformin with metformin combined with canagliflozin for diarrhea and nausea.156 Four RCTs (published in three articles) compared metformin to metformin combined with dapagliflozin for diarrhea and nausea.88, 169, 170 One RCT compared metformin to metformin combined with empagliflozin for diarrhea and nausea.153 There were no consistent differences in diarrhea between arms. Nausea tended to be more common with combination therapy, although the finding was not statistically significant (Figure 81). (SOE: Moderate; Neither favored for diarrhea) (SOE: Low; metformin favored for nausea)
- 277. 220 Figure 81. Pooledodds ratio of gastrointestinal adverse events comparing metformin with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; Group 1 = metformin; Group 2 = metformin plus a sodium-glucose co-transporter-2 inhibitor; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Four RCTs compared metformin with metformin plus a GLP-1 receptor agonist.141, 159, 174, 256 Metformin plus a GLP-1 receptor agonist showed no clear between-group differences in GI adverse events compared with metformin alone in one study,256 but there were more GI adverse events in the combination arm in the other studies (Figure 82).141, 159, 174 We did not pool data in meta-analyses due to insufficient numbers of studies and due to differences in study duration. (SOE: Low; Neither favored)
- 278. 221 Figure 82. Oddsratio of gastrointestinal adverse events comparing metformin with a combination of metformin plus a GLP-1 receptor agonist CI = confidence interval; GLP-1 = glucagon-like peptide-1; Group 1 = metformin; Group 2 = metformin plus a glucagon-like peptide-1 agonist; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study. The width of the horizontal lines represents the 95 percent confidence intervals for each study. Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea Five RCTs compared GI adverse events in the combination of metformin plus a thiazolidinedione versus metformin plus a sulfonylurea, with little between-group difference (Table 75). The RD between-groups ranged from -5.0% to 2.1%.175, 178, 180, 181, 265 (SOE: Moderate; Neither favored)
- 279. 222 Table 75. Randomizedcontrolled trials comparing a combination of metformin plus a thiazolidinedione with a combination of metformin plus a sulfonylurea on gastrointestinal adverse events Author, Year Outcome Event Rates (Metformin Plus Thiazolidinedione Versus Metformin Plus Sulfonylurea) Derosa, 2005 265 Flatulence 4.2% (2/48) versus 2.1% (1/47) Garber, 2006 180 Combined GI events Diarrhea Abdominal pain 10% (16/155) versus 11% (18/159) 3% (5/155) versus 6% (10/159) 4% (6/155) versus 6% (10/159) Umpierrez, 2006 178 Diarrhea 4.7% (5/104) versus 6% (6/96) (no difference) Hamann, 2008 175 Combined GI events 13% (38/294) versus 18% (54/301) Maffioli, 2013 181 Withdrawal owing to nausea 1% (1/86) versus 1% (1/84) GI = gastrointestinal Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor Three RCTs compared GI adverse events in the combination of metformin plus pioglitazone or rosiglitazone versus the combination of metformin plus sitagliptin with mixed results.118, 187, 188 Diarrhea and nausea tended to be more common with sitagliptin in one trial,188 but there was only one occurrence of each (4%) in the sitagliptin group (and none reported in the pioglitazone group) in the other trial.187 In the third trial, six of 87 patients in the metformin plus rosiglitazone arm and one of 94 patients in the metformin plus sitagliptin arm experienced a GI adverse event, a difference that was close to statistically significant.118 (SOE: Low; Neither favored) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist One RCT compared diarrhea, nausea, and vomiting in the combination of metformin plus pioglitazone versus the combination of metformin plus exenatide at 26 weeks.188 There was more diarrhea (7% versus 18%), nausea (5% versus 24%), and vomiting (3% versus 11%) in the group receiving exenatide; all differences are statistically significant. The range in ORs was 2.9 to 6.3, and range in RDs were 8% to 19%. (SOE: Moderate; Combination of metformin plus a thiazolidinedione favored) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor Seven RCTs compared diarrhea, abdominal pain, nausea, vomiting, and unspecified GI adverse events in the combination of metformin plus glipizide or glimepiride versus metformin plus alogliptin, linagliptin, sitagliptin, or saxagliptin, with no difference between treatments (Figure 83).139, 141, 192-195, 197 The OR for the four trials with similar study duration of 104 weeks for diarrhea was 0.97 (95% CI, 0.76 to 1.24; I-squared, 0 percent).141, 194, 195, 197 No single study strongly influenced the results, and no substantial heterogeneity was identified. One trial was identified in ClinicalTrials.gov (NCT00856284) that randomized 869 people to metformin plus 5 mg of glipizide and 378 people to metformin plus 25 mg of alogliptin. Metformin plus alogliptin had more GI adverse events (32 nausea, 1 severe vomiting, 60
- 280. 223 diarrhea) than metforminplus glipizide (20 nausea and 63 diarrhea). No severe diarrhea was described in either group. (SOE: High; Neither favored) Figure 83. Odds ratio of gastrointestinal adverse events comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = a combination of metformin plus a sulfonylurea; Group 2 = a combination of metformin plus a dipeptidyl peptidase-4 inhibitor; Met = metformin; OR = odds ratio; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study. The width of the horizontal lines represents the 95 percent confidence intervals for each study. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three RCTs compared metformin plus glipizide or glimepiride with metformin plus dapagliflozin or empagliflozin on the outcomes of diarrhea and nausea.54, 199, 200 There was little difference between treatments. A meta-analysis was not performed because the trials reported adverse events at 1, 2, and 4 years after randomization. One trial identified from ClinicalTrials.gov (NCT01368081) reported no significant difference in the number of individuals with diarrhea with metformin plus sulfonylurea (5 of 63) and metformin plus empagliflozin (4 of 65). (SOE: Low; Neither favored) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three RCTs compared metformin plus glibenclamide or glimepiride with metformin plus albiglutide or exenatide for diarrhea, nausea, or vomiting, with a similar and low incidence of adverse events in one trial of exenatide (diarrhea: 1/65 versus 2/63; vomiting: 1/65 versus 1/63),205 but a greater percentage of individuals with GI adverse events with exenatide in another trial (diarrhea: 7% versus 12%; nausea: 2% versus 29%; vomiting: 2% versus 9%).53 Diarrhea and nausea were more common with albiglutide (diarrhea: 11.9% versus 8.6%; nausea: 8.9%
- 281. 224 versus 5.2%; vomiting:5.6% versus 3.6%; unspecified adverse events: 36.4% versus 37.6%).141 A meta-analysis was not performed because the trials differed in duration. (SOE: Low; Combination of metformin plus a sulfonylurea favored) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Two trials compared metformin plus sitagliptin with metformin plus different doses of canagliflozin or empagliflozin.153, 156 Diarrhea and nausea were reported for the canagliflozin comparisons, and dyspepsia was reported for the empagliflozin comparisons. There were no clear differences in GI side effects for either medication. (SOE: Low; Neither favored) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Four trials reported on GI adverse events comparing metformin plus sitagliptin with metformin plus albiglutide, dulaglutide, or exenatide.141, 159, 188, 256 There were more GI adverse events with GLP-1 receptor agonists in three of the four trials, with a range in RD of 0% to 23% (Figure 84). (SOE: Moderate; Combination of metformin plus a DPP-4 inhibitor favored) Figure 84. Odds ratio of gastrointestinal adverse events comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; Group 1 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; Group 2 = combination of metformin plus a glucagon-like peptide-1 agonist; Met = metformin; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study. The width of the horizontal lines represents the 95 percent confidence intervals for each study.
- 282. 225 Combination of MetforminPlus a GLP-1 Receptor Agonist Versus a Combination of Metformin Plus a Premixed Insulin One trial compared metformin plus exenatide with metformin plus insulin aspart 70/30 at 26 weeks for diarrhea, dyspepsia, nausea, and vomiting.213 There was no difference in diarrhea between the treatments (11% versus 8%). Differences between groups for the other outcomes could not be evaluated because they were only reported for exenatide (6% dyspepsia; 19% nausea; 10% vomiting). (SOE: Insufficient) Combination of Metformin Plus a Basal Insulin Versus a Combination of Metformin Plus a Premixed Insulin One RCT compared metformin in a combination regimen with either insulin glargine or lispro for diarrhea; neither arm was favored.214 One trial was reported only in ClinicalTrials.gov (NCT01068652). Thirteen of 200 people who received metformin plus insulin detemir had diarrhea compared with 15 of 203 people in the metformin plus biphasic insulin aspart 30 group. (SOE: Low; Neither favored) Strength of Evidence for Gastrointestinal Side Effects The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 76, Table 77, and Table 78 and summarized in the key points. All studies were RCTs. Study limitations for most comparisons in the strength of evidence table were graded as low or medium; only two comparisons were graded as having high study limitations owing to lack of description of randomization or blinding or failure to describe withdrawals or dropouts. In general, we did not find strong relative differences in outcomes by study quality. We did not find any evidence of publication bias comparing results published in peer-reviewed journals to results published on ClinicalTrials.gov. However, for the comparison of thiazolidinediones with GLP-1 receptor agonists, an additional trial was found in ClinicalTrials.gov with consistent findings to the two published studies favoring thiazolidinediones. This study would likely increase the strength of evidence from low to moderate. We considered GI side effects a direct outcome, because they were measured directly from patient report. The most common reasons for downgrading the evidence were imprecision, inconsistency, and study limitations.
- 283. 226 Table 76. Strengthof evidence domains for monotherapy comparisons in terms of gastrointestinal side effects among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. TZD 8 (6,250) Medium Consistent Direct Precise Undetected Moderate TZD favored for diarrhea, 0.24 (0.17 to 0.34) and nausea Metformin vs. SU 12 (6094) Medium Consistent Direct Imprecise Undetected Moderate SU favored for diarrhea, 0.42 (0.24 to 0.72); abdominal pain, 0.44 (0.29 to 0.67); nausea and vomiting, 0.45 (0.31 to 0.65); and any GI adverse events, 0.45 (0.28 to 0.72) Metformin vs. DPP-4 inhibitors 6 (5,842) Low Consistent Direct Precise Undetected High DPP-4 inhibitors favored for nausea, 0.37 (0.15 to 0.91) and diarrhea, 0.38 (0.18 to 0.83) Metformin vs. SGLT-2 inhibitors 4 (2,041) Medium Consistent Direct Imprecise Undetected Low SGLT-2 inhibitors favored for diarrhea and nausea Metformin vs. GLP-1 receptor agonists 3 (879) Low Inconsistent Direct Imprecise for diarrhea; Precise for nausea/ vomiting Undetected Low for diarrhea; Moderate for nausea/ vomiting GLP-1 receptor agonists favored for diarrhea; Metformin favored for nausea/vomiting, 1.73 (1.01 to 2.95) TZD vs. SU 5 (6,432) Low Consistent Direct Precise Undetected High Neither favored; range in OR, 0.8 to 2.0; range in RD, -1.2% to 1.7% TZD vs. DPP-4 inhibitors 2 (1,031) Low Inconsistent Direct Imprecise Undetected Low Neither favored TZD vs. GLP-1 receptor agonists 2 (1,236) Low Consistent Direct Imprecise Undetected ‡ Low TZD favored SU vs. DPP-4 inhibitors 2 (653) Low Inconsistent Direct Imprecise Undetected Low Neither favored
- 284. 227 Table 76. Strengthof evidence domains for monotherapy comparisons in terms of gastrointestinal side effects among adults with type 2 diabetes (continued) Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † SU vs. GLP-1 receptor agonists 3 (1,568) High Consistent Direct Precise for diarrhea; Imprecise for all other GI adverse events Undetected Moderate for diarrhea; Low for abdominal pain, any GI adverse event, nausea and vomiting SU favored; range in OR for diarrhea, 1.5 to 2.4; range in RD for diarrhea, 3% to 9% DPP-4 inhibitors vs. GLP-1 receptor agonists 1 (820) Low Not applicable Direct Not evaluated Undetected Low DPP-4 inhibitors favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; OR = odds ratio; RD = risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence. ‡ An additional article was found in clinical trials.gov which was consistent with the two other studies favoring thiazolidinediones over GLP-1 receptor agonists. Inclusion of this study may have increased our strength of evidence from low to moderate.
- 285. 228 Table 77. Strengthof evidence domains for metformin versus metformin combination comparisons in terms of gastrointestinal side effects among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + TZD 10 (3,878) Medium Consistent Direct Imprecise Undetected Low Metformin + TZD favored for diarrhea; Neither favored for other GI-related outcomes Metformin vs. metformin + SU 12 (4,317) Low Inconsistent Direct Imprecise Undetected Low Neither drug favored for diarrhea or any GI adverse events Metformin vs. metformin + DPP- 4 inhibitors 26 (14,324) Low Consistent for nausea, any GI adverse event, and vomiting; Inconsistent for all others Direct Precise for nausea, any GI adverse event, and vomiting for shorter studies; Imprecise for all others Undetected Moderate for any GI adverse event, nausea, and vomiting (shorter studies) Low for abdominal pain, diarrhea, and vomiting (longer studies) Neither favored for shorter studies; 0.9 (0.6 to 1.3) for nausea; 0.9 (0.7 to 1.3) for any GI adverse event; 1.1 (0.6 to 2.0) for vomiting; and for abdominal pain For diarrhea, the combination was favored in the shorter studies and metformin monotherapy was favored in the longer studies For vomiting in the longer studies, metformin monotherapy was favored Metformin vs. metformin + SGLT-2 inhibitors 6 (2,918) Low Consistent Direct Precise Undetected Moderate for diarrhea; Low for nausea Neither favored for diarrhea, 0.9 (0.5 to 1.5), Metformin favored for nausea Metformin vs. metformin + GLP- 1 receptor agonists 4 (2,713) Low Inconsistent Direct Imprecise Undetected Low Neither favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 286. 229 Table 78. Strengthof evidence domains for metformin-based combination comparisons in terms of gastrointestinal side effects among adults with type 2 diabetes Comparison Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + TZD vs. metformin + SU 5 (1,382) Low Consistent Direct Precise Undetected Moderate Neither favored; range in OR, 0.5 to 2.0; range in RD, -5% to 2.1% Metformin + TZD vs. metformin + DPP-4 inhibitors 3 (747) Low Inconsistent Direct Imprecise Undetected Low Neither favored Metformin + TZD vs. metformin + GLP-1 receptor agonists 1 (514) Low Not applicable Direct Imprecise Undetected Moderate Metformin + TZD favored; range in OR, 2.9 to 6.3; range in RD, 8% to 19% Metformin + SU vs. metformin + DPP-4 inhibitors (long-term studies) 7 (8,321) Low Consistent Direct Precise Undetected High Neither favored for diarrhea at 104 weeks; 1.0 (0.8 to 1.2) Metformin + SU vs. metformin + SGLT-2 inhibitors 3 (3,177) Low Consistent Direct Imprecise Undetected Low Neither favored Metformin + SU vs. metformin + GLP-1 receptor agonists 3 (2,018) Medium Inconsistent Direct Imprecise Undetected Low Metformin + SU favored Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors 2 (946) Medium Not applicable Direct Not evaluated Undetected Low No difference Metformin + DPP-4 inhibitors vs. metformin + GLP-1 receptor agonists 4 (2,891) Medium Consistent Direct Precise Undetected Moderate Metformin + DPP-4 inhibitors favored; range in OR, 1.0 to 5.1; range in RD, 0% to 23% Metformin + GLP-1 receptor agonists vs. metformin + premixed insulin 1 (363) High Not applicable Direct Imprecise Undetected Insufficient Unable to determine Metformin + basal insulin vs. metformin + premixed insulin 1 (317) Medium Not applicable Direct Imprecise Undetected Low Neither favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; OR = odds ratio; RD = risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack
- 287. 230 of available evidence.Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 288. 231 Evidence for Cancer MonotherapyComparisons Metformin Versus Thiazolidinediones A single retrospective cohort study, from the England Cancer Registry, reported no difference in non-melanoma cancer risk for thiazolidinediones users (N=31,372) versus metformin users (N=109,708) (adjusted RR, 0.96; 95% CI, 0.81 to 1.13, P not reported) over 4 years of followup.255 (Not graded) Metformin Versus Sulfonylureas Four retrospective, cohort studies compared cancer outcomes for metformin and sulfonylurea users (Table 79).225, 254, 255, 266 Three studies reported no difference between metformin and sulfonylurea users.225, 255, 266 The other study only provided results stratified by statin use, indicating a possible interaction with statin use.254 (SOE: Low; Neither favored for long-term cancer risk) Table 79. Retrospective cohort studies comparing metformin with sulfonylureas on cancer Author, Year Population Followup Outcome Results van Staa, 2012 255 England Cancer Registry [n=68,209 (sulfonylurea); n=109,708 (metformin)] 4-5 years Non-melanoma cancer HR 1.03; 95% CI, 0.91 to 1.17 Reference=metformin Andersson, 2010 225 Danish Patient Registry Patients with heart failure (N=5,852) 10 years Death from cancer HR 1.01; 95% CI, 0.72 to 1.43 Reference=sulfonylurea Lehman, 2012 254 Veterans Health Administration [n=533 (metformin- statin); n=2404 (sulfonylurea-statin); n=175 (metformin-no statin); n=1,930 (sulfonylurea-no statin)] 270.4 weeks Incident prostate cancer Statin users HR 0.69; 95% CI, 0.5 to 0.92, P = 0.01 Reference=sulfonylurea Non users of statins HR 2.15; 95% CI, 1.83 to 2.52, P < 0.0001 Reference=sulfonylurea Kowall, 2015 266 German Disease Analyzer (IMS Health) – primary care clinics (N=22,556) 4.8 years Incident cancer by ICD-10 code Adjusted HR, 1.09; 95% CI, 0.87 to 1.36 Reference=metformin CI = confidence interval; HR = hazard ratio; ICD-10 = International Classification of Diseases Metformin Versus DPP-4 Inhibitors Two RCTs of metformin plus placebo (N=510) versus a DPP-4 inhibitor plus placebo (N=514) evaluated cancer outcomes.85, 87 Each study reported one cancer event: one death due to pancreatic neoplasm/sepsis87 and one occurrence of esophageal cancer,85 in the metformin arms. Neither study reported on cancer in the DPP-4 inhibitor arm.85, 87 Followup ranged from 7687 to 104 weeks.85 (SOE: Insufficient)
- 289. 232 Metformin Versus SGLT-2Inhibitors A single RCT (N=404) of metformin plus placebo versus dapagliflozin plus placebo, with 24 weeks of followup, reported on occurrence of cancer.88 The RCT reported no bladder cancer in either arm and a single case of breast cancer in the SGLT-2 arm; breast cancer was not reported on in the metformin arm.88 (SOE: Insufficient) Thiazolidinediones Versus Sulfonylureas A 56-week, multi-center trial in the US, including Puerto Rico, reported two events of stage IV colon cancer (2/251, 0.8%) in the sulfonylurea arm and none in the thiazolidinedione arm (0/251, 0.0%).95 (SOE: Low; Thiazolidinediones favored) Sulfonylureas Versus DPP-4 Inhibitors Two short-term RCTs compared sulfonylurea with DPP-4 inhibitor monotherapy and reported on cancer outcomes collected through passive ascertainment.106, 107 One 52-week RCT reported one case of colon cancer in the glimepiride arm (1/76, 1.3%) and did not report on cancer in the linagliptin arm (NR/151).106 The other RCT compared glipizide with sitagliptin among participants with at least moderate renal insufficiency and reported five cases of cancer in the sitagliptin arm (5/210, 2.3%) and none in the sulfonylurea arm (0/212, 0.0%), over 58 weeks of followup.107 In that study, cancer cases were chronic myeloid leukemia in a participant with baseline leukocytosis, breast cancer diagnosed after 4 days of sitagliptin initiation, lung cancer in a participant with 40 pack-years of smoking, a pancreatic mass, and a case of polycythemia vera in a participant with a germline JAK-2 mutation.107 (SOE: Insufficient) Sulfonylureas Versus GLP-1 Receptor Agonists One RCT compared glimepiride with liraglutide and reported on cancer outcomes with 104 weeks of followup but did not report if ascertainment was active.113 Two cases of breast cancer occurred in the liraglutide 1.8 mg arm (2/251, 0.8%) and two cases of thyroid tumors (one benign thyroid neoplasm and one papillary thyroid cancer) in the liraglutide 1.2 mg arm (1/247, 0.4%); cancer outcomes were not reported on in the sulfonylurea arm (N=248).113 These events were only reported if they were considered to be possibly related to the trial drug.113 (SOE: Insufficient) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a Sulfonylurea A single RCT compared metformin (N=101) with metformin plus glimepiride (N=307) and reported no cases of thyroid cancer in either arm at 104 weeks.141 Of note, withdrawal rates were greater than 30 percent across arms, and the investigators did not use an intention-to-treat analysis. (SOE: Insufficient) Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Nine RCTs compared metformin with metformin plus a DPP-4 inhibitor.85, 87, 105, 141, 142, 154, 159, 160, 162 We did not combine these studies in a meta-analysis because of lack of consistent ascertainment of and reporting on cancer outcomes; and, when reported, heterogeneous definitions of cancer outcomes (Table 80). Two studies had long-term follow up (104 weeks).85, 141 Of these, the RCT with active ascertainment reported two cases of thyroid cancer in the
- 290. 233 metformin plus DPP-4inhibitor arm (0.7%) and none in the metformin arm (0.0%) at 104 weeks.141 Results were mixed across the other studies, with many arms not reporting on cancer outcomes. (SOE: Insufficient) Table 80. Randomized controlled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on cancer Author, Year Followup (Weeks) Outcome Active/Passive Ascertainment Metformin Events/N (%) Metformin + DPP- 4 Inhibitor Events/N (%) Ji, 2015 162 14 Pancreatic cancer NR 0/345 Linagliptin 5 mg: 0/344 (0) Wang, 2015 160 24 Gastric cancer NR 0/100 (0) 1/205 (0.5) Pancreatic cancer NR 0/100 (0) 0/205 (0) Nauck, 2009 154 26 Discontinuation because of prostate cancer Passive NR/104 Alogliptin 12.5 mg: 1/213 (0.5) Alogliptin 25 mg: NR/210 Discontinuation because of endometrial cancer Passive NR/104 Alogliptin 12.5 mg: 1/213 (0.5) Alogliptin 25 mg: NR/210 Nauck, 2014 159 * 26 Thyroid cancer NR 0/177 (0) 0/315 (0) Raz, 2008 142 30 Cases of cancer Active 3/94 (3) 0/96 (0) Xu, 2015 105 48 Cholangiocarcinoma NR 0/136 (0) 1/142 (0.7) Pfutzner, 2011 87 76 Death due to pancreatic neoplasm/sepsis NR 1/328 (0.3) Saxagliptin 5 mg: NR/320 Saxagliptin 10 mg: NR/323 Ahren, 2014 141 104 Thyroid cancer Active 0/101 (0) 2/302 (0.7) Williams-Herman, 2010 85 104 Esophageal cancer NR Metformin 1000 mg: 1/182 (0.5) Metformin 2000 mg: NR/182 Metformin 2000 mg + sitagliptin 100 mg: NR/182 Metformin 1000 mg + sitagliptin 100 mg: NR/190 DPP-4 = dipeptidyl peptidase-4; mg = milligram; NR = not reported *Cancer outcome at 52 weeks not reported in the metformin arm and none reported in the MET+DPP-4 arm Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Four RCTs compared metformin with metformin plus an SGLT-2 inhibitor and reported on cancer outcomes (Table 81).88, 165, 169, 170 Reporting of cancer was incomplete for many studies, and studies did not report on whether there was active ascertainment for cancer outcomes. Therefore, we did not perform a meta-analysis for this comparison (Table 81). Cancer outcomes were rare but appeared to occur at similar rates in the treatment arms; most studies were small and less than 1 year in duration. (SOE: Low; Neither favored)
- 291. 234 Table 81. Randomizedcontrolled trials comparing metformin with a combination of metformin plus a SGLT-2 inhibitor on cancer Author, Year Followup (Weeks) Outcome Active/Passive Ascertainment Metformin Events/N (%) Metformin + SGLT-2 Inhibitor Events/N (%) Qiu, 2014 165 18 Colon cancer NR NR/93 Canagliflozin 100 mg: NR/93 Canagliflozin 300 mg: 1/93 (1.1) Henry, 2012 88 24 Bladder malignancy NR 0/201 (0) 0/194 (0) Bolinder, 2012 169 50 Prostatic cancer or prostatic adenoma NR 1/91 (1.1) 1/91 (1.1) Basal cell carcinoma NR 1/91 (1.1) NR/91 Breast cancer leading to discontinuation NR 0/91 (0) 1/91 (1.1) Bailey, 2013 170 102 Unspecified adverse event (lung cancer) NR 1/137 (0.7) Dapagliflozin 2.5 mg: NR/137 Dapagliflozin 5 mg: 1/137 (0.7) (bladder cancer) Dapagliflozin 10 mg: 1/135 (0.7) (breast cancer) mg = milligram; NR = not reported; SGLT-2 = sodium-glucose co-transporter-2 Metformin Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two RCTs compared metformin with metformin plus a GLP-1 receptor agonist and reported on cancer outcomes.141, 159 One trial did active surveillance for thyroid cancer and reported one case of follicular thyroid cancer in the metformin plus albiglutide arm (1/302, 0.3%) and no cases (0/101, 0.0%) in the metformin arm at 104 weeks.141 The 52-week RCT reported no cases of thyroid cancer in the metformin plus dulaglutide arms (metformin plus dulaglutide 0.75 mg/week, n=302; metformin plus dulaglutide 1.5 mg/week, n=304) and did not report on thyroid cancer in the metformin arm (n=177).159 (SOE: Low; Metformin favored) Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea One 24-week RCT compared the combination of moderately-dosed metformin plus pioglitazone with the combination of moderately-dosed metformin plus glimepiride and reported on cancer outcomes, but whether ascertainment was active was not reported.185 A single case of prostate cancer occurred in the metformin plus glimepiride arm (1/142, 0.7%), and cancer was not reported on in the metformin plus pioglitazone arm (n=146).185 More than 20 percent of participants withdrew from each arm.185 (SOE: Insufficient)
- 292. 235 Combination of MetforminPlus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor A single 26-week RCT compared the combination of metformin plus pioglitazone with the combination of metformin plus sitagliptin and reported on cancer outcomes; the method of ascertainment was not described.188 A single case of papillary thyroid cancer occurred in the metformin plus sitagliptin arm (1/166, 0.6%), and no events were reported in the metformin plus pioglitazone arm (0/165, 0.0%).188 More participants withdrew from the metformin plus pioglitazone arm (21%) than the metformin plus sitagliptin arm, and the investigators did not use an intention-to-treat analysis for this outcome.188 (SOE: Insufficient) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist A single 26-week RCT compared the combination of metformin plus pioglitazone with the combination of metformin plus weekly exenatide and reported on cancer outcomes; the method of ascertainment was not described.188 No cases of thyroid cancer were reported in either arm (metformin plus pioglitazone: 0/165, 0.0% and metformin plus exenatide: 0/160, 0.0%).188 (SOE: Insufficient) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor Three RCTs, with 104 weeks of followup, compared the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor and reported on cancer outcomes (Table 82).141, 194, 195 Cancer incidence was slightly higher in the metformin plus DPP- 4 arms than the metformin plus sulfonylurea arms, in these 2-year trials. More than 20 percent of participants withdrew from these studies. Two of the studies used an intention-to-treat analysis.194, 195 An additional RCT with only 52 weeks of followup also reported a higher incidence of cancer in metformin plus DPP-4 inhibitor arm than in the metformin plus a sulfonylurea.193 (SOE: Low; Combination of metformin plus a sulfonylurea favored) Table 82. Randomized controlled trials comparing a combination of metformin with a sulfonylurea with a combination of metformin plus a DPP-4 inhibitor on cancer Author, Year Followup (Weeks) Outcome Active/Passive Ascertainment ITT Analysis Metformin + SU Events/N (%) Metformin + DPP-4 Inhibitor Events/N (%) Ahren, 2014 141 104 Thyroid cancer Active No 0/307 (0.0%) 2/302 (0.7%) Gallwitz, 2012 194 104 Prostate, breast, and colon cancer* NR Yes 7/775 (0.9%) 10/776 (1.3%) Goke, 2010 195 104 Acute myeloid leukemia NR Yes NR/430 1/428 (0.2%) Schernthaner, 2015 193 52 Neoplasm NR No 3/360 (0.8) 10/360 (2.8%) DPP-4 = dipeptidyl peptidase-4; ITT = intention-to-treat; NR = not reported; SU = sulfonylurea * Unclear if ascertained for specific types of cancer
- 293. 236 Combination of MetforminPlus a Sulfonylurea Versus a Combination of Metformin Plus a SGLT-2 Inhibitor Two RCTs compared the combination of metformin plus an sulfonylurea with the combination of metformin plus an SGLT-2 inhibitor and reported on cancer outcomes.199, 201 In one 104-week trial using passive ascertainment, the authors reported more cases of cancer (prostate cancer, n=3; breast cancer, n=1; gastric cancer, n=1; and pancreatic cancer, n=2) in the metformin plus dapagliflozin arm (7/406, 1.7%) than in the metformin plus glipizide arm (prostate cancer, basal cell skin cancer, and lung cancer; 3/408, 0.7%).199 In the other RCT (also with 104 weeks of followup), a single death due to cervical cancer was reported in the metformin plus sulfonylurea arm; the study did not report on this outcome for the metformin plus SGLT-2 inhibitor arms.201 (SOE: Insufficient) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Two RCTs compared the combination of metformin plus glimepiride with the combination of metformin plus a GLP-1 receptor agonist and reported on cancer outcomes.53, 141 Both trials reported thyroid cancer events in the metformin plus GLP-1 receptor agonist arm.53, 141 In the study by Ahren 2014, the investigators actively ascertained for thyroid cancer with 104 weeks of followup and found no thyroid cancer in the metformin plus glimepiride arm (0/307, 0.0%) and one case in the metformin plus albiglutide arm (1/302, 0.3%).141 In the other trial (maximum followup of 3 years), which did not report on the method of ascertainment of cancer outcomes, the authors reported three cases of thyroid cancer (3/511, 0.6%) in the metformin plus exenatide arm and did not report on thyroid cancer for the metformin plus sulfonylurea arm. In that trial, the authors also reported a single case of breast cancer in the metformin plus sulfonylurea arm (1/508, 0.2%) and did not report on breast cancer for the metformin plus exenatide arm.53 (SOE: Low; Combination of metformin plus a sulfonylurea favored for long-term risk of thyroid cancer) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist Three short-term RCTs159, 188, 210 and one long-term RCT compared the combination of metformin plus a GLP-1 receptor agonist with the combination of metformin plus a DPP-4 inhibitor and found conflicting results on risk of thyroid cancer (Figure 85). None of the short- term studies reported active ascertainment of thyroid cancer, and one did not provide results for the intention-to-treat population.188 The long-term RCT (104 weeks) actively ascertained thyroid cancer and reported two events in the metformin plus DPP-4 arm and one event in the metformin plus GLP-1 receptor agonist arm; the authors did not evaluate this outcome in the intention-to- treat population.141 Withdrawal rates were high across the study arms (range, 13% to 77%) with most arms having more than 30 percent losses to followup.141, 159, 188, 210 (SOE: Low; Combination of metformin plus a GLP-1 receptor agonist favored)
- 294. 237 Figure 85. Pooledodds ratio of cancer events comparing the combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus a GLP-1 receptor agonist CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; Met = metformin; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a Basal Insulin A single 25-week RCT compared the combination of metformin plus sitagliptin with the combination of metformin plus insulin glargine and reported two cases of cancer (Kaposi’s sarcoma and prostate cancer) in the metformin plus sitagliptin arm (2/264, 0.8%) and none in the metformin plus insulin glargine arm (0/237, 0.0%).211 (SOE: Low; Combination of metformin plus a basal insulin favored) Strength of Evidence for Cancer We found low or insufficient strength of evidence on cancer outcomes for all comparisons of interest, as described in the Key Points and Table 83, Table 84, and Table 85. The major limitation of the evidence on cancer for the diabetes medication comparisons was the lack of studies. For RCTs, major study limitations included high rates of withdrawals (>20%) combined with lack of an intention-to-treat approach and lack of active ascertainment of, or reporting on, cancer outcomes. We usually could not determine consistency because of a lack of studies (i.e., one study available for a given comparison), or evidence was graded as inconsistent based on only a few studies for each comparison. The evidence on all comparisons was imprecise because of insufficient sample size for cancer outcomes. We identified several unpublished studies that may have affected our grading of the evidence. For the comparison of sulfonylurea monotherapy and DPP-4 inhibitor monotherapy, two unpublished studies favored sulfonylurea monotherapy. These results could have moved our evidence grade from “insufficient” to “low” for this comparison and suggested that sulfonylurea monotherapy is favored over DPP-4 inhibitors. Also, addition of an unpublished study with long-
- 295. 238 term followup ofthe published study comparing metformin to metformin plus a sulfonylurea would have led to the conclusion that metformin is favored over the combination of metformin plus a sulfonylurea, although with low strength of evidence. The same unpublished study also provided additional results for the comparison of metformin to combination therapy with metformin and a DPP-4 inhibitor and to the combination of metformin plus a GLP-1 receptor agonist; results suggested that metformin was favored over both combination therapies. These would also have been conclusions based on low strength of evidence for both comparisons. Two unpublished studies with long-term follow up supported the published evidence that the combination of metformin plus a sulfonylurea is favored over the combination of metformin plus a DPP-4 inhibitor. We found two additional unpublished studies of comparisons for which there were no published studies: thiazolidinedione vs. DPP-4 inhibitors and metformin plus basal insulin vs. metformin plus premixed insulin. Finally, most evidence for the comparisons of interest included studies that did not report on cancer events in all arms. While this limited our ability to synthesize data quantitatively, we do not believe that this was a source of selective analysis reporting bias, as much as a reflection of a lack of a focus on active ascertainment and reporting of cancer outcomes.
- 296. 239 Table 83. Strengthof evidence domains for monotherapy comparisons and cancer outcomes among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. SU Observational: 4 (211,367) Medium Consistent Direct Imprecise NA Low Neither favored Metformin vs. DPP-4 inhibitors RCTs: 2 (1,014) Low Consistent Direct Imprecise Undetected Insufficient Unable to determine Metformin vs. SGLT-2 inhibitors RCT: 1 (404) Medium Unknown Direct Imprecise Undetected Insufficient Unable to determine TZD vs. SU RCT: 1 (502) Medium Unknown Direct Imprecise Undetected Low TZD favored SU vs. DPP-4 inhibitors RCTs: 2 (653) Low Inconsistent Direct Imprecise Suspected Insufficient Unable to determine SU vs. GLP-1 receptor agonists RCT: 1 (746) Medium Unknown Direct Imprecise Undetected Insufficient Unable to determine DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; NA = not applicable; RCT = randomized controlled trial; SGLT- 2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating the outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 297. 240 Table 84. Strengthof evidence domains for metformin versus metformin-based combination comparisons and cancer outcomes among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + SU RCT: 1 (1049) High Unknown Direct Imprecise Suspected Insufficient Unable to determine Metformin vs. metformin + DPP-4 inhibitors RCTs: 8 (6266) Low Inconsistent Direct Imprecise Suspected Insufficient Unable to determine 4 RCTs did not report on events in all arms Metformin vs. metformin + SGLT-2 inhibitors RCTs: 4 (1610) Low Consistent Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + GLP-1 receptor agonists RCTs: 2 (2147) High Inconsistent Direct Imprecise Suspected Low Metformin favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; NA = not applicable; RCT = randomized controlled trial; SGLT- 2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating the outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 298. 241 Table 85. Strengthof evidence domains for combination therapy comparisons and cancer among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + pio vs. metformin + SU RCT: 1 (305) High Unknown Direct Imprecise Undetected Insufficient Unable to determine Metformin + pio vs. metformin + DPP-4 inhibitors RCT: 1 (514) High Unknown Direct Imprecise Undetected Insufficient Unable to determine Metformin + pio vs. metformin + GLP-1 receptor agonists RCT: 1 (514) High Unknown Direct Imprecise Undetected Insufficient Unable to determine Metformin + SU vs. metformin + DPP-4 inhibitors (long-term studies) RCTs: 4 (4,179) Medium Consistent Direct Imprecise Suspected Low Metformin + SU favored for longer-term cancer risk Metformin + SU vs. metformin + SGLT-2 inhibitors (long-term studies) RCT: 2 (2264) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Metformin + SU vs. metformin + GLP-1 receptor agonists (long-term studies) RCT: 2 (2,078) High Consistent Direct Imprecise Undetected Low Metformin + SU favored for long-term thyroid cancer risk Metformin + DPP-4 inhibitors vs. metformin + GLP-1 receptor agonists RCT: 4 (3,107) High Consistent Direct Imprecise Undetected Low Metformin + GLP-1 receptor agonists favored Metformin + DPP-4 inhibitor vs. metformin + basal insulin RCT: 1 (515) Medium Unknown Direct Imprecise Undetected Low Metformin + basal insulin favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; pio = pioglitazone; RCT = randomized controlled trial; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating the outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 299. 242 Evidence for CongestiveHeart Failure Monotherapy Comparisons Metformin Versus Thiazolidinediones Three RCTs50, 70, 76 and two observational studies233, 243 examined heart failure for the comparison of metformin versus thiazolidinediones (Table 86). We did not conduct a meta- analysis because of differences in study duration and design. The two RCTs, each lasting less than 1 year, showed no events of heart failure in either arm.70, 76 The third RCT, the ADOPT study,50 had over 1,400 subjects in each arm with a median duration of treatment of 4 years. In this study, the investigators compared metformin with rosiglitazone on the primary outcome of time to monotherapy failure. While the study was not powered to detect differences in cardiovascular events and excluded patients with heart failure at baseline, there was no statistically significant difference between the incidence of investigator-reported heart failure in these two arms (22/1456 for rosiglitazone versus 19/1454 for metformin).50 Two observational studies with 6 to 8 years of followup also compared metformin with thiazolidinediones.233, 243 Both studies reported point estimates suggesting harm from thiazolidinediones compared with metformin; the results were only close to statistically significant for the comparison of pioglitazone versus metformin in one of the two studies (Table 86).233 (SOE: Low; Metformin favored) Table 86. Studies comparing metformin with thiazolidinediones on congestive heart failure Author, Year Study Design Comparison Heart Failure Incidence (Metformin as Reference Group) Kahn, 2006 50 RCT Rosiglitazone versus metformin 22/1456 versus 19/1454 versus; OR, 1.2 (95% CI, 0.6 to 2.3) Erem, 2014 70 RCT Pioglitazone versus metformin 0/19 versus 0/19 Esposito, 2011 76 RCT Pioglitazone versus metformin 0/55 versus 0/55 Pantalone, 2009 233 Observational study Rosiglitazone versus Metformin HR, 1.16 (95% CI, 0.78 to 1.73) Pioglitazone versus metformin HR, 1.38 (95% CI, 1.00 to 1.90) Hsiao, 2009 243 Observational study Rosiglitazone versus metformin HR, 1.30 (95% CI, 0.89 to 1.89) Pioglitazone versus metformin HR, 1.54 (95% CI, 0.65 to 3.64) CI = confidence interval; HR = hazard ratio for thiazolidinediones with metformin as reference group; OR = odds ratio; RCT = randomized controlled trial Metformin Versus Sulfonylureas Two studies (one RCT and one observational study) reported on the risk of heart failure events with metformin compared with the sulfonylureas, with both point estimates favoring metformin over sulfonylureas (Table 87).231, 233 The 144-week RCT compared metformin with glipizide in adults with diabetes and a history of coronary artery disease, and reported a small, non-significant, greater number of events in the glipizide arm (10/148) compared to the metformin arm (9/156).231 Rescue therapy was insulin and was initiated in about 20 percent of each arm. The larger retrospective observational study (N=20,450) compared metformin with
- 300. 243 sulfonylurea in patientswithin one health care system in the United States from 1998 to 2006.233 After adjusting for differences in baseline patient characteristics (e.g., gender, race, age, smoking status, and medications), sulfonylureas were associated with a greater risk of heart failure than metformin.233 (SOE: Low; Metformin favored) Table 87. Studies comparing metformin with sulfonylureas on congestive heart failure Author, Year Study Design Comparison Heart Failure Incidence (Sulfonylurea as Reference Group) Hong, 2013 231 RCT Metformin versus glipizide HR, 0.82 (95% CI, 0.31 to 2.13) Pantalone, 2009 233 Observational study Metformin versus sulfonylurea (unspecified drug type) HR, 0.76 (95% CI, 0.64 to 0.91) CI = confidence interval; HR = hazard ratio for metformin with sulfonylureas as the reference group; RCT = randomized controlled trial Metformin Versus DPP-4 Inhibitors One RCT, lasting 26 weeks, compared metformin with alogliptin, with no heart failure events in either arm.84 (SOE: Low; Neither favored) Thiazolidinediones Versus Sulfonylureas Four trials49, 50, 52, 217 and two observational studies233, 243 examined heart failure outcomes for the comparison of thiazolidinediones versus sulfonylureas (Table 88), finding no clear between- group differences. A meta-analysis of the four RCTs49, 50, 52, 217 showed an increased risk of congestive heart failure with thiazolidinediones compared with sulfonylureas, which did not reach statistical significance (pooled OR, 1.62; 95% CI, 0.95 to 2.76) (Figure 86). There was no evidence of statistical heterogeneity among the included studies (I2 = 0%). Consistent with the meta-analysis of the RCTs, the two observational studies also showed increased risk of heart failure which did not reach statistical significance in three of the four thiazolidinedione arms compared with the sulfonylurea arms.233, 243 (SOE: Low; Sulfonylureas favored) Table 88. Observational studies comparing thiazolidinediones with sulfonylureas on congestive heart failure Author, Year Study Design Comparison Heart Failure Incidence (Sulfonylurea as Reference Group) Pantalone, 2009 233 Observational study Rosiglitazone versus sulfonylurea HR, 0.88 (95% CI, 0.60 to 1.31), p = 0.55 Pioglitazone versus sulfonylurea HR, 1.05 (95% CI, 0.77 to 1.43), p = 0.76 Hsiao, 2009 243 Observational study Rosiglitazone versus sulfonylurea HR, 1.22 (95% CI, 0.86 to 1.74), p = 0.26 Pioglitazone versus sulfonylurea HR, 1.37 (95% CI, 0.58 to 3.20), p = 0.46 CI = confidence interval; HR = hazard ratio
- 301. 244 Figure 86. Pooledodds ratio of congestive heart failure events comparing thiazolidinediones with sulfonylureas CI = confidence interval; Group 1 = sulfonylureas; Group 2 = thiazolidinediones; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Thiazolidinediones Versus DPP-4 Inhibitors One 26-week RCT compared pioglitazone with alogliptin, reporting no heart failure events in either arm.104 (SOE: Low; Neither drug favored) Sulfonylureas Versus DPP-4 Inhibitors One 58-week RCT comparing glipizide with sitagliptin reported four of 212 patients having heart failure events in the glipizide arm compared with none of 210 patients in the sitagliptin arm.107 The only rescue therapy was insulin, which was initiated in about 10 percent of participants in each arm. (SOE: Insufficient) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a Thiazolidinedione Three RCTs, lasting from 26 to 80 weeks, compared metformin alone with the combination of metformin plus a thiazolidinedione, showing a small, non-significant, greater number of heart failure events in the metformin plus thiazolidinedione arms in two of the three studies (Table 89).116, 126, 127 (SOE: Low; Metformin favored)
- 302. 245 Table 89. Randomizedcontrolled trials comparing metformin with a combination of metformin plus a thiazolidinedione on congestive heart failure Author, Year Study Design Comparison Heart Failure Incidence Leiter, 2005 116 RCT Metformin versus metformin plus rosiglitazone 0/78 versus 0/158 Borges, 2011 127 RCT Metformin versus metformin plus rosiglitazone 0/334 versus 1/344 DeFronzo, 2012 126 RCT Metformin versus metformin plus pioglitazone NR/129 versus 1/129 RCT = randomized controlled trial Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Four 26-week RCTs lasting 24 to 26 weeks compared metformin alone with the combination of metformin plus a DPP-4 inhibitor, showing no significant increased risk of heart failure in either arm.84, 126, 154, 160 Two RCTs reported no events in either arm.84, 160 One RCT reported no events in the combination arm but did not report on events in the metformin monotherapy arm.126 The third RCT reported one event in the combination arm and did not report on events in the metformin monotherapy arm.154 We combined these four RCTs in a meta-analysis, using zero events for the arms where no data were reported, and found no significant increased risk of heart failure between groups (pooled OR, 1.5; 95% CI, 0.06 to 37) (Figure 87).84, 126, 154, 160 (SOE: Low; Neither favored) Figure 87. Pooled odds ratio of congestive heart failure events comparing metformin with a combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin; Group 2 = combination of metformin plus a dipeptidyl peptidase-4 inhibitor; OR = odds ratio Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events.
- 303. 246 Metformin-Based Combination Comparisons Combinationof Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea One 24-week RCT comparing metformin plus pioglitazone with metformin plus glipizide reported two of 146 patients with heart failure events in the metformin plus pioglitazone arm and did not report whether there were any events in the 142 patients in the metformin plus glipizide arm.185 (SOE: Insufficient) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor One 26-week RCT comparing different doses of metformin plus pioglitazone with different doses of metformin plus alogliptin reported two heart failure events in the 258 patients in the metformin plus pioglitazone arms and did not report on heart failure events in the metformin plus alogliptin arms.126 (SOE: Insufficient) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor One double-blind moderately-sized 52-week RCT compared fixed dose metformin plus titration of glimepiride (mean dose 3.3 mg) with fixed dose metformin plus fixed dose saxagliptin (5 mg daily) in adults 65 years or older.193 They reported six heart failure events (1.7%) in the metformin plus glimepiride arm compared with three events (0.8%) in the metformin plus saxagliptin arm. The study had about 20 percent loss to followup in each arm. (SOE: Insufficient) Combination of Metformin Plus a Basal Insulin Versus a Combination of Metformin Plus a Premixed Insulin In a RCT that compared a combination of insulin glargine daily plus metformin with a combination of insulin lispro 75/25 plus metformin, hospitalization due to heart failure was reported in a single patient on the insulin lispro 75/25 and metformin combination.223 (SOE: Insufficient) Strength of Evidence for Congestive Heart Failure The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 90, Table 91, and Table 92 and summarized in the key points. Most studies were RCTs, although five medium-quality observational studies were included. Study limitations for all comparisons were low or medium. In general, we did not find strong differences in outcomes in the lower- versus higher-quality studies. However, many comparisons only had one or two studies, making these quality comparisons difficult. We did not find any evidence of publication bias in any of the comparisons for congestive heart failure. We also did not find any evidence of publication bias or reporting bias in the grey literature review, which would substantially alter our findings (Appendix E). Three studies reported events in one arm only; therefore, we were unable to draw firm conclusions from those studies. While this raises concerns for reporting bias, we expect arms with reporting on this outcome are likely to be the arms where more events occurred. For instance, two of the three studies reported the
- 304. 247 congestive heart failureevents in the thiazolidinedione arms. However, this inconsistent reporting remains problematic.
- 305. 248 Table 90. Strengthof evidence domains for monotherapy comparisons in terms of congestive heart failure among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. TZD (shorter studies) RCTs: 2 (170) Medium Consistent Direct Imprecise Undetected Low Neither drug arm favored Metformin vs. TZD (longer studies) RCTs: 1(4360) Medium Consistent with observational studies Direct Imprecise Undetected Low Metformin favored Obs: 2 (65,237) Medium Consistent with RCT Direct Imprecise Metformin vs. SU RCT: 1 (304) Low Consistent with observational study Direct Imprecise Undetected Low Metformin favored Obs: 1 (17,863) Medium Consistent with RCT Direct Precise Metformin vs. DPP- 4 inhibitors RCT: 1 (784) Medium Unable to determine Direct Imprecise Undetected Low Neither drug favored TZD vs. SU RCTs: 4 (11,130) Low Consistent Direct Imprecise Undetected Low SU favored Obs: 2 (116,625) Medium Consistent Direct Imprecise TZD vs. DPP-4 inhibitors RCT: 1 (655) Low Unable to determine Direct Imprecise Undetected Low Neither drug favored SU vs. DPP-4 inhibitors RCT: 1 (426) Low Unable to determine Direct Imprecise Suspected Insufficient Unable to determine DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; Obs = observational study; RCT = randomized controlled trial; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 306. 249 Table 91. Strengthof evidence domains for monotherapy versus metformin-based combination comparisons in terms of congestive heart failure among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + TZD RCT: 3 (2947) Medium Consistent Direct Imprecise Undetected Low Metformin favored Metformin vs. metformin + DPP-4 inhibitors RCT: 4 (3170) Medium Consistent Direct Imprecise Undetected Low Neither drug arm favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; RCT = randomized controlled trial; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 307. 250 Table 92. Strengthof evidence domains for metformin-based combination comparisons in terms of congestive heart failure among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + TZD vs. metformin + SU RCT: 1 (305) Medium Unable to determine Direct Imprecise Suspected ‡ Insufficient Unable to determine Metformin + TZD vs. metformin + DPP-4 inhibitors RCT: 1 (1554) Medium Unable to determine Direct Imprecise Suspected ‡ Insufficient Unable to determine Metformin + SU vs. metformin + DPP-4 inhibitors RCT: 1 (720) Low Unable to determine Direct Imprecise Undetected Insufficient Unable to determine Metformin + basal insulin vs. metformin + premixed insulin RCT: 1 (105) Medium Unable to determine Direct Imprecise Suspected ‡ Insufficient Unable to determine DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; RCT = randomized controlled trial; SGLT-2 inhibitors = sodium- glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence. ‡ Reporting bias was listed as suspected for each of these comparisons, because they did not report events in one of the study arms; however, the arm where the event was not reported is often in the drug arm where one might anticipate that there is likely to be no events.
- 308. 251 Evidence for LiverInjury Monotherapy Comparisons Metformin Versus Thiazolidinediones Three RCTs comparing metformin with thiazolidinediones reported on liver injury. Two studies compared metformin with pioglitazone,62, 70 and one study compared metformin with rosiglitazone.74 Followup and liver injury definitions varied across studies, and results were mixed. All studies targeted at least 2,000 mg daily in their metformin arms, and doses of thiazolidinediones varied (Table 93). The longer studies were of poorer quality and did not find differences in liver injury between arms. The shortest and largest RCT was a high-quality trial which used the highest doses of the drugs (metformin 2,550 daily maximum dose and pioglitazone 45 mg daily maximum dose) and found more liver injury in the metformin than the pioglitazone arm. (SOE: Low; Neither favored) Table 93. Randomized controlled trials comparing metformin with thiazolidinediones on liver injury Author, Year Study Size (Total N) Followup TZD Definition of Liver Injury Metformin Events/N (%) TZD Events/N (%) Yoon, 2011 74 349 48 weeks Rosiglitazone 5.9 mg daily (mean) Abnormal liver function not defined 0/114 (0) 1/117 (0.85%) Erem, 2014 70 60 48 weeks Pioglitazone started at 15 mg daily (most participants on ≤30 mg daily at end) Liver enzymes > 2 times ULN 0/13 (0) 0/12 (0) Schernthaner, 2004 62 1,199 26 weeks Pioglitazone started at 30 mg daily; 45 mg daily (maximum) Liver enzymes > 3 times ULN 2.2% 0.9% mg = milligrams; TZD = thiazolidinedione; ULN = upper limit normal Metformin Versus Sulfonylureas Two RCTs compared metformin with sulfonylureas and reported on liver injury.50, 74 Neither study provided a specific definition of liver injury, and both studies used sub-maximal doses of the sulfonylurea and comparable doses of metformin (titration to maximum of 2,000 mg daily). ADOPT, the study with long-term followup, found similar rates of liver injury in the two arms,50 and the other study reported more liver abnormalities in the sulfonylurea arm (Table 94).74 (SOE: Low; Neither favored)
- 309. 252 Table 94. Randomizedcontrolled trials comparing metformin with sulfonylureas on liver injury Author, Year Study Size (Total N) Followup SU Definition of Liver Injury Metformin Events/N (%) SU Events/N (%) Kahn, 2006 50 ADOPT Study 4360 Not reported* Glyburide; started 2.5 mg; maximum 15 mg Not defined NR/1341 (1.1) NR/1441 (0.8) Yoon, 2011 74 349 48 weeks Glimepiride 4.5 mg daily (mean) Abnormal liver function not defined 0/114 (0) 5/118 (4.24%); P = 0.05 ADOPT = A Diabetes Outcome Progression Trial; mg = milligrams; NR = not reported; SU = sulfonylurea *Study was 6.1 years in duration, but followup for this outcome was not reported. Thiazolidinediones Versus Sulfonylureas Three RCTs comparing thiazolidinediones with sulfonylureas reported on liver injury.50, 52, 74 Followup and liver injury definitions varied (Table 95). One study reported an non-significant increased risk of liver injury for sulfonylurea versus submaximal rosiglitazone,74 and the other two RCTs did not find substantial differences in livery injury between arms.50, 52 Of note, the highest-quality, largest and longest study reported no liver toxicity in either arm.50 (SOE: Low; Neither favored) Table 95. Randomized controlled trials comparing thiazolidinediones with sulfonylureas on liver injury Author, Year Study Size (Total N) Followup TZD SU Definition of Liver Injury TZD Events/N (%) SU Events/N (%) Kahn, 2006 50 ADOPT Study 4360 Not reported* Rosiglitazone 8 mg (maximum) Glyburide 15 mg daily (maximum) Not defined 0/1456 (0%) 0/1441 (0%) Yoon, 2011 74 349 48 weeks Rosiglitazone 5.9 mg daily (mean) Glimepiride 4.5 mg daily (mean) Abnormal liver function not defined 1/117 (0.85%) 5/118 (4.2%); P = 0.05 Tolman, 2009 52 2120 24 weeks Pioglitazone 45 mg daily (maximum) Glyburide 15 mg daily (maximum) Liver enzymes > 3 times ULN with confirmation 0/1051 (0%) 4/1046 (0.4%) P =0.06 ADOPT = A Diabetes Outcome Progression Trial; mg = milligrams; SU = sulfonylurea; TZD = thiazolidinedione; ULN = upper limit of normal *Study was 6.1 years in duration, but followup time for this outcome was not reported Sulfonylureas Versus GLP-1 Receptor Agonists One RCT examined liver injury (defined as hepatobiliary disorders) as an adverse event for this comparison.110 Seven of 132 participants treated with submaximally-dosed glibenclamide (fixed dose of 1.25 to 2.5 mg daily) developed liver injury at 52 weeks compared with 11 participants of 268 treated with liraglutide titrated to a maximum of 0.9 mg daily (5.3% versus 4.1%).110 (SOE: Low; Neither favored)
- 310. 253 Metformin Versus Metformin-BasedComparisons Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Four RCTs compared metformin with the combination of metformin plus a DPP-4 inhibitor.146, 152, 160, 164 Followup and liver injury definitions varied (Table 96). In the longest study (52 weeks), both treatments were associated with similar rates of hepatic adverse events (not specified), at a dose of metformin 1000 mg/day.164 In the shorter studies (12 to 24 weeks), which used higher doses of metformin (1500 mg/day), events were rare, and occurred slightly more often in the metformin plus DPP-4 inhibitor arms.146, 152, 160 Overall, the lack of clarity on the definition of liver injury precluded conclusions on this outcome for this comparison. (SOE: Insufficient) Table 96. Randomized controlled trials comparing metformin with metformin plus DPP-4 inhibitors on liver injury Author, Year Study Size (Total N) Followup DPP-4 Inhibitor Definition of Liver Injury Metformin Events/N (%) Metformin + DPP-4 Inhibitor Events/N (%) Haak, 2013 164 567 52 weeks Linagliptin 5 mg daily Unspecified hepatic adverse events 13 /170 (7.6%) 11/171 (6.4%) Wang, 2015 160 305 24 weeks Linagliptin 5 mg dialy Alanine transaminase increase considered to be drug-related 0/100 (0%) 1/205 (0.5%) Ross, 2012 152 491 12 weeks Linagliptin 5 mg daily Unspecified elevation of liver enzymes 0/44 (0%) 2/224 (0.9%) Yang, 2011 146 570 24 weeks Saxagliptin 5 mg daily Abnormal liver function 0/142 (0%) 1/146 (0.6%) DPP-4 = dipeptidyl peptidase-4; mg = milligrams Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor One RCT examined liver injury (defined as hepatic impairment) as an adverse event for the comparison of metformin versus a combination of metformin plus a SGLT-2 inhibitor.168 None of the 101 participants treated with metformin nor the 199 participants treated with either 5 or 10 mg of canagliflozin developed liver injury at 20 weeks. (SOE: Low; Neither favored) Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea Two RCTs compared the combination of metformin plus a thiazolidinedione with the combination of metformin plus a sulfonylurea and reported on liver injury.179, 185 One trial reported no cases of liver injury (defined as hepatic failure) in the metformin plus pioglitazone arm (0/146; 0%) and one case in the metformin plus glimepiride arm (1/142; 0.7%), at 24 weeks.185 A smaller 48-week trial reported no cases of liver injury (defined as liver enzymes
- 311. 254 values greater thanthree times the upper limit of normal) in the metformin plus rosiglitazone (0/48) or metformin plus glimepiride (0/47) arms.179 (SOE: Insufficient) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor A single RCT examined liver injury (defined as alanine aminotransferase or aspartate aminotransferase values greater than three times the upper limit of normal) as an adverse event for the comparison of a combination of metformin plus a sulfonylurea versus a combination of metformin plus a SGLT-2 inhibitor.201 Three of 482 (0.6%) participants treated with glimepiride (mean daily dose 5.6 mg) developed liver injury (alanine aminotransferase values greater than three times the upper limit of normal) at 104 weeks compared with six of 483 (1.3%) treated with 100 mg canagliflozin and seven of 485 (1.5%) treated with 300 mg canagliflozin. Two participants (0.4%) in the glimepiride arm, five participants (1.1%) in the 100 mg canagliflozin arm, and three participants (0.6%) in the 300 mg canagliflozin arm had aspartate aminotransferase values greater than three times the upper limit of normal. (SOE: Low; Neither favored) Strength of Evidence for Liver Injury We found low strength of evidence for the monotherapy comparisons for which there was evidence on liver injury and insufficient evidence for all combination therapy comparisons for this outcome (Table 97). The evidence was limited by a small number of studies with a high risk of bias based on assessment of randomization, masking, and withdrawals. Studies addressing liver injury were generally small and did not use maximal dosing of medications, especially for the non-metformin arms. Also, heterogeneity in definitions of liver injury (or lack of reporting specific definitions) limited the strength of evidence and our ability to make conclusions. A single unpublished study supported findings that neither the combination of metformin plus a sulfonylurea or the combination of metformin plus an SGLT-2 inhibitor were favored for liver injury.
- 312. 255 Table 97. Strengthof evidence domains for comparisons in terms of liver injury among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. TZD 3 (1608) High Inconsistent Indirect Imprecise Undetected Low Neither favored Metformin vs. SU 2 (4709) High Inconsistent Indirect Imprecise Undetected Low Neither favored TZD vs. SU 3 (6829) Medium Inconsistent Indirect Imprecise Undetected Low Neither favored SU vs. GLP-1 receptor agonists 1 (400) High Unknown Indirect Imprecise Undetected Low Neither favored Metformin vs. metformin + DPP-4 inhibitors 4 (1933) Low Unknown Indirect Imprecise Undetected Insufficient Unable to determine Metformin vs. metformin + SGLT-2 inhibitors 1 (299) Low Unknown Indirect Imprecise Undetected Low Neither favored Metformin + TZD vs. metformin +SU 2 (723) High Inconsistent Indirect Imprecise Undetected Insufficient Unable to determine Metformin + SU vs. metformin + SGLT-2 inhibitors (long-term study) 1 (1450) Medium Unknown Indirect Imprecise Undetected Low Neither favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 313. 256 Evidence for LacticAcidosis Monotherapy Comparisons Metformin Versus Sulfonylureas We identified two short RCTs (lasting 18 and 16 weeks) reporting the rates of lactic acidosis for metformin and sulfonylureas. These RCTs reported no cases of lactic acidosis in any of the treatment arms.130, 131 (SOE: Low; Neither favored) Metformin Versus Metformin-Based Comparisons Metformin Versus a Combination of Metformin Plus a Sulfonylurea We identified two RCTs (lasting 18 and 16 weeks) reporting the rates of lactic acidosis for metformin and the combination of metformin and a sulfonylurea. These RCTs reported no cases of lactic acidosis in any of the treatment arms.130, 131 (SOE: Low; Neither favored) Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor One 12-week RCT reported that increased lactic acid blood levels were more frequent in participants treated with metformin alone (3/100; 3%) than in those treated with the combined regimen of metformin with alogliptin (1/96; 1%); the study did not provide a statistical comparison of these rates.157 Of note, metformin doses were very small in this study (500 to 750 mg daily). (SOE: Insufficient) Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a Sulfonylurea One 24-week RCT compared the rates of lactic acidosis between the combination of metformin and pioglitazone and the combination of metformin and glimepiride. One case of lactic acidosis was reported in the 142 participants (0.7%) receiving metformin plus glimepiride and none were reported in the 146 participants (0%) who received metformin plus pioglitazone.185 Of note, the participant with lactic acidosis was noted to have had multiple serious adverse events including heart failure, liver failure, renal failure, and electrolyte disturbances. (SOE: Low; Combination of metformin plus a thiazolidinedione favored) Strength of Evidence for Lactic Acidosis Few studies addressed lactic acidosis, and evidence was of low strength or insufficient when present (Table 98). The evidence was at low or medium risk of bias, and studies were small and brief in duration; thus the evidence was imprecise and consistency unknown. One of the four studies addressing lactic acidosis only reported on elevated blood levels of lactic acidosis and not on the clinical syndrome of lactic acidosis.
- 314. 257 Table 98. Strengthof evidence domains for comparisons in terms of lactic acidosis among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. SU 2 (886) Medium Consistent Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + SU 2 (886) Medium Consistent Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + DPP-4 inhibitors 1 (288) Low Unknown Indirect Imprecise Undetected Insufficient Unable to determine Metformin + TZD vs. metformin +SU 1 (288) High Unknown Direct Imprecise Undetected Low Metformin + TZD favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 315. 258 Evidence for Pancreatitis MonotherapyComparisons Metformin Versus DPP-4 Inhibitors A 26-week trial compared metformin (1,000 mg daily; n=109), metformin (2,000 mg daily; n=111), and alogliptin (25 mg daily; n=112) and actively ascertained for pancreatitis and found no cases of pancreatitis in these arms.84 (SOE: Low; Neither favored) Metformin Versus GLP-1 Receptor Agonists A 52-week RCT that compared metformin with dulaglutide reported no cases of pancreatitis (defined as a lipase increase higher than three times the upper limit) in any of the 268 participants receiving metformin or the 269 participants receiving dulaglutide.91 (SOE: Low; Neither favored) Thiazolidinediones Versus GLP-1 Receptor Agonists A single RCT compared pioglitazone (maximum tolerated dose up to 45 mg/day) with exenatide titrated to 10 µg twice daily and reported no cases (0/136, 0%) of pancreatitis in the pioglitazone arm and a single case (1/142, 0.7%) in the exenatide arm at 48 weeks.105 Pancreatitis was not defined, and the method of ascertainment was not reported.105 (SOE: Low; Pioglitazone favored) Sulfonylureas Versus DPP-4 Inhibitors We identified one RCT comparing the incidence of pancreatitis between sulfonylurea and DPP-4 inhibitors at 52 weeks.106 There were no cases of pancreatitis in any of the 76 participants receiving glimepiride or the 151 participants receiving linagliptin. The definition of pancreatitis was unspecified. (SOE: Low; Neither favored) Sulfonylureas Versus GLP-1 Receptor Agonists Two RCTs compared sulfonylureas with GLP-1 receptor agonists and reported on pancreatitis.109, 112 One trial reported two cases of pancreatitis in the liraglutide arm (2/498; 0.4%) and no cases in the glimepiride arm (0/248; 0%) at 104 weeks.112 A 24-week trial reported no cases of pancreatitis in the liraglutide (n=272) or glibenclamide (n=139) arms.109 The criteria for a diagnosis of pancreatitis was unspecified in both studies. (SOE: Low; Sulfonylureas favored) DPP-4 Inhibitors Versus GLP-1 Receptor Agonists A 26-week RCT that compared liraglutide (n = 446) with sitagliptin (n = 219) reported no episodes of pancreatitis.210 The definition of pancreatitis was unspecified. (SOE: Low; Neither favored)
- 316. 259 Metformin Versus Metformin-BasedCombination Comparisons Metformin Versus a Combination of Metformin Plus a Sulfonylurea A single RCT with long-term followup compared the incidence of pancreatitis with metformin to the combination of metformin plus a sulfonylurea.141 There were no cases of pancreatitis, at 104 weeks of followup, among the 100 participants who received monotherapy or the 302 participants who received combined therapy. Criteria for pancreatitis was enzymatic elevation at three times the upper limit plus clinical symptoms. (SOE: Low; Neither favored) Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor We identified 11 RCTs comparing the incidence of pancreatitis with metformin with the combination of metformin plus a DDP-4 inhibitor.51, 84, 141, 151, 152, 157, 159, 160, 162, 164, 256 Definitions of pancreatitis and duration of followup differed across studies (Table 99). Four RCTs did not describe active ascertainment of pancreatitis,51, 160, 164, 256 and three RCTs had substantial losses to followup.51, 84, 141 Pancreatitis was rare, with events in only three of the 11 studies. In the study of longer duration, rates of pancreatitis were similar (0.6%) across arms at 52 weeks.159 Events were reported in only the metformin plus DPP-4 inhibitor arms in the two shorter studies.84, 162 (SOE: Low; Neither favored) Table 99. Randomized controlled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on pancreatitis Author, Year Study Size Followup DPP-4 Inhibitor Definition of Pancreatitis Metformin Events/N DPP-4 Inhibitor Events/N Bergenstal, 2012 51 666 156 weeks Sitagliptin 100 mg daily Unspecified 0/93 0/184 Ahren, 2014 141 1049 104 weeks Sitagliptin 100 mg daily Enzymes elevation > 3ULN + clinical symptoms Adjudicated 0/100 0/299 Nauck, 2014 159 1098 52 weeks Sitagliptin 100 mg daily Enzymes elevation > 3ULN + clinical symptoms Adjudicated 1/177 (0.6%) 2/315 (0.6%) Skrivanek, 2014 256 230 26 weeks Sitagliptin 100 mg daily Unspecified elevation of enzymes 0/38 0/42 Pratley, 2014 84 784 26 weeks Alogliptin 25 mg daily Unspecified elevation of enzymes 0/222 2/220 (0.9%) 1 case confirmed Seino, 2012 157 288 12 weeks Alogliptin 12.5 or 25 mg daily Unspecified 0/100 12.5 mg: 0/92 25 mg: 0/96 Haak, 2013 164 567 52 weeks Linagliptin 5 mg Clinical diagnosis 0/170 0/396 Wang, 2015 160 305 24 weeks Linagliptin 5 mg Unspecified 0/100 0/205 Ji, 2015 162 689 14 weeks Linagliptin 5 mg Unspecified 0/345 1/344 (0.3%) Ross, 2012 152 491 12 weeks Linagliptin 5 mg Unspecified 0/44 0/447 White, 2014 151 160 12 weeks Saxagliptin 5 mg Unspecified 0/78 0/66 DPP-4 = dipeptidyl peptidase-4; mg = milligrams; ULN = upper limit of normal
- 317. 260 Metformin Versus aCombination of Metformin Plus a GLP-1 Receptor Agonist Three RCTs compared the incidence of pancreatitis with metformin and the combination of metformin plus a GLP-1 receptor agonist (Table 100).141, 159, 256 The longest study actively ascertained for pancreatitis but had substantial losses to followup. In this study with 104 weeks of followup, two cases of pancreatitis were reported in the metformin plus GLP-1 receptor agonist arm (2/296, 0.7%) and none were reported in the metformin arm (0/100, 0%).141 Results from the other two RCTs, which were brief in duration, were mixed. One reported no pancreatitis in either arm at 26 weeks but did not report on active ascertainment,256 and the other reported a single case of pancreatitis in the metformin monotherapy arm and no cases of pancreatitis in the combination arm and did actively ascertain for pancreatitis.159 (SOE: Low; Metformin favored) Table 100. Randomized controlled trials comparing metformin with a combination of metformin plus a GLP-1 receptor agonist on pancreatitis Author, Year Study Size (Total N) Followup GLP-1 Receptor Agonist Definition of Pancreatitis Metformin Events/N (%) DPP-4 inhibitor Events/N Skrivanek, 2014 256 230 26 weeks Dulaglutide 0.75, 1.0, and 1.5 mg weekly Unspecified elevation of enzymes 0/38 0.75 mg: 0/21 1.0 mg: 0/10 1.5 mg: 0/25 Nauck, 2014 159 1098 52 weeks Dulaglutide 0.75 and 1.5 mg weekly Enzymes elevation > 3ULN + clinical symptoms Adjudicated 1/177 (0. 5%) 0.75 mg: 0/302 1.5 mg: 0/304 Ahren, 2014 141 1049 104 weeks Albiglutide 50 mg weekly (maximum) Enzymes elevation > 3ULN + clinical symptoms Adjudicated 0/100 2/296 (0.7%) DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; mg = milligrams; ULN = upper limit of normal Metformin-Based Combination Comparisons Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a DPP-4 Inhibitor The 26-week DURATION-2 RCT reported two cases of pancreatitis in the 165 participants who were treated with the metformin plus pioglitazone (2/165, 1.2%) combination compared with none of the 166 participants who received the metformin plus sitagliptin combination (0/166, 0%).188 Pancreatitis was not actively ascertained, and criteria for diagnosis were unspecified; this study had differential losses to followup across the arms (metformin plus thiazolidinedione, 21%; metformin plus DPP-4 inhibitor, 13%). (SOE: Low; Combination of metformin plus a DPP-4 inhibitor favored for short-term risk of pancreatitis) Combination of Metformin Plus a Thiazolidinedione Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist The DURATION-2 RCT described above had an additional arm with 160 participants who received metformin plus weekly exenatide, and none were reported to have pancreatitis during the study.188 Again, two of 165 participants had pancreatitis in the metformin plus
- 318. 261 thiazolidinedione arm; pancreatitiswas not actively ascertained, and criteria for diagnosis were unspecified. This study had large losses to followup across the arms (21% in both the metformin plus thiazolidinedione and metformin plus exenatide arms).188 (SOE: Low; Combination of metformin and GLP-1 receptor agonist favored for short-term risk of pancreatitis) Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor We identified four RCTs which compared the incidence of pancreatitis for the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor.193- 195, 197 Event rates were low in both arms across studies. Results were inconsistent across the studies of longer duration (104 weeks) and across the shorter studies (52 weeks) (Table 101, Figure 88). Only one study reported active ascertainment of pancreatitis,197 and losses to followup were substantial in all four studies.193-195, 197 (SOE: Insufficient for long-term and short- term risk) Table 101. Randomized controlled trials comparing the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor for pancreatitis Author, Year Study Size (Total N) Followup SU DPP-4 Inhibitor Definition of Pancreatitis SU Events/N (%) DPP-4 Inhibitor Events/N Del Prato, 2014 197 2620 104 weeks Glipizide 5 mg mean daily dose Alogliptin 12.5 or 25 mg arms Confirmed by laboratory and imaging tests (not defined) 3/869 (0.3%) Alogliptin 12.5 mg: 0/873 (0%) Alogliptin 25 mg: 1/878 (0.1%) Goke, 2010 195 858 52 weeks Glipizide 14.7 mg mean daily dose Saxagliptin 5 mg daily Not defined 1/428 (0.2%) 0/430 (0%) Gallwitz, 2012 194 1552 104 weeks Glimepiride 3 mg mean daily dose Linagliptin 5 mg daily Not defined 0/775 (0%) 1/776 (0.1%) Schernthaner, 2015 193 718 52 weeks Glimepiride 3.3 mg mean daily dose Saxagliptin 5 mg daily Not defined 0/359 (0%) 0/359 (0%) DPP-4 = dipeptidyl peptidase-4; mg = milligrams; SU = sulfonylurea
- 319. 262 Figure 88. Oddsratio of pancreatitis comparing the combination of metformin plus a sulfonylurea with the combination of metformin plus a DPP-4 inhibitor CI = confidence interval; DPP-4 = dipeptidyl peptidase-4; Group 1 = metformin plus a sulfonylurea; Group 2 = metformin plus a DPP-4 inhibitor; Met = metformin; OR = odds ratio; SU = sulfonylurea Boxes indicate individual study point estimates. The width of the horizontal lines represents the 95 percent confidence intervals for each study. Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist We identified two relevant RCTs.53, 204 A 104-week RCT comparing metformin plus glimepiride (N=508) versus metformin plus exenatide (N = 511) reported one case of pancreatitis in each arm (0.2% in each arm).53 A 16-week RCT comparing metformin plus glimepiride (N=231) versus metformin plus liraglutide (N = 467) reported no cases of pancreatitis in either arm.204 Studies did not report active ascertainment of pancreatitis, and the criteria for pancreatitis diagnosis were unspecified. (SOE: Low; Neither favored) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor A single RCT (N=355), with 24 weeks of followup, reported no cases of pancreatitis for metformin plus saxagliptin or metformin plus dapagliflozin.209 Pancreatitis was not defined, and the method of ascertainment was not described. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus a GLP-1 Receptor Agonist The 26-week DURATION-2 RCT reported no cases of pancreatitis in either the metformin plus weekly exenatide (n=160) or metformin plus sitagliptin arm (n=166). Pancreatitis was not actively ascertained, and criteria for diagnosis of pancreatitis were unspecified. This study had differential losses to followup across the arms (metformin plus GLP-1 receptor agonist, 21%; metformin plus DPP-4 inhibitor, 13%).188 (SOE: Low; Neither favored for short-term pancreatitis risk)
- 320. 263 Strength of Evidencefor Pancreatitis The published evidence on the comparative safety of the medications of interest was of low- strength or insufficient (Table 102, Table 103, and Table 104). The evidence was mainly limited by a lack of studies and further limited by the short duration of studies and low (expected) event rates. All evidence came from RCTs but tended to be at medium to high risk of bias, mainly because of the availability of only a few fair- to poor-quality studies for each comparison. Consistency tended to be indeterminate because of a lack of more than one study for many comparisons. All evidence was direct, although active ascertainment and definitions were not usually provided in studies. The small number of studies and their small sample sizes contributed to the evidence being imprecise for all comparisons for which we had studies. We identified unpublished studies which could have affected our grading of the evidence, but the evidence would likely only have been strengthened to a rating of low. One unpublished study confirmed the findings of the single published study that thiazolidinediones are favored over GLP-1 receptor agonists for pancreatitis. Two unpublished studies with long-term followup would have likely supported a conclusion of metformin plus a DPP-4 inhibitor being favored over metformin plus a sulfonylurea regarding long-term pancreatitis risk with low strength of evidence. One of these unpublished studies with long-term followup suggested increased risk of pancreatitis (long- term) for metformin plus a GLP-1 receptor agonist compared with metformin plus a DPP-4 inhibitor; this was in contrast to the single published study that suggested no difference in short- term risk of pancreatitis. We identified an unpublished study comparing thiazolidinediones to DPP-4 inhibitors, a comparison for which we had no published evidence.
- 321. 264 Table 102. Strengthof evidence domains for monotherapy comparisons in terms of pancreatitis among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. DPP-4 inhibitors 1 (784) Medium Unknown Direct Imprecise Undetected Low Neither treatment favored Metformin vs. GLP-1 receptor agonists 1 (495) Low Unknown Direct Imprecise Undetected Low Neither treatment favored TZD vs. GLP-1 receptor agonists 1 (278) High Unknown Direct Imprecise Suspected Low TZD favored SU vs. DPP-4 inhibitors 1 (227) Low Unknown Direct Imprecise Undetected Low Neither treatment favored SU vs. GLP-1 receptor agonists 2 (1210) Medium Inconsistent Direct Imprecise Undetected Low SU favored DPP-4 inhibitors vs. GLP-1 receptor agonists 1 (661) Medium Unknown Direct Imprecise Undetected Low Neither treatment favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating the outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 322. 265 Table 103. Strengthof evidence domains for metformin monotherapy versus metformin-based combination comparisons in terms of pancreatitis among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + SU (long-term study) 1 (1049) Medium Unknown Direct Imprecise Undetected Low Neither treatment favored Metformin vs. metformin + DPP-4 inhibitors (long-term and short-term studies) 11 (6327) Medium Consistent Direct Imprecise Undetected Low Neither treatment favored for long-term or short-term pancreatitis risk Metformin vs. metformin + GLP-1 receptor agonists 3 (2377) Medium Inconsistent Direct Imprecise Undetected Low Metformin favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating the outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 323. 266 Table 104. Strengthof evidence domains for metformin-based combination comparisons in terms of pancreatitis among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin + TZD vs. metformin + DPP-4 inhibitors 1 (491) Medium Unknown Direct Imprecise Undetected Low Metformin + DPP-4 inhibitor favored for short- term risk of pancreatitis Metformin + TZD vs. metformin + GLP-1 receptor agonists 1 (491) Medium Unknown Direct Imprecise Undetected Low Metformin + GLP-1 receptor agonist favored for short-term risk of pancreatitis Metformin + SU vs. metformin + DPP-4 inhibitors (longer duration studies) 2 (4172) High Inconsistent Direct Imprecise Suspected Insufficient Unable to determine for long-term risk of pancreatitis Metformin + SU vs. metformin + DPP-4 inhibitors (shorter duration study) 2 (1576) High Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Metformin + SU vs. metformin + GLP-1 receptor agonists 2 (2481) High Consistent Direct Imprecise Undetected Low Neither treatment favored Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors (shorter duration study) 1 (355) Low Unknown Direct Imprecise Undetected Low Neither treatment favored Metformin + DPP-4 inhibitors vs. metformin + GLP-1 receptor agonists (short duration study) 1 (491) Medium Unknown Direct Imprecise Undetected Low Neither treatment favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating the outcome. † Unless otherwise specified, the estimates are the pooled odds ratio (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 324. 267 Evidence for SevereAllergic Reactions Monotherapy Comparisons Metformin Versus GLP-1 Receptor Agonists A single 52-week RCT (N=495) that compared metformin with dulaglutide reported no systemic hypersensitivity reaction in either arm.91 (SOE: Low; Neither favored) Thiazolidinediones Versus GLP-1 Receptor Agonists A single RCT (N=278) that compared pioglitazone with exenatide reported no systemic hypersensitivity reaction in either arm at 48 weeks of followup.105 (SOE: Low; Neither favored) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus a DPP-4 Inhibitor Four RCTs compared the incidence of severe allergic reactions between metformin and the combination of metformin plus a DDP-4 inhibitor (Table 105).146, 151, 152, 164 Heterogeneity in definitions of severe allergic reactions and duration of followup precluded a meta-analysis. Three of the four RCTs reported slightly higher rates of hypersensitivity reaction events in the metformin plus DPP-4 inhibitor versus the metformin monotherapy arms. The longest RCT was a 54-week extension study164 in which the arms included participants from the initial 6-month study86 and participants who were re-randomized for the extension study. Among participants who were newly randomized for the 54-week extension study, hypersensitivity reactions occurred in 0 percent of the metformin 2000 mg arm, 0 percent of the metformin 1000 mg plus linagliptin 5 mg arm, and in 1.7 percent of the metformin 2000 mg plus linagliptin 5 mg arm. (SOE: Low; Metformin favored)
- 325. 268 Table 105. Randomizedcontrolled trials comparing metformin with a combination of metformin plus a DPP-4 inhibitor on severe allergic reactions Author, Year Sample Size (Total N) Followup DPP-4 Inhibitor Definition of Severe Allergic Reaction Active Asc. Metformin Events/N (%) DPP-4 Inhibitor Events/N (%) Haak, 2013 164 567 54 weeks Linagliptin 5 mg Hypersensitivity reactions (e.g., angioedema, anaphylaxis) Severe cutaneous reactions Yes Yes 1/170 (0.6%)* 0/170 2/171 (1.2%)* 0/171 White, 2014 151 160 12 weeks Saxagliptin 5 mg Hypersensitivity reactions NR 0/78 0/66 Yang, 2011 146 570 24 weeks Saxagliptin 5 mg Hypersensitivity reactions Yes 0/287 3/283 (1.1%) Ross, 2012 152 491 12 weeks Linagliptin Hypersensitivity reactions (angioedema, anaphylaxis, angioedema-like) Yes 0/44 5 mg: 1/224 (0.4%) 2.5 mg: 0/223 Asc = ascertainment; DPP-4 = dipeptidyl peptidase-4; mg = milligrams * Data are shown for the metformin 2000 mg monotherapy arm and the metformin 2000 mg plus linagliptin 5 mg combination arm; there were no events reported in the metformin 1000 mg plus linagliptin combination arm. Metformin-Based Combination Comparisons Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus a DPP-4 Inhibitor A single 52-week RCT compared metformin plus glipizide with metformin plus saxagliptin and reported on hypersensitivity adverse events.195 The authors reported a hypersensitivity adverse event in one participant in the metformin plus saxagliptin arm and in two participants in the metformin plus glipizide arm. One of the events in the metformin plus glipizide arm was noted to be related to ciprofloxacin. This study did not provide information on the method of ascertainment or definition of hypersensitivity. This RCT also had high rates of discontinuation based partly on increasingly strict glycemic control criteria for maintaining eligibility in the study.195 (SOE: Low; Neither favored) Strength of Evidence for Severe Allergic Reactions We identified evidence for three comparisons for the outcome of allergic reactions (Table 106). The published studies were on comparisons that included GLP-1 receptor agonists and DPP-4 inhibitors. All evidence was low or insufficient for this outcome. Because of the limited numbers of studies and their samples sizes, evidence was imprecise. We did not detect reporting bias, but our assessment of this was also limited by the small number of studies.
- 326. 269 Table 106. Strengthof evidence domains for comparisons in terms of severe allergic reactions among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. GLP-1 receptor agonists 1 (495) Low Unknown Direct Imprecise Undetected Low Neither favored TZD vs. GLP-1 receptor agonists 1 (278) Medium Unknown Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + DPP- 4 inhibitors 4 (1788) Low Consistent Direct Imprecise Undetected Low Metformin favored Metformin + SU vs. metformin + DPP-4 inhibitors 1 (858) High Unknown Direct Imprecise Undetected Low Neither favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating the outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 327. 270 Evidence for MacularEdema or Decreased Vision Monotherapy Comparisons Thiazolidinediones Versus GLP-1 Receptor Agonists A single RCT examined blurred vision.105 Three of 136 participants (2.2%) treated with pioglitazone and two of 142 participants (1.4%) treated with exenatide had blurred vision over 48 weeks of followup. (SOE: Low; Neither favored) Sulfonylureas Versus GLP-1 Receptor Agonists One 104-week RCT compared glimepiride with liraglutide at two different doses (1.2 and 1.8 mg) and reported on decreased vision.113 At the end of the study, the incidence of decreased vision was comparable in all arms, with 7 percent of the glimepiride participants (n=248) having decreased vision compared with 6 percent of the liraglutide participants (n=251 for liraglutide at 1.2 mg and n=247 for liraglutide at 1.8 mg). (SOE: Low; Neither favored) Strength of Evidence for Macular Edema or Decreased Vision We identified only two studies for the outcomes of macular edema or decreased vision which evaluated different comparisons. Therefore, the evidence on these outcomes was insufficient because of a lack of studies (Table 107).
- 328. 271 Table 107. Strengthof evidence domains for monotherapy comparisons in terms of macular edema or decreased vision among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † TZD vs. GLP-1 receptor agonists 1 (278) Medium Unknown Direct Imprecise Undetected Low Neither favored SU vs. GLP-1 receptor agonists (long-term study) 1 (746) Medium Unknown Direct Imprecise Undetected Low Neither favored GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the safety outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 329. 272 Evidence for UrinaryTract Infections Monotherapy Comparisons Metformin Versus SGLT-2 Inhibitors Three short RCTs (published in two articles), 12 to 24 weeks in duration, compared dapagliflozin with metformin and reported on UTIs.88, 89 Since ORs did not appear to vary by gender in these short-term studies, we present results for men and women combined. We found significant statistical heterogeneity using a random effects meta-analysis (pooled OR, 1.54; 95% CI, 0.56 to 4.22; I-squared, 61.1%). Exclusion of any one study did not change the inference of the meta-analysis. We found similar non-significant increased odds for SGLT-2 inhibitors versus metformin for UTIs (pooled OR, 1.5; 95% CI, 0.5 to 5.0 (Figure 89) using the profile likelihood method. One of the RCTs used a lower dose of dapagliflozin of 5 mg88 relative to 10 mg in the other two RCTs. We did not include one RCT in the meta-analysis because it was longer (78 weeks) than the other three studies.90 This study compared metformin to 10 mg of empagliflozin and 25 mg of empagliflozin and reported similar overall incidences of UTIs with both doses of empagliflozin. UTI rates among men receiving 25 mg of empagliflozin (4/57; 7.0%) were non-significantly higher than among those receiving 10 mg of empagliflozin (0/49; 0%) and metformin (0/28; 0%) and approached UTI rates among women [empagliflozin 10 mg: 4/57 (7.0%), empagliflozin 25 mg: 3/52 (5.8%), and metformin: 2/28 (7.1%)].90 (SOE: Low; Neither favored) Figure 89. Pooled odds ratio of urinary tract infections comparing metformin with SGLT-2 inhibitors CI = confidence interval; Group 1 = metformin; Group 2 = sodium-glucose co-transporter-2 inhibitors; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate.
- 330. 273 DPP-4 Inhibitors VersusSGLT-2 Inhibitors One 24-week RCT compared 100 mg of sitagliptin with 10 mg and 25 mg of empagliflozin and reported UTI events separately by gender.114 UTI occurrences among men were similar in those receiving sitagliptin (4/141; 3%) versus 10 mg of empagliflozin (3/142; 2%) and 25 mg of empagliflozin (2/144; 1%).114 UTI events were non-significantly lower in women receiving sitagliptin (7/82; 9%) versus 10 mg of empagliflozin (12/82; 15%) and 25 mg of empagliflozin (10/78; 13%). The study did not test for an interaction by gender. (SOE: Low; Neither favored in men, DPP-4 inhibitors favored in women) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Nine short-term RCTs (published in six articles) compared metformin with the combination of metformin plus an SGLT-2 inhibitor and showed similar rates of UTIs for the combination compared to metformin (pooled OR, 1.2; 95% CI, 0.7 to 1.9) (Figure 90).88, 153, 156, 165, 166, 168 No single study markedly influenced the results, and we did not find evidence of substantial statistical heterogeneity (I2 = 0.0%). One study also reported that no events of urosepsis or pyelonephritis occurred in either arm.166 The definitions for UTIs varied across studies (Table 108). Table 108. Definitions of urinary tract infections used in randomized controlled trials comparing metformin with a combination of metformin and SGLT-2 inhibitor Author, Year Definition of UTI Outcome (Actively Ascertained Unless Otherwise Noted) Bailey, 2013 170 UTI (does not include events suggestive of UTI) Rosenstock, 2012 156 UTI, not otherwise specified Henry, 2012 88 Events suggestive of UTI Bolinder, 2014 267 MedDRA definition of UTI Rosenstock, 2013 153 UTI, not otherwise specified Qiu, 2014 165 UTI, not otherwise specified Haring, 2014 (a) 166 MedDRA definition for UTI Henry, 2012 (b) 88 Based on a predefined list of signs, symptoms and other events suggestive of UTI Schumm-Draeger, 2015 168 MedDRA definition of UTI MedDRA = Medical Dictionary for Regulatory Activities; UTI = urinary tract infection Based on three studies providing gender-stratified results, meta-analyses stratified by gender showed that women had non-significantly increased odds of UTIs for the combination of metformin plus an SGLT-2 inhibitor versus metformin (pooled OR, 1.4; 95% CI, 0.8 to 2.3) and no difference in UTI odds between treatment groups for men (pooled OR, 1.0; 95% CI, 0.4 to 2.8).88, 166
- 331. 274 Figure 90. Pooledodds ratio of short-term risk of urinary tract infections comparing metformin with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. We did not include two moderately-sized RCTs (total sample size, 182 to 546) in the meta- analysis because they were longer (102 weeks).170, 267 One compared metformin with metformin plus dapagliflozin and reported UTI rates of 5.8 percent and 11.9 percent in the metformin and metformin plus dapagliflozin arms, respectively.170 The other had similar UTI rates across arms (metformin, 4.4% and dapagliflozin, 3.3%).267 Of note, the former had very high losses to followup with 47 percent losses in the metformin arm compared with 30 percent to 40 percent in the other arms.170 (SOE: Low; Neither favored) Metformin-Based Combination Comparisons Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three RCTs compared metformin plus a sulfonylurea with metformin plus an SGLT-2 inhibitor and reported inconsistent results regarding UTIs (Table 109).54, 200, 201 We did not combine these studies in a meta-analysis because of the heterogeneity in the definition of UTI and study durations. Two 104-week RCTs compared metformin plus glimepiride to metformin plus an SGLT-2 inhibitor and found similar incidences of UTIs across arms.200, 201 A 208-week RCT reported more UTIs in the metformin plus dapagliflozin arm compared to the metformin plus glipizide arm.54 This study reported high withdrawal rates of greater than 60 percent in each
- 332. 275 arm.54 UTI rates werelower among men compared to women when reported by gender.54, 200 (SOE: Low; Neither favored) Table 109. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor on urinary tract infections Author, Year Definition of UTI Outcome* Results Events/N (%) Leiter, 2015 201 Cystitis, pyelonephritis chronic, and UTI 198 Metformin + glimepiride: 33/482 (7%) Metformin + canagliflozin 100 mg: 51/483 (11%) Metformin + canagliflozin 300 mg: 42/485 (9%) Del Prato, 2015 54 Confirmed UTI, not otherwise defined Female Metformin + glipizide: 25/408 (13.5%) Metformin + dapagliflozin: 35/406 (19.4%) Male Metformin + glipizide: 13/408 (5.8%) Metformin + dapagliflozin: 20/406 (8.8%) Ridderstrale, 2014 200 MedDRA definition of UTI (passive ascertainment) Female Metformin + glimepiride: 81/359 (23%) Metformin + empagliflozin: 74/333 (22%) Male Metformin + glimepiride: 21/421 (5%) Metformin + empagliflozin: 31/432 (7%) MedDRA = Medical Dictionary for Regulatory Activities; mg = milligrams; UTI = urinary tract infections * Outcomes are actively ascertained unless otherwise noted. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Five RCTs compared metformin plus DPP-4 with metformin plus an SGLT-2 inhibitor and reported inconsistent results (Table 110). We did not combine these studies because of differences in study duration and dosing.90, 153, 156, 158, 209 The longest RCT (78 weeks) was of low quality and reported UTI rates stratified by sex. Among women, UTI rates were higher for the metformin plus empagliflozin 25 mg arm relative to the other arms; among men, UTI rates were highest in the metformin plus sitagliptin arm relative to the other arms.90 A medium-quality, 52-week RCT compared metformin plus sitagliptin to metformin plus canagliflozin at doses of 100 mg and 300 mg and reported slightly lower UTI rates in the highest-dose (300 mg) canagliflozin arm and slightly higher rates in the lower-dose (100 mg) canagliflozin arm.158 Three short-term studies also conflicted. One medium-quality study reported a higher UTI rate with metformin plus 200 mg of canagliflozin compared with the other arms.156 Two high- quality short-term RCTs found similar UTI rates in the metformin plus DPP-4 inhibitor arms and metformin plus SGLT-2 inhibitor arms.153, 209 The evidence was graded as being at medium risk of bias because of failure to report clearly on the randomization scheme. (SOE: Low; Neither favored)
- 333. 276 Table 110. Randomizedcontrolled trials comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor on urinary tract infections Author, Year Followup Definition of UTI Outcome* Results Events/N (%) Rosenstock, 2012 156 12 weeks UTI, not otherwise specified Metformin + sitagliptin 100 mg: 1/65 (2%) Metformin + canagliflozin 100 mg: 2/64 (3%) Metformin + canagliflozin 200 mg: 6/65 (9%) Metformin + canagliflozin 300 mg: 2/64 (3%) Rosenstock, 2013 153 12 weeks UTIs, including cystitis, excluding signs and symptoms Metformin + sitagliptin 100 mg: 4.2% Metformin + empagliflozin 10 mg: 4.2% Metformin + empagliflozin 25 mg: 5.7% Rosenstock, 2015 209 24 weeks UTI, not otherwise specified Metformin + saxagliptin 5 mg: 9/176 (5%) Metformin + dapagliflozin 10 mg: 7/179 (5%) Lavalle-Gonzalez, 2013 158 52 weeks UTI, not otherwise specified Metformin + sitagliptin 100 mg: 23/366 (6.3%) Metformin + canagliflozin 100 mg: 29/368 (7.9%) Metformin + canagliflozin 300 mg: 18/367 (4.9%) Ferrannini, 2013 90 78 weeks MedDRA definition of UTI Female Metformin + sitagliptin 100 mg: 4/27 (14.8%) Metformin + empagliflozin 10 mg: 13/83 (15.7%) Metformin + empagliflozin 25 mg: 18/78 (23.1%) Male Metformin + sitagliptin 100 mg: 3/29 (10.3%) Metformin + empagliflozin 10 mg: 2/83 (2.4%) Metformin + empagliflozin 25 mg: 3/88 (3.4%) MedDRA = Medical Dictionary for Regulatory Activities; mg = milligrams; UTI = urinary tract infections * Outcomes are actively ascertained unless otherwise noted. Strength of Evidence for Urinary Tract Infections The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 111 and summarized in the key points. All studies were RCTs. Study limitations for all the comparisons were low or medium. Although evidence of gender differences in UTI rates was limited, the data suggest that there may be higher rates of UTIs among females (particularly noted in the comparisons of metformin versus metformin plus SGLT-2 inhibitor). In general, we did not find strong differences in outcomes in the lower- versus higher-quality studies. We did not find any evidence of publication bias in any of the comparisons for UTI. The grey literature was consistent with our findings in the metformin plus SU vs. metformin plus SGLT-2 inhibitor comparison, with one unpublished study that found no UTIs in either arm.
- 334. 277 Table 111. Strengthof evidence domains for monotherapy and metformin-based combination comparisons in terms of urinary tract infections among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. SGLT-2 inhibitors 4 (2,292) Medium Inconsistent Direct Imprecise Undetected Low Neither favored for short- term UTI risk DPP-4 inhibitors vs. SGLT-2 inhibitors 1 (899) Low Unknown Direct Imprecise Undetected Low Neither favored in men DPP-4 inhibitors favored in women Metformin vs. metformin + SGLT-2 inhibitors 9 (4,035) Low Consistent Direct Imprecise Undetected Low Neither favored for short- term UTI risk Metformin + SU vs. metformin + SGLT-2 inhibitors (longer studies) 3 (3,815) Low Inconsistent Direct Imprecise Undetected Low Neither favored for long- term risk of UTI Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors 5 (3,423) Medium Inconsistent Direct Precise Undetected Low Neither favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 335. 278 Evidence for ImpairedRenal Function Monotherapy Comparisons Metformin Versus SGLT-2 Inhibitors Three RCTs compared metformin with SGLT-2 inhibitors and reported on impaired renal function (Table 112).89, 90, 239 We did not combine the results of these RCTs in a meta-analysis because they varied greatly in their definitions of impaired renal function. Two trials evaluated the change in estimated glomerular filtration rate (eGFR) and found no substantial differences between the arms.89, 90 One 12-week trial (N=408) evaluated incident microalbuminuria, change in the urine albumin-to-creatinine ratio, and incident diabetic nephropathy.239 The investigators did not find substantial differences in urine albumin-to-creatinine ratio across arms, and those in the low-dose empagliflozin arm had more incident microalbuminuria and diabetic nephropathy compared with the metformin and high-dose empagliflozin arm. (SOE: Low; Neither favored) Table 112. Randomized controlled trials comparing metformin with SGLT-2 inhibitors on impaired renal function Author, Year Followup Definition of Impaired Renal Function* Results List, 2009 89 12 weeks eGFR NR for any arm Qualitative statement of no difference across groups Ferrannini, 2013 90 90 weeks eGFR (ml/min/1.73 m 2 ) Metformin vs. empagliflozin 10 mg: between-group difference from baseline to final, 0.13 (95% CI, -4.5 to 4.7) Metformin vs. empagliflozin 25 mg: between-group difference from baseline to final, 2.66 (95% CI, -1.8 to 7.1) Ferrannini, 2013 239 12 weeks Microalbuminuria, not further defined Metformin: 1.3% Empagliflozin 10 mg: 3.7% Empagliflozin 25 mg: 0% 12 weeks Diabetic nephropathy (unclear if actively ascertained) Metformin: 1.3% Empagliflozin 10 mg: 2.5% Empagliflozin 25 mg: 1.2% 12 weeks Urinary albumin to creatinine ratio (mg/mmol) Metformin vs. empagliflozin 10 mg: between-group difference from baseline to final, 0.08 mg/mmol Metformin vs. empagliflozin 25 mg: between-group difference 0.6 mg/mmol CI = confidence interval; eGFR = estimated glomerular filtration rate; mg = milligrams; mg/mmol = milligrams per millimole; mil/min*1.73 m2 = milliliters per minute per 1.73 meters squared; NR = not reported * Outcomes are actively ascertained unless otherwise noted. Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Six RCTs comparing metformin with a combination of metformin plus SGLT-2 inhibitor reported on albuminuria and eGFR (Table 113).153, 165, 166, 168, 170, 267
- 336. 279 Two 102-week trialsevaluated renal impairment or failure as a categorical outcome, with conflicting results.170, 267 One found slightly more events of renal impairment in the metformin plus SGLT-2 inhibitor arm compared to metformin (3.3% vs. 0%),267 and the other RCT did not find a clear pattern of differences across arms.170 A 16-week RCT also evaluated renal impairment or failure events and reported similar rates of events across arms both arms.168 Three short-term trials evaluated eGFR and found no meaningful differences between the metformin and metformin plus SGLT-2 inhibitor arms.165, 166, 168 One 12-week RCT compared metformin with metformin plus empagliflozin and stated qualitatively that treatment with empagliflozin did not significantly change creatinine clearance or urine albumin compared with metformin.153 (SOE: Low; Neither favored) Table 113. Randomized controlled trials comparing metformin with a combination of metformin plus an SGLT-2 inhibitor on impaired renal function Author, Year Followup Definition of Impaired Renal Function* Results Rosenstock, 2013 153 12 weeks Creatinine clearance, microalbuminuria NR for any arm Qualitative statement of no difference across groups Schumm-Draeger, 2015 168 16 weeks Renal impairment/failure as specified in protocol Metformin: 4/101 (4%) Metformin + dapagliflozin 5 mg twice daily: 3/100 (3%) Metformin + dapagliflozin 10 mg daily: 3/99 (3%) Schumm-Draeger, 2015 168 16 weeks eGFR (ml/min/1.73 m 2 ) Metformin vs. metformin + dapagliflozin 5 mg twice daily: between-group difference from baseline to final, 4.0 (95%CI, 1.6 to 4.4) Metformin vs. metformin + dapagliflozin 10 mg: between-group difference from baseline to final, 0.8 (95%CI, -1.5 to 4.1) Haring, 2014 166 24 weeks eGFR (ml/min/1.73 m 2 ) Metformin vs. metformin + empagliflozin 10 mg: between-group difference from baseline to final, 0.9 (95% CI, -1.5 to 3.3) Metformin vs. metformin + empagliflozin 25 mg: between-group difference from baseline to final, 2.7 (95% CI, 0.6 to 4.8) Qiu, 2014 165 18 weeks eGFR % reduction in eGFR Metformin: 0.3% Metformin + canagliflozin 50 mg twice daily: 0.7% Metformin + canagliflozin 150 mg twice daily: 3.8% Bailey, 2013 170 102 weeks Renal impairment or failure, not otherwise specified Metformin: 2/137 (1.5%) Metformin + dapagliflozin 5 mg: 4/137 (2.9%) Metformin + dapagliflozin 10 mg: 2/135 (1.5%) Bolinder, 2014 267 102 weeks MedDRA definition of renal impairment, renal failure Metformin: 0/91 (0%) Metformin + dapagliflozin 10 mg: 3/91 (3.3%) CI = confidence interval; eGFR = estimated glomerular filtration rate; MedDRA = Medical Dictionary for Regulatory Activities; mg = milligrams; mil/min*1.73 m2 = milliliters per minute per 1.73 meters squared; NR = not reported * Outcomes are actively ascertained unless otherwise noted.
- 337. 280 Metformin-Based Combination Comparisons Combinationof Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three RCTs with long-term followup compared the effects of metformin plus a sulfonylurea to metformin plus a SGLT-2 inhibitor on renal impairment or failure, changes in eGFR, and albuminuria (Table 114).54, 200, 201 One 208-week trial evaluated reduced creatinine clearance and renal impairment and found similar rates of events between arms.54 Three trials evaluated changes in eGFR and found no meaningful differences in eGFR changes across arms.54, 200, 201 Two trials evaluated albuminuria and found no differences between arms.200, 201 (SOE: Low; Neither favored)
- 338. 281 Table 114. Randomizedcontrolled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor on impaired renal function Author, Year Followup Definition of Impaired Renal Function* Results Notes Leiter, 2015 201 104 weeks eGFR (ml/min/1.73 m 2 ) Metformin + glimepiride vs. metformin + canagliflozin 100 mg: between-group difference from baseline to final, -2.4 (95%CI, -6.3 to 1.5) Metformin + glimepiride vs. metformin + canagliflozin 300 mg: between-group difference from baseline to final, -4.2 (95%CI, -8.1 to -0.3) Subjects meeting eGFR withdrawal criteria: n=6 in glimepiride arm, n=5 in canagliflozin 100 mg arm, and n=6 for canagliflozin 300 mg arm Unclear if included in analysis Ridderstrale, 2014 200 104 weeks eGFR (ml/min/1.73 m 2 ) Metformin + glipizide vs. metformin + empagliflozin 25 mg; between-group difference, 3.5 (95% CI, 2.2 to 4.8) Del Prato, 2015 54 208 weeks MedDRA defined decreased eGFR Metformin + glipizide: 4/408 (1.0 %) Metformin + dapagliflozin: 2/406 (0.5 %) Del Prato, 2015 54 208 weeks MedDRA defined reduced creatinine clearance Metformin + glipizide: 13/408 (3.2 %) Metformin + dapagliflozin: 20/406 (4.9%) Del Prato, 2015 54 208 weeks MedDRA defined renal impairment Metformin + glipizide: 11/408 (2.7 %) Metformin + dapagliflozin: 10/406 (2.5%) Leiter, 2015 201 104 weeks Renal failure leading to medication discontinuation Metformin + glimepiride: NR Metformin + canagliflozin 100 mg: NR Metformin + canagliflozin 300 mg: 3/385 (0.6%) Leiter, 2015 201 104 weeks Urine albumin-to- creatinine ratio (mg/g) Metformin + glimepiride vs. metformin + canagliflozin 100 mg: between-group difference, 13.9 mg/g Metformin plus glimepiride vs. metformin plus canagliflozin 300 mg: between- group difference, 16.1 mg/g Ridderstrale, 2014 200 104 weeks Urine albumin-to- creatinine ratio (mg/g) Metformin + glimepiride vs. metformin + empagliflozin 25 mg: between-group difference, 1.9 mg/g (95% CI, -5.1 to 8.9 mg/g) Subgroup with no albuminuria at baseline CI = confidence interval; eGFR = estimated glomerular filtration rate; MedDRA = Medical Dictionary for Regulatory Activities; mg = milligrams; mg/g = milligrams per gram; mil/min*1.73 m2 = milliliters per minute per 1.73 meters squared; NR = not reported * Outcomes are actively ascertained unless otherwise noted. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Four RCTs compared metformin plus a DPP-4 inhibitor with metformin plus a SGLT-2 inhibitor on renal outcomes (Table 115).90, 153, 158, 209 Three trials evaluated changes in eGFR and found no substantial differences across arms.90, 158, 209 One 12-week RCT evaluated changes in creatinine clearance and microalbuminuria comparing metformin plus sitagliptin and metformin
- 339. 282 plus empagliflozin andfound no significant differences between arms.153 (SOE: Low; Neither favored) Table 115. Randomized controlled trials comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor on impaired renal function Author, Year Followup Definition of Impaired Renal Function* Results Ferrannini, 2013 90 90 weeks eGFR (ml/min/1.73 m 2 ) Metformin + sitagliptin vs. metformin + empagliflozin 10 mg: between-group difference, 4.1 (95% CI, 0.3 to 8.6) Metformin + sitagliptin vs. metformin + empagliflozin 25 mg: between-group difference, 2.8 (95% CI, -1.5 to 7.1) Lavalle-Gonzalez, 2013 158 52 weeks Decreased eGFR Metformin + sitagliptin vs. metformin + canagliflozin 100 mg: between-group difference, 1.0% Metformin + sitagliptin vs. metformin + canagliflozin 300 mg: between-group difference, 0.9% Rosenstock, 2013 153 12 weeks Creatinine clearance, microalbuminuria NR for any arm Qualitative statement of no difference across groups Rosenstock, 2015 209 24 weeks GFR decrease, not otherwise defined Metformin + saxagliptin: 1/176 (0.6%) Metformin + dapagliflozin: 0/179 (0%) CI = confidence interval; eGFR = estimated glomerular filtration rate; mg = milligrams; mil/min*1.73 m2 = milliliters per minute per 1.73 meters squared; NR = not reported * Outcomes are actively ascertained unless otherwise noted. Strength of Evidence for Impaired Renal Function The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 116 and summarized in the key points. All studies were RCTs. Study limitations for all the comparisons were low or medium. The evidence was generally imprecise because of small event rates and sample sizes. In general, we did not find strong differences in outcomes in the lower- versus higher-quality studies. We did not find any evidence of publication bias in any of the comparisons for renal outcomes. We also did not find any evidence of publication bias or reporting bias in the grey literature review.
- 340. 283 Table 116. Strengthof evidence domains for monotherapy and metformin-based combination comparisons in terms of impaired renal function among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. SGLT-2 inhibitors 3 (1,456) High Consistent Direct Imprecise Undetected Low Neither favored DPP-4 inhibitors vs. SGLT- 2 inhibitors 2 (1,394) Medium Consistent Direct Precise Undetected Low Neither favored Metformin vs. metformin + SGLT-2 inhibitors 6 (2,340) Low Consistent Direct Imprecise Undetected Low Neither favored Metformin + SU vs. metformin + SGLT-2 inhibitors (longer studies) 3 (3,815) Medium Inconsistent Direct Imprecise Undetected Low Neither favored Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors 4 (2,972) Low Consistent Direct Imprecise Undetected Low Neither favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 341. 284 Evidence for GenitalMycotic Infections Monotherapy Comparisons Metformin Versus SGLT-2 Inhibitors Three medium- to high-quality, short RCTs (reported in two articles) compared metformin with SGLT-2 inhibitors and found more genital infections in the SGLT-2 inhibitor vs. metformin arms (pooled OR, 4.1; 95% CI, 2.0 to 8.3) (Figure 91).88, 89 ORs did appear to vary by gender. No single study markedly influenced the results, and we did not find significant statistical heterogeneity (I2 = 0.0%). Figure 91. Pooled odds ratio of genital or mycotic infections comparing metformin with SGLT-2 inhibitors CI = confidence interval; Group 1 = metformin; Group 2 = sodium-glucose co-transporter-2 inhibitors; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. We did not include a low-quality, 78-week RCT in the meta-analysis because of its longer duration.90 This study compared metformin with empagliflozin and reported slightly higher rates of genital infections among females for SGLT-2 inhibitor therapy [1/28 (3.6%) with metformin versus 3/57 (5.3%) with metformin plus low-dose empagliflozin and 3/52 (5.8%) with metformin plus high-dose empagliflozin] and more genital infections among males with SGLT-2 inhibitors [0/28 (0%) with metformin versus 2/49 (4.1%) with metformin plus low dose empagliflozin and 3/57 (5.3%) with metformin plus high- dose empagliflozin]. (SOE: Moderate; Metformin favored) DPP-4 Inhibitors Versus SGLT-2 Inhibitors Two RCTs (24 to 26 weeks) compared outcomes from use of 100 mg of sitagliptin daily to an SGLT-2 inhibitor, by gender.114, 240 Both trials reported higher rates of genital infections among both women and men with use of SGLT-2 inhibitors compared with sitagliptin, with
- 342. 285 some of thecomparisons statistically significant (Table 117).114, 240 (SOE: Low; DPP-4 inhibitors favored) Table 117. Randomized controlled trials comparing DPP-4 inhibitors with SGLT-2 inhibitors on genital infections Author, Year Medication Dose Women Events/N (%) Men Events/N (%) Stenlof, 2014 240 Sitagliptin 100 mg Canagliflozin 100 mg Canagliflozin 300 mg 1/155 (1.2) 3/170 (3) 3/170 (3.2) 0/155 (0) 3/170 (3) 5/170 (6.5) Roden, 2013 114 Sitagliptin 100 mg Empagliflozin 10 mg Empagliflozin 25 mg 1/82 (1) 3/82 (4) 7/79 (9) 1/141 (1) 4/142 (3) 2/144 (1) mg = milligrams Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Seven RCTs (reported in six articles) compared metformin with metformin plus a SGLT-2 inhibitor and found increased odds of genital infections for combination therapy over metformin monotherapy with no clear differences by gender: pooled OR, 3.0 (95% CI, 1.2 to 7.2) for women and pooled OR, 2.7 (95% CI, 0.8 to 9.0) for men (Figure 92).88, 156, 165, 166, 168 No single study markedly influenced the results, and we did not find significant statistical heterogeneity (I2 = 15.4% for women and I2 = 0.0% for men). An additional 12-week RCT did not provide sex- stratified analyses so was not included in the meta-analysis.153 This study reported more genital infection events in one of the groups receiving empagliflozin compared with the other two arms (metformin plus sitagliptin 100 mg: 0%; metformin plus empagliflozin 10 mg: 9.9%; metformin plus empagliflozin 25 mg: 0%).
- 343. 286 Figure 92. Pooledodds ratio of genital or mycotic infections comparing metformin with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events. We excluded two RCTs from the meta-analysis because of their longer durations.170, 267 The results of those RCTs are included in Table 118 and are consistent with the findings from the short-term studies. (SOE: High; Metformin favored)
- 344. 287 Table 118. Randomizedcontrolled trials comparing metformin with a combination of metformin plus an SGLT-2 inhibitor on genital infections Author, Year Followup Genital Infection Outcome* Results (Metformin Versus Metformin + SGLT-2 Inhibitor) Rosenstock, 2012 156 12 weeks Symptomatic of genital infections Metformin: 1/65 (2%) Metformin + canagliflozin 100 mg: 4/64 (6%) Metformin + canagliflozin 200 mg: 2/65 (3%) Metformin + canagliflozin 300 mg: 2/64 (3%) Rosenstock, 2013 153 12 weeks MedDRA definition Metformin: 0% Metformin + empagliflozin 10 mg: 9.9% Metformin + empagliflozin 25 mg: 0% Schumm-Draeger, 2015 168 16 weeks MedDRA definition Metformin: 1/101 (1%) Metformin + dapagliflozin 5 mg twice daily: 5/100 (5%) Metformin + dapagliflozin 10 mg: 3/99 (3%) Qiu, 2014 165 18 weeks Males: balanitis candida and genital infection fungal. Females: vaginal infection, vulvovaginal candidiasis, vulvovaginal mycotic infection, and vulvovaginitis Males Metformin: 1/46 (2.2%) Metformin + canagliflozin 100 mg: 1/40 (2.5%) Metformin + canagliflozin 300 mg: 0/44 (0%) Females Metformin: 2/47 (4.3%) Metformin + canagliflozin 100 mg: 6/53 (11.3%) Metformin + canagliflozin 300 mg: 1/49 (2.0%) Henry, 2012 (a) 88 24 weeks Events suggestive of vulvovaginitis, balanitis, and related genital infection Males Metformin: 0% Metformin + dapagliflozin 5 mg: 5.1% Females Metformin: 3.8% Metformin + dapagliflozin 5 mg: 7.8% Haring, 2014 166 24 weeks MedDRA definition Males Metformin: 0/116 (0%) Metformin + empagliflozin 10 mg: 1/125 (0.8%) Metformin + empagliflozin 25 mg: 1/120 (0.8%) Females Metformin: 0/91 (0%) Metformin + empagliflozin 10 mg: 7/92 (7.6%) Metformin + empagliflozin 25 mg: 9/93 (9.7%) Henry, 2012 (b) 88 24 weeks Based on a predefined list of signs, symptoms and other events suggestive of genital infection Males Metformin: 2/97 (2.1%) Metformin + dapagliflozin 10 mg: 6/106 (5.7%) Females Metformin: 3/111 (2.7%) Metformin + dapagliflozin 10 mg: 12/105 (11.4%) Bolinder, 2014 267 102 weeks Genital infections Metformin: 1/91 (1.1%) Metformin + dapagliflozin 10 mg: 2/91 (2.2%) Bailey, 2013 170 102 weeks Events suggestive of genital infection Metformin: 7/137 (5.1%) Metformin + dapagliflozin 2.5 mg: 16/137 (11.7%) Metformin + dapagliflozin 5 mg: 20/137 (14.6%) Metformin + dapagliflozin 10 mg: 17/135 (12.6%) MedDRA = Medical Dictionary for Regulatory Activities; mg = milligrams * Outcomes are actively ascertained unless otherwise noted.
- 345. 288 Metformin-Based Combination Comparisons Combinationof Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three RCTs compared metformin plus a sulfonylurea with metformin plus a SGLT-2 inhibitor, suggesting increased odds of genital infections for metformin plus SGLT-2 inhibitors and differences in relative odds by gender: pooled OR, 5.2 (95% CI, 3.4 to 7.8) for women and pooled OR, 7.6 (95% CI, 4.0 to 14.4) for men (Figure 93).200, 201, 219 No single study markedly influenced the results, and we did not find significant statistical heterogeneity (I2 = 0.0% for women and I2 = 0.0% for men). A 208-week extension of Nauck, 2014 et al.219 also reported higher rates of genital infection in the SGLT-2 inhibitor combination arm and among women compared with men (Table 119).54 Losses to followup were high (>60% in both arms).54 (SOE: High; Combination of metformin plus a sulfonylurea favored) Table 119. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor on genital infections Author, Year Definition Medication Dose Women Events/N (%) Men Events/N (%) Ridderstrale, 2014 200 MedDRA definition of genital infection Metformin + glimepiride Metformin + empagliflozin 25 mg 12/359 (3%) 49/333 (15%) 5/421 (1%) 41/432 (9%) Leiter, 2015 201 Males: balanitis, balanitis candida, balanoposthitis, genital candidiasis, genital infection fungal, and posthitis Females: genital infection fungal, vaginal infection, vulvitis, vulvovaginal candidiasis, vulvovaginal mycotic infection, and vulvovaginitis Metformin + glimepiride Metformin + canagliflozin 300 mg 6/219 (2.7%) 38/244 (15.6%) 5/263 (1.9%) 22/241 (9.1%) Nauck, 2014 219 MedRA definition of genital infection Metformin + glipizide Metformin + dapagliflozin 11/185 (5.9) 42/180 (23.3) 1/223 (0.4) 18/226 (8.0) Del Prato, 2015 54 * Confirmed genital infection Metformin + glipizide Metformin + dapagliflozin 11/408 (5.9%) 41/406 (22.8%) 1/408 (0.4%) 17/406 (7.5%) MedDRA = Medical Dictionary for Regulatory Activities; mg = milligrams * Del Prato, 2015 is the 208-week extension study of Nauck, 2014.
- 346. 289 Figure 93. Pooledodds ratio of genital or mycotic infections comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; Group 1 = combination of metformin plus a sulfonylurea; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; Met = metformin; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2; SU = sulfonylurea Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Five RCTs compared metformin plus a DPP-4 inhibitor with metformin plus a SGLT-2 inhibitor and generally found more genital infections with the SGLT-2 inhibitor arms.90, 153, 156, 158, 209 These studies were not pooled owing to differences in study duration. Two of the studies stratified outcomes by gender (Table 120).90, 158 (SOE: Moderate; Combination of metformin plus a DPP-4 inhibitor favored)
- 347. 290 Table 120. Randomizedcontrolled trials comparing a combination of metformin plus a DPP-4 inhibitor with a combination of metformin plus an SGLT-2 inhibitor on genital infections Author, Year Followup (Weeks) Medication Dose Women Events/N (%) Men Events/N (%) Total Events/N (%)* Comments Rosenstock, 2012 156 12 weeks Metformin + sitagliptin 1/27 (3.7) NR NR Metformin + canagliflozin 100 mg 2/28 (7.1) NR NR Metformin + canagliflozin 200 mg 4/32 (12.5) NR NR Metformin + canagliflozin 300 mg 1/28 (3.6) NR NR Rosenstock, 2013 153 12 weeks Metformin + sitagliptin NR NR 2/71 (2.8) Metformin + empagliflozin 10 mg NR NR 7/71 (9.9) Metformin + empagliflozin 25 mg NR NR 0/70 (0) Rosenstock, 2015 209 24 weeks Metformin + saxagliptin NR NR 1/176 (0.6) Metformin + dapagliflozin 10 mg NR NR 10/179 (6.0) Lavalle-Gonzalez, 2013 158 52 weeks Metformin + sitagliptin 5/194 (2.6) 2/172 (1.2) NR ITT analysis not performed Metformin + canagliflozin 100 mg 22/194 (11.3) 9/174 (5.2) NR Metformin + canagliflozin 300 mg 20/202 (9.9) 4/165 (2.4) NR Ferrannini, 2013 90 78 weeks Metformin + sitagliptin 0/29 (0) 0/27 (0) NR ITT analysis not performed Metformin + empagliflozin 10 mg 2/83 (2.4) 3/83 (3.6) NR Metformin + empagliflozin 25 mg 3/88 (3.4) 3/78 (3.8) NR DPP-4 = dipeptidyl peptidase-4; ITT = intention-to-treat; mg = milligrams; NR = not reported; SGLT-2 = sodium-glucose co- transporter-2 * Results for both genders provided if sex-stratified results not reported Strength of Evidence for Genital Mycotic Infections The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 121 and summarized in the key points. All studies were RCTs. Study limitations for all the comparisons were low or medium. In general, we did not find strong differences in outcomes in the lower- versus higher-quality studies. We did not find any evidence of publication bias in any of the comparisons for genital infections. We also did not find any evidence of publication bias or reporting bias in the grey literature review.
- 348. 291 Table 121. Strengthof evidence domains for monotherapy and metformin-based combination comparisons in terms of genital mycotic infections among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. SGLT-2 inhibitors 4 (2,292) Medium Consistent Direct Imprecise Undetected Moderate Metformin favored; 4.1 (2.0 to 8.3) for SGLT-2 inhibitors vs. metformin DPP-4 inhibitors vs. SGLT-2 inhibitors 2 (1,394) Medium Consistent Direct Imprecise Undetected Low DPP-4 inhibitors favored Metformin vs. metformin + SGLT-2 inhibitors 9 (4,035) Low Consistent Direct Precise Undetected High Metformin favored; 3.0 (1.2 to 7.2) for females and 2.7 (0.8 to 9.0) for males Metformin + SU vs. metformin + SGLT-2 inhibitors (longer studies) 3 (3,815) Medium Consistent Direct Precise Undetected High Metformin + SU favored; 5.2 (3.4 to 8.0) for females and 7.6 (4.0 to 14.4) for males Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors 5 (3,423) Medium Consistent Direct Imprecise (n for metformin insufficient) Undetected Moderate Metformin + DPP-4 inhibitors favored; range in OR, 1.0 to 10.4; range in RD, -3% to 9% CI = confidence interval; DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; OR = odds ratio; RD = risk difference; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 349. 292 Evidence for Fracture MetforminVersus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three RCTs compared metformin with metformin plus an SGLT-2 inhibitor and found no differences in fractures.168, 170, 267 Two of these had followup for longer than one year: One 102- week RCT compared metformin with metformin plus dapagliflozin and reported a slightly higher incidence of fractures in the highest-dose dapagliflozin arm [2/137 (1.5%) for metformin versus 2/137 (1.5%) for dapagliflozin 2.5 mg, 2/137 (1.5%) for dapagliflozin 5 mg, and 3/135 (2.2%) for dapagliflozin 10 mg]. There was a high loss to followup in this study, ranging from 30 percent to 47 percent across arms.170 Another 102-week RCT compared metformin with metformin plus 10 mg of dapagliflozin and reported one fracture (1.1%) in each treatment arm (n=91 for both arms).267 A single 16-week RCT compared metformin with metformin plus dapagliflozin and reported no fractures in either arm.168 (SOE: Low; Neither favored for shorter studies; SOE: Low; Neither favored for longer studies) Metformin-Based Combination Comparisons Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Two studies compared metformin plus a GLP-1 receptor agonist with metformin plus a SGLT-2 inhibitor, showing no differences in fracture risk.200, 219 One 104-week RCT compared metformin plus glipizide with metformin plus dapagliflozin and showed a slightly higher incidence of fractures in the metformin plus glipizide arm [9/408 (2.2%) for metformin plus glipizide versus 6/406 (1.5%) for metformin plus dapagliflozin].219 Another 104-week RCT compared metformin plus glimepiride to metformin plus empagliflozin showing similar incidences of fractures in both arms (2%).200 (SOE: Insufficient) Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor One 24-week RCT compared metformin plus a DPP-4 inhibitor with metformin plus a SGLT-2 inhibitor and reported a slightly higher risk of fracture in the metformin arm [2/176 (1.0%) for metformin plus saxagliptin versus 1/179 (0.6%) for metformin plus dapagliflozin].209 (SOE: Low; Metformin plus SGLT-2 inhibitor favored short-term fracture risk) Strength of Evidence for Fracture The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 122 and summarized in the key points. All studies were RCTs. Study limitations for all the comparisons were low. We did not find any evidence of publication bias in any of the comparisons for fractures. A single 52-week, unpublished trial reported no fractures in either arm (NCT01368081), which is consistent with the other studies. We also did not find any evidence of publication bias or reporting bias in the grey literature review.
- 350. 293 Table 122. Strengthof evidence domains for monotherapy and metformin-based combination comparisons in terms of fracture among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. metformin + SGLT-2 inhibitors (shorter studies) 1 (200) Low Unknown Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + SGLT-2 inhibitors (longer studies) 2 (728) Low Consistent Direct Imprecise Undetected Low Neither favored Metformin + SU vs. metformin + SGLT-2 inhibitors (longer studies) 2 (2,363) Medium Inconsistent Direct Imprecise Undetected Insufficient Unable to determine Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors (shorter study) 1 (534) Low Unknown Direct Imprecise Undetected Low Metformin + SGLT-2 inhibitor favored SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † We only include estimates for comparisons with high or moderate strength of evidence.
- 351. 294 Evidence for VolumeDepletion Monotherapy Comparisons Metformin Versus SGLT-2 Inhibitors Two moderately-sized, short RCTs compared metformin with SGLT-2 inhibitors and reported inconsistent results.88, 89 One 12-week RCT comparing metformin with dapagliflozin reported a higher incidence of hypotensive events with metformin (4% for metformin versus 0% for dapagliflozin 5 mg and 10 mg).89 One 24-week RCT comparing metformin with dapagliflozin 5 mg reported significantly more events of hypotension or syncope with dapagliflozin [0/201 (0%) for metformin versus 4/203 (2%) for dapagliflozin 5 mg].88 (SOE: Low; Conflicting results) Metformin Versus Metformin-Based Combination Comparisons Metformin Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Six RCTs compared metformin with the combination of metformin plus an SGLT-2 inhibitor for this outcome.88, 156, 165, 168, 170, 267 We did not combine these in a meta-analysis because of differences in study duration and definition of volume depletion events (Table 123, Figure 94). The two RCTs with long-term followup had large losses to followup and were conflicting with one suggesting a higher risk of hypotension in the SGLT-2 inhibitor-based arm267 and the other suggesting similar rates of volume depletion across arms.170 Volume depletion events were rare in the four short-duration RCTs (Figure 94).88, 156, 165, 168 (SOE: Low; Neither favored for shorter studies; SOE: Low; Neither Favored for longer studies) Figure 94. Pooled odds ratio of volume depletion comparing metformin with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; Group 1 = metformin; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Studies were excluded because they did not contribute any events.
- 352. 295 Table 123. Randomizedcontrolled trials comparing metformin with a combination of metformin plus an SGLT-2 inhibitor on volume depletion Author, Year Followup Definition of Volume Depletion Outcome Results (Metformin Versus Metformin + SGLT2 Inhibitor) Comments Rosenstock, 2012 156 12 weeks Adverse events possibly related to hypovolemia (dizziness, dizziness postural, heart rate increased, tachycardia, and urine output decreased) Metformin: 1/65 (2%) Metformin + canagliflozin 100 mg: 4/64 (6%) Metformin + canagliflozin 200 mg: 3/65 (5%) Metformin + canagliflozin 300 mg: 1/64 (2%) Included in meta- analysis Schumm-Draeger, 2015 168 16 weeks MedDRA definition for hypotension, dehydration, or hypovolemia Metformin: 0/101 (0%) Metformin + dapagliflozin 5 mg twice daily: 0/100 (0%) Metformin + dapagliflozin 10 mg: 0/99 (0%) Included in meta- analysis Qiu, 2014 165 18 weeks Orthostatic hypotension, postural dizziness Metformin: 0/93 (0%) Metformin + canagliflozin 100 mg: 0/93 (0%) Metformin + canagliflozin 300 mg: 0/93 (0%) Included in meta- analysis Henry, 2012 88 24 weeks Hypotension or syncope Metformin: 0/201 (0%) Dapagliflozin 5 mg: 1/194 (0.5%) No ITT analysis performed Included in meta- analysis Bolinder, 2014 267 102 weeks Hypotension Metformin: 0/91 (0%) Metformin + dapagliflozin 10 mg: 1/91 (1.1%) Unclear if ITT analysis performed High losses to follow up Bailey, 2013 170 102 weeks MedDRA definition for hypotension, dehydration, or hypovolemia Metformin: 2/137 (1.5%) Metformin + dapagliflozin 2.5 mg: 0/137 (0%) Metformin + dapagliflozin 5 mg: 3/137 (2.2%) Metformin + dapagliflozin 10 mg: 2/135 (1.5%) High losses to follow up ITT = intention-to-treat; MedDRA = Medical Dictionary for Regulatory Activities; mg = milligrams; SGLT-2 = sodium-glucose co-transporter-2 Metformin-Based Combination Comparisons Combination of Metformin Plus a Sulfonylurea Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Three 104-week RCTs compared metformin plus a sulfonylurea to metformin plus a SGLT-2 inhibitor and described volume depletion events, with varying definitions in each study. The evidence suggested little difference between arms (pooled OR for metformin plus sulfonylurea versus metformin plus SGLT-2 inhibitor, 1.0; 95% CI, 0.6 to 1.7) (Figure 95 and Table 124).200, 201, 219 No single study markedly influenced the results, and we did not detect substantial heterogeneity (I2 = 0.0%). (SOE: Low; Neither favored)
- 353. 296 Figure 95. Pooledodds ratio of volume depletion comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor CI = confidence interval; Group 1 = combination of metformin plus a sulfonylurea; Group 2 = combination of metformin plus a sodium-glucose co-transporter-2 inhibitor; OR = odds ratio; SGLT-2 = sodium-glucose co-transporter-2 Boxes indicate individual study point estimates. The box size denotes the weight of the study, with larger boxes contributing more to the pooled estimate. The width of the horizontal lines represents the 95 percent confidence intervals for each study. The diamond at the bottom of the graph indicates the 95 percent confidence interval for the random-effects pooled estimate. Table 124. Randomized controlled trials comparing a combination of metformin plus a sulfonylurea with a combination of metformin plus an SGLT-2 inhibitor on volume depletion Author, Year Followup Definition of Volume Depletion Outcome Results Events/N (%) Leiter, 2015 201 104 weeks Decreased blood pressure, dehydration, postural dizziness, hypotension, orthostatic hypotension, presyncope, and syncope Metformin + glimepiride: 11/482 (2.3%) Metformin + canagliflozin 100 mg: 8/483 (1.7%) Metformin + canagliflozin 300 mg: 12/485 (2.5%) Ridderstrale, 2014 200 104 weeks MedDRA definition Metformin + glimepiride: 8/780 (1%) Metformin + empagliflozin 25 mg: 11/765 (1%) Nauck, 2014 219 104 weeks Hypotension, dehydration, hypovolemia Metformin + glipizide: 7/408 (1.7%) Metformin + dapagliflozin 10 mg: 6/406 (1.5%) MedDRA = Medical Dictionary for Regulatory Activities; mg = milligrams Combination of Metformin Plus a DPP-4 Inhibitor Versus a Combination of Metformin Plus an SGLT-2 Inhibitor Two RCTs compared metformin plus a DPP-4 inhibitor with metformin plus a SGLT-2 inhibitor and showed no clear differences in volume depletion outcomes.156, 158 Both studies reported active ascertainment of the outcome. One 52-week RCT comparing metformin plus sitagliptin with metformin plus canagliflozin 100 mg and 300 mg reported similar incidences in orthostatic hypotension (0% to 0.3% across all three arms).158 One 12-week RCT comparing metformin plus sitagliptin with metformin plus canagliflozin reported slightly more events related to hypovolemia in the arms receiving the lower doses of canagliflozin [1/65 (2%) for
- 354. 297 metformin plus sitagliptinversus 4/64 (6%) for metformin plus canagliflozin 100 mg, 3/65 (5%) for metformin plus canagliflozin 200 mg, and 1/64 (2%) for metformin plus canagliflozin 300 mg].156 (SOE: Low; Neither favored) Strength of Evidence for Volume Depletion The strength of evidence for the comparative effects of monotherapy and metformin-based combinations are presented in Table 125 and summarized in the key points. All studies were RCTs. Study limitations for all the comparisons were low. Where quality influences results, we describe that under the appropriate comparisons. In general, we did not find strong differences in outcomes in the lower- versus higher-quality studies. We did not find any evidence of publication bias in any of the comparisons for volume depletion. We also did not find any evidence of publication bias or reporting bias in the grey literature review.
- 355. 298 Table 125. Strengthof evidence domains for monotherapy and metformin-based combination comparisons in terms of volume depletion among adults with type 2 diabetes Comparison* Number of Studies (Subjects) Study Limitations Consistency Directness Precision Reporting Bias Strength of Evidence Summary † Metformin vs. SGLT-2 inhibitors 2 (992) Medium Inconsistent Direct Imprecise Undetected Low Conflicting results from 2 RCTs Metformin vs. metformin + SGLT-2 inhibitors (shorter studies) 4 (1533) Medium Consistent Direct Imprecise Undetected Low Neither favored Metformin vs. metformin + SGLT-2 inhibitors (longer studies) 2 (728) Low Consistent Direct Imprecise Undetected Low Neither favored Metformin + SU vs. metformin + SGLT-2 inhibitors (longer studies) 3 (3,815) Medium Inconsistent Direct Imprecise Undetected Low Neither favored Metformin + DPP-4 inhibitors vs. metformin + SGLT-2 inhibitors 2 (1,735) Medium Inconsistent Direct Imprecise Undetected Low Neither favored DPP-4 inhibitors = dipeptidyl-peptidase 4 inhibitors; GLP-1 receptor agonists = glucagon-like peptide-1 agonists; SGLT-2 inhibitors = sodium-glucose co-transporter 2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione * We only list comparisons that were evaluated by at least one randomized controlled trial. All other comparisons were considered to have insufficient evidence because of a lack of available evidence. Unless otherwise specified, conclusions for the intermediate outcomes are short-term (1 year or shorter) because there are few longer-duration studies evaluating this outcome. † Unless otherwise specified, the estimates are the pooled odds ratios (95 percent confidence intervals) from randomized controlled trials. We only include estimates for comparisons with high or moderate strength of evidence.
- 356. 299 Key Question 4:Subgroups Although thirty-two studies reported on the comparative effectiveness and safety for sub- populations relevant to Key Question 4 (Appendix D, Table D14), few studies had sufficient power to assess comparative effectiveness or safety by subgroup. The evidence favoring one medication over another across subgroups is unclear. We included 29 RCTs and five cohort studies that addressed this Key Question. The majority of these trials (n=22) evaluated subgroup effects on the outcome of HbA1c.54, 77, 80, 82, 84, 104, 107, 118, 126, 139, 141, 142, 145, 149, 151, 154, 160, 193 41187, 194, 241, 268, 269 RCTs also included subgroup results on weight gain,77, 169, 241, 270, 271 hypoglycemia,190, 195, 272 and fractures.273 The cohort studies included subgroup results for mortality,234 cardiovascular events,245, 246 fractures,274 and kidney disease progression.249 We were unable to draw conclusions about the differential effects of medications in the specified sub-populations because of the small number of studies available for any one outcome for the included comparisons. Subgroups Defined by Age Hemoglobin A1c Sixteen RCTs, out of the 21 reporting on subgroups for this outcome, did not find differences in the effects of diabetes medications on HbA1c by age. Cardiovascular Mortality and Morbidity We included two retrospective cohort studies which reported on cardiovascular outcomes by age. One study compared metformin users, glimepiride users, and glyburide users.245 Metformin use was associated with lower risk of nonfatal CVD events compared with glyburide use among older participants (>51 years old) (Adjusted HR for metformin vs. glyburide among those age 51 to 70 years, 0.28; 95% CI, 0.20 to 0.39; adjusted HR among those age 71 years or older, 0.30; 95% CI, 0.18 to 0.48). In the younger age group, only metformin was associated with a decreased risk of cardiovascular events (adjusted HR for metformin vs. glyburide, 0.39; 95% CI, 0.21 to 0.73). Another retrospective cohort study compared metformin with sulfonylureas and found no difference in the incidence of death or cardiovascular events across age groups.246 Hypoglycemia Two RCTs evaluating the risk of hypoglycemia by age reported no differences by age for the combination of metformin and a sulfonylurea versus the combination of metformin and a DPP-4 inhibitor.190, 195 Kidney Function Decline A retrospective cohort study compared the effect of metformin, rosiglitazone and sulfonylureas on kidney function and found no differences by age for kidney disease progression.249
- 357. 300 Subgroups Defined bySex Hemoglobin A1c Seventeen RCTs examined the impact of sex on glycemic control (HbA1c) for the comparisons of interest and found no differences by sex.77, 80, 82, 84, 104, 107, 118, 126, 139, 141, 142, 145, 149, 151, 154, 194, 268 Weight One trial reported a greater weight reduction among men compared with women for the combination of metformin plus dapagliflozin versus metformin; the mean decrease at 24 weeks attributable to the addition of dapagliflozin was 2.76 kg for men and 1.22 kg for women, p for interaction = 0.048).169 A second study comparing metformin with pioglitazone reported that while both men and women in the metformin arm had a slight but not significant weight loss, those in the pioglitazone arm differentially gained weight; the mean increase at 12 weeks was 1.78 kg for women (p = 0.039) and 0.86 kg for men (p=0.151).77 Another trial that compared metformin with metformin plus pioglitazone over 24 weeks reported no treatment differences by sex.270 Long-Term Clinical Outcomes The two retrospective cohort studies described above comparing the effect of metformin and different sulfonylureas on cardiovascular risk found no association between treatments and cardiovascular outcomes by sex.245, 246 However, a retrospective cohort study of new monotherapy users found that compared with those on metformin, women on rosiglitazone had a higher risk of death (RR, 6.21; 95% CI, 1.22 to 19.65) than men (RR, 1.76; 95% CI, 1.41 to 2.18).234 The p-value for the interaction between treatment and sex was 0.034. Hypoglycemia Two studies that compared the combination of metformin and sulfonylureas with the combination of metformin and a DPP-4 inhibitor found no differences by gender for this outcome.190, 272 Fractures A retrospective analysis of the ADOPT trial found that women treated with rosiglitazone had an increased risk of fracture relative to those treated with metformin or glyburide (HR, 1.57 and 1.61, respectively); the investigators did not find an increased risk of fracture among men in this study with a median of 4 years of followup.273 One cohort study reported that women have a higher risk of peripheral fractures when treated with pioglitazone than with sulfonylureas (adjusted HR, 1.77; 95% CI, 1.32 to 2.38).274 However, the study did not find a statistically significant increased risk of peripheral fractures for women treated with rosiglitazone compared with sulfonylureas (adjusted HR, 1.17; 95% CI 0.91 to 1.50). Men treated with thiazolidinediones had an increased risk compared with men treated with sulfonylureas (adjusted HR, 1.61; 95% CI, 1.18 to 2.20).
- 358. 301 Subgroups Defined byRace/Ethnicity Hemoglobin A1c Thirteen RCTs examined the impact of race on HbA1c reduction, and found no differences by race for the comparisons studied.82, 84, 104, 107, 126, 139, 141, 145, 151, 154, 160, 194 Kidney Function Decline A single retrospective cohort evaluating the effects of metformin, glyburide, and glimepiride on progression of chronic kidney disease found no differences by race for this outcome.249 Subgroups Defined by Body Mass Index Hemoglobin A1c Sixteen RCTs found no differences by baseline BMI on the effects of diabetes medications on HbA1c reduction for the comparisons studied.54, 80, 82, 84, 104, 107, 118, 126, 139, 141, 145, 154, 194, 241, 269 Weight Two studies reported on effects of weight in a subgroup of obese patients, but they did not report on this outcome in the non-obese patients.241, 271 One RCT found that obese patients treated with metformin lost an average of 1.3 kg, but those treated with sulfonylureas gained an average of 3.7 kg.271 In another RCT, obese patients allocated to metformin lost an average of 0.9 kg but those allocated to a combination of metformin plus rosiglitazone gained an average of 2.5 kg.241
- 359. 302 Discussion This systematic reviewaddresses the comparative effectiveness and safety of diabetes medications used most frequently in the United States as monotherapy and compares therapies in combination with metformin to each other. This review updates and adds to two previous comparative effectiveness reviews (CER) published in 200715 and 2011,16 by focusing on the head-to-head comparisons of medications most relevant to clinicians and their patients (Table 2), particularly those for which evidence was previously lacking. We broadened the scope by including seven medications newly approved by Food and Drug Administration (FDA), including one new medication class, the sodium-glucose co-transporter-2 (SGLT-2) inhibitors. We identified 107 new studies, which included 87 trials and 20 observational studies, published since we completed our 2011 review. Our comprehensive review of the newer medication classes in comparison with other medications and comparisons of combination therapies is an important contribution to the literature, because it is the first to address this many comparisons for a wide range of outcomes in patients with type 2 diabetes mellitus. Key Findings in Context Intermediate Outcomes One hundred sixty-two RCTs evaluated the intermediate outcomes of hemoglobin A1c (HbA1c), weight, systolic blood pressure, and heart rate. Studies mainly measured these intermediate outcomes at 1 year or less; only six studies were longer than 2 years. The few longer studies had results consistent with the results from the shorter-term studies. Rarely, results from a longer study conflicted with results from a short-term study, but the high losses to followup (generally >20%) made these uninterpretable. Therefore, short-term results are discussed below, unless otherwise specified in the figure or text. Hemoglobin A1c HbA1c is unequivocally linked to microvascular disease,10, 275, 276 making it a good proximal outcome to measure. Consistent with the prior 2011 report,16 most oral diabetes medications (thiazolidinediones, sulfonylureas, and metformin) had similar efficacy in reducing HbA1c when used as monotherapy (Figure 96). The one exception was that metformin yielded a greater reduction in HbA1c compared with DPP-4 inhibitors, consistent with the prior report.16 In the last report,16 metformin versus sulfonylurea was graded as having a high strength of evidence showing no significant between-group differences in HbA1c; therefore, it was not updated in this report. In this report, metformin versus GLP-1 receptor agonists and metformin versus SGLT-2 inhibitors also showed no clear between-group differences in HbA1c. They were graded as low strength of evidence, because the three studies in each comparison were imprecise and inconsistent. In this update, we found inconsistent findings in the studies of GLP-1 receptor agonists. It may be that the individual GLP-1 receptor agonists have different effects on HbA1c. A 2011 Cochrane systematic review showed small between-group differences in HbA1c, around 0.3%, for different GLP-1 receptor agonists.277 The strength of evidence was graded as insufficient for the other monotherapy comparisons of SGLT-2 inhibitors and GLP-1 receptor agonists, and this warrants further study. All metformin combination therapies were better at reducing HbA1c than metformin monotherapy regimens, with between-group differences of around 0.7 to 1 absolute percentage
- 360. 303 points (Figure 96).While several moderate strength of evidence combination comparisons were significantly favored over the comparator combination (Figure 96), most between-group differences were small (<0.3%), with questionable clinical relevance. Only one combination comparison with moderate strength of evidence was favored by >0.3% over any other combination comparison: the combination of metformin plus a GLP-1 receptor agonist reduced HbA1c more than metformin plus a DPP-4 inhibitor. Despite the clinical interest in comparing metformin plus injectables, there was insufficient or low strength of evidence on glycemic control for the following comparisons: metformin plus the GLP-1 receptor agonists versus metformin plus basal or premixed insulin, and metformin plus premixed insulin versus metformin plus basal insulin. Two prior network meta-analyses278, 279 showed that most metformin-based combination comparisons had similar reductions in HbA1c. However, the results of the direct comparisons evaluated in this report are more precise allowing us to detect smaller between group differences than the results of the indirect comparisons found in the network meta-analyses. Figure 96. Pooled between-group differences in hemoglobin A1c and strength of evidence for monotherapy and metformin-based combination comparisons BL = baseline; CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; ES = effect size (mean between-group difference in HbA1c); GLP1 = glucagon-like peptide-1 agonists; HbA1c = hemoglobin A1c; Met = metformin; PLE = profile likelihood estimate; SGLT2 = sodium-glucose co-transporter-2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione; wks = weeks The width of the horizontal lines represents the 95 percent confidence intervals for each pooled analysis.
- 361. 304 Weight Monotherapy and combinationmedication comparisons generally showed significant between-group differences when comparing medications anticipated to increase weight (sulfonylurea, thiazolidinediones, and insulin) with medications expected to maintain or decrease weight (metformin, DPP-4 inhibitors, GLP-1 receptor agonists, and SGLT-2 inhibitors) (Figure 97). In the prior report, metformin versus thiazolidinediones and metformin versus sulfonylureas were found to favor metformin, with differences of 2.5 kg, with high strength of evidence; therefore, these comparisons were not updated. In this report, metformin decreased weight more than DPP-4 inhibitors, while sulfonylureas caused slightly less weight gain than thiazolidinediones (Figure 97). Compared with metformin plus a DPP-4 inhibitor, the combinations of metformin plus GLP-1 receptor agonists and metformin plus SGLT-2 inhibitors were favored (Figure 97). Several comparisons discussed below had moderate strength of evidence but were unable to be pooled owing to differences among the studies. SGLT-2 inhibitors decreased weight more than metformin and more than DPP-4 inhibitors (between- group differences ranging from -1.3 kg to -2.7 kg). DPP-4 inhibitors and GLP-1 receptor agonists both decreased weight more than thiazolidinediones (between-group differences ranging from -2.3 kg to -3.5 kg). Lastly, metformin plus a sulfonylurea was favored over metformin plus a premixed or long-acting insulin (between-group difference ranging from -0.5 kg to -1.7 kg). Despite the clinical interest in comparing metformin plus injectables, there was low strength of evidence on weight for the comparisons of metformin plus the GLP-1 receptor agonists versus metformin plus basal or premixed insulin and metformin plus premixed insulin versus metformin plus basal insulin. Notably, weight differences were small to moderate in these mainly short trials. However, even small to moderate amounts of weight gain (5 percent to 10 percent of body weight) may be associated with increased insulin resistance.280 In addition, weight loss and glycemic control were reported as the primary drivers of patient preferences for diabetes medications when compared with treatment burden and side effects in a recent systematic review.281 Drug effects on weight, therefore, have a strong impact on the choice of the drug added for second-line combination therapy in a patient not well controlled on a single agent. Our findings about diabetes medications effects on weight are similar to other prior systematic reviews.16, 277, 282 As monotherapy and in combination with metformin, thiazolidinediones, sulfonylureas, and basal or premixed insulin are associated with weight gain; DPP-4 inhibitors with weight maintenance; and SGLT-2 inhibitors and GLP-1 receptor agonists with weight loss.16, 277, 282, 283 Our systematic review builds on prior work by adding more direct comparative data about metformin combinations, which further confirm the known weight effects of the individual medications.
- 362. 305 Figure 97. Pooledbetween-group differences in weight and strength of evidence for monotherapy and metformin-based combination comparisons BL = baseline; CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; ES = effect size (mean between-group difference in weight); GLP1 = glucagon-like peptide-1 agonists; kg = kilograms; Met = metformin; PLE = profile likelihood estimate; SGLT2 = sodium-glucose co-transporter-2 inhibitors; SU = sulfonylurea; TZD = thiazolidinedione; wks = weeks; wt = weight The width of the horizontal lines represents the 95 percent confidence intervals for each pooled analysis. Systolic Blood Pressure and Heart Rate Systolic blood pressure and heart rate were evaluated for the newer classes of medications, SGLT-2 inhibitors and GLP-1 receptor agonists, owing to suspected effects of these medications based on prior literature.277, 282 Blood pressure control is important in adults with diabetes.220, 284- 286 The United Kingdom Prospective Diabetes Study (UKPDS) showed that for every 10 mmHg decrease in systolic blood pressure, there is a 15 percent decrease in diabetes-related deaths.285 Consistent with a prior systematic review on SGLT-2 inhibitors,282 the SGLT-2 inhibitors consistently reduced systolic blood pressure in all comparisons where there were sufficient studies (significant between-group differences of 3 to 5 mmHg when compared with other diabetes medications that have no effect on blood pressure). Our review builds on this work by evaluating direct comparisons of specific medication classes as comparators, as opposed to grouping all active comparators together. This is especially important, because thiazolidinediones and GLP-1 receptor agonists also have been associated with decreases in systolic blood pressure by 3 to 5 mmHg.15, 277 Metformin plus a GLP-1 receptor agonist had a greater reduction in systolic blood pressure compared with metformin alone (pooled between- group difference, 3.1 mmHg; 95% CI, 1.4 to 4.9 mmHg), with moderate strength of evidence.
- 363. 306 While the clinicalrelevance of these small differences is unclear, a change of 3-5 mmHg is about half the effect of a low sodium diet (around 7-11 mmHg) and about one-third the effect of blood pressure medications (around 10-15 mmHg).287, 288 Future research will be needed to determine whether there are any links between these small differences in blood pressure and micro- and macrovascular outcomes, given the prevalent use of effective medications to reduce cardiovascular risk (e.g., aspirin, blood pressure, and cholesterol medications). Increased heart rate is associated with increased mortality.289 However, whether heart rate is an independent predictor of long-term clinical outcomes such as mortality is less clear.44 We opted to evaluate heart rate for the newer medications, SGLT-2 inhibitors and GLP-1 receptor agonists, given their association with mortality. In addition, we wanted to see if the benefits from blood pressure reduction might be offset by a concomitant increase in heart rate. We did not identify any prior systematic reviews that have evaluated this outcome for the diabetes comparisons of interest. Only two comparisons had sufficient data to grade the evidence as more than insufficient or low. The SGLT-2 inhibitors in combination with metformin were found to decrease heart rate by 1.5 bpm (95% CI, 0.6 bpm to 2.3 bpm) when compared with metformin plus a sulfonylurea; metformin and GLP-1 receptor agonist trials showed no between-group differences in heart rate. Therefore, these early findings support the findings of minimal to no effects on heart rate and no increase in heart rate for the newer medications. All-Cause Mortality and Macrovascular and Microvascular Outcomes Ninety-six RCTs and 22 observational studies evaluated these clinical outcomes. Compared to the prior report,16 the evidence on mortality, cardiovascular mortality, and cardiovascular morbidity was strengthened for many comparisons, although most of this evidence was of low strength. The evidence regarding treatment effect on microvascular outcomes remains largely insufficient. Overall, the evidence base on these outcomes was limited by the short duration of RCTs (<2 years) and a lack of high-quality observational studies that would allow detection of treatment differences for infrequent outcomes. Also, none of the included RCTs were designed to evaluate these long-term outcomes. All-Cause Mortality, Cardiovascular Mortality, and Cardiovascular Morbidity This report builds on our prior results for the comparison of metformin and sulfonylurea monotherapy.16 In this update, we found moderate strength of evidence that sulfonylurea monotherapy was associated with a 50 percent to 70 percent increased relative risk (absolute risk difference, 0.1% to 2.9% in RCTs; number needed to treat, 34 to 1,000) of long-term cardiovascular mortality compared with metformin monotherapy. Our findings on all-cause mortality and cardiovascular morbidity were supportive of this conclusion, with low strength of evidence suggesting that metformin is favored over sulfonylurea monotherapy for both long-term mortality and cardiovascular morbidity. Our findings on sulfonylurea monotherapy enhance findings from meta-analyses published in 2012 and 2013, which relied more heavily on observational data or did not report explicitly on head-to-head comparisons of metformin and sulfonylurea monotherapy.290, 291 Importantly, we cannot know the absolute impact of individual drug classes on cardiovascular outcomes from this comparative effectiveness analysis. Our results suggest that cardiovascular mortality is lower in the metformin than sulfonylurea arms of the studies; however, we do not know if metformin actually decreases cardiovascular mortality
- 364. 307 or just increasescardiovascular mortality less than sulfonylureas, and likewise, we do not know if sulfonylureas actually increase cardiovascular mortality or just decrease cardiovascular mortality less than metformin. We evaluated pioglitazone and rosiglitazone separately when they were compared with other drug classes for all-cause mortality and cardiovascular outcomes, given concerns about cardiovascular risk and mortality associated with rosiglitazone raised previously292 and acted upon by the FDA through recommendations regarding the prescription of this medication in 2010.293, 294 While the strength of evidence was low, metformin was favored over the combination of metformin plus rosiglitazone for short-term all-cause mortality, as well as cardiovascular mortality and cardiovascular morbidity; metformin was also favored, albeit with low strength of evidence, over rosiglitazone monotherapy for long-term mortality and long-term cardiovascular morbidity. We had little evidence on the combination of metformin plus a sulfonylurea or metformin plus a thiazolidinedione. In The Rosiglitazone Evaluated for Cardiovascular Outcomes in oral agent combination therapy for type 2 Diabetes (RECORD) trial, 4,447 participants with diabetes were randomized to a rosiglitazone-based combination (with either metformin or sulfonylurea) or to the combination of metformin plus a sulfonylurea.49 We excluded this study from this outcome assessment, because it did not stratify results by a comparison of interest. This trial did not find a difference between rosiglitazone-based two-drug therapy and metformin plus a sulfonylurea for the primary endpoint, cardiovascular hospitalization or death. After 5.5 years, all-cause mortality, cardiovascular death, and stroke were slightly higher in the metformin plus sulfonylurea arm, and myocardial infarction rates were slightly (non-statistically significantly) higher in the rosiglitazone combination therapy arm. Of note, the FDA commissioned an independent re-adjudication and analysis of the data from RECORD and subsequently (in 2013) lifted their restrictions on rosiglitazone usage.295 Although strength of evidence was low, compared with the combination of metformin plus a sulfonylurea, the combination of metformin plus a DPP-4 inhibitor was associated with lower rates of long-term all-cause mortality and cardiovascular mortality and morbidity; an unpublished study with long-term followup was supportive of these conclusions. We also identified many new studies evaluating short-term all-cause mortality and cardiovascular outcomes for DPP-4 inhibitor comparisons, but most of this evidence was of low strength. Several meta-analyses published since the 2011 report have evaluated DPP-4 inhibitors and all-cause mortality and cardiovascular outcomes; they tended to combine comparators (active, placebo, or both) against DPP-4 inhibitors so were not conclusive about specific direct comparisons.296-301 The conclusions of these reviews have been based mainly on trials with less than 2 years of followup and have reported mixed results on short-term mortality and cardiovascular risk. Outside of these meta-analyses, three large RCTs have evaluated DPP-4 inhibitors compared with placebo, added to the standard diabetes treatment per routine clinical care, and cardiovascular outcomes.302-304 None of these RCTs evaluated direct head-to-head comparisons of interest, and none were included in this review. All of these trials designated a composite cardiovascular outcome as the primary outcome and reported non-inferiority (but not superiority) for the DPP-4 inhibitor addition compared with placebo addition for the composite cardiovascular outcome.302-304 Rates of the primary endpoint were lower in the DPP-4 inhibitor versus placebo arm in two of the three trials [risk difference (statistical comparison not provided): -0.5%,303 -0.2%,302 0.1%302 ]. Participants in all three trials were treated per clinical
- 365. 308 standards with diabetesand cardiovascular medications, and, in the two trials reporting on this, participants in the placebo arm were more likely to have an increase or addition of a diabetes medication and were more likely to start insulin.302, 303 There are several reasons that these RCTs demonstrating “non-inferiority” must be interpreted with caution: 1) differential diabetes medication use in the DPP-4 versus placebo arms raises the undeniable possibility that we are observing the effects of other diabetes medications (e.g., insulin) on the cardiovascular outcomes; 2) the assumptions for the non-inferiority sample size/power calculations are based on relative measures when an absolute risk difference may be quite relevant (e.g., an absolute risk difference of 0.1% corresponds to a number needed to harm of 1,000 – a number that has public health relevance given the potential population to be exposed); 3) effects of DPP-4 inhibitors versus placebo on individual outcomes (e.g., mortality, cardiovascular mortality) varied across the studies (Table 126); and 4) followup times of the studies were still short relative to the likely duration of use in actual clinical populations. Table 126. Placebo-controlled RCTs evaluating DPP-4 inhibitors added to standard treatment with composite cardiovascular primary outcome Trial N Study Population Median Followup All-Cause Mortality CVD Death Nonfatal MI SAVOR- TIMI 53 302 16,492 CVD or risk factors 2.1 years HR, 1.11 (0.96 to 1.27; P=0.15) RD: 0.3% HR, 1.03 (0.87 to 1.22; P=0.72) RD: 0.3% HR, 0.95 (0.80 to 1.12; P=0.52) RD: -0.2% EXAMINE 304 5,380 Recent CVD* 18 months HR, 0.88 (0.71 to 1.09; P=0.23) RD: -0.8% HR, 0.85 (0.66 to 1.10; P=0.21) RD: -0.8% HR, 1.08 (0.88 to 1.33); P=0.47) RD: 0.4% TECOS 303 14,671 CVD 3.0 years HR, 1.01 (0.90 to 1.14; P=0.88) RD: 0.2% HR, 1.03 (0.89 to 1.19; P=0.71) RD: 0.2% HR, NR RD: 0.1% CVD = cardiovascular disease; EXAMINE = Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care; HR = hazards ratio; MI = myocardial infarction; RD = risk difference; SAVOR-TIMI 53 = Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus Thrombolysis in Myocardial Infarction; TECOS = Trial Evaluating Cardiovascular Outcomes with Sitagliptin HR displayed with 95% CI and P value and RD with placebo arm as reference. *Acute myocardial infarction or unstable angina in the past 15-90 days. The evidence remains largely insufficient or of low strength regarding mortality and cardiovascular benefits or harms associated with GLP-1 receptor agonists. A meta-analysis suggested no difference in mortality rates for GLP-1 receptor agonists compared with other agents but did not make explicit monotherapy or combination comparisons.305 The evidence on all-cause mortality and cardiovascular mortality and morbidity for SGLT-2 inhibitor comparisons was limited. A meta-analysis of 25 studies comparing SGLT-2 inhibitor monotherapy with placebo or active monotherapy reported a non-statistically significant decrease in the risk of cardiovascular events for SGLT-2 inhibitors versus placebo or active monotherapy (OR, 0.90; 95% CI, 0.72 to 1.13); most studies were 52 weeks or shorter in duration.282 Retinopathy, Nephropathy, and Neuropathy While we found more evidence on microvascular outcomes compared to the 2011 report,16 all evidence was of low strength or inconclusive, thereby limiting substantial conclusions. We
- 366. 309 did not identifyany other evidence syntheses of these microvascular outcomes published since the prior report. Adverse Events One hundred thirty-seven RCTs and eight observational studies evaluated adverse events. The RCTs mainly measured adverse events at 1 year or earlier. Five percent of RCTs were longer than 2 years. Of the few RCTs that evaluated longer time frames, most (75%) had at least 20 percent losses to followup, making it challenging to draw firm longer-term conclusions. Therefore, results discussed below are generally for the short term (less than 2 years) unless otherwise specified in the figure or text. As adverse events sometimes accrue over time, the mainly short-term differences in adverse events could be larger in the long-term. In addition, short-term studies measuring rare adverse events with no between-group differences could develop between-group differences when evaluated over the longer term or in those with higher baseline risk. Hypoglycemia Severe hypoglycemia is associated with increased morbidity (e.g., reduced cognition), increased avoidable health care use (e.g., emergency room visits for hypoglycemia), and increased mortality in clinical trials and observational studies.11, 306-308 We added new information on this important outcome in this report. We found moderate strength of evidence that sulfonylureas had an increased risk of severe hypoglycemia when compared with metformin (for RCTs: range in ORs, 1.4 to 2.0; range in RDs, 1% to 23%) or thiazolidinedione monotherapy (OR 8.1, RD, 0.5% from ADOPT).50 Similarly, in combination with metformin, sulfonylurea use had a greater risk of severe hypoglycemia when compared with the combination of metformin plus DPP-4 inhibitors or SGLT-2 inhibitors (for RCTs: range in ORs, 6 to 14; range in RDs, 0% to 3%). In this report, we confirmed the elevated risk of mild, moderate, or total hypoglycemia associated with sulfonylureas, either alone or in combination, compared with both the older and newer hypoglycemic agents (Figure 98). Of the combination comparisons, we confirmed that metformin plus basal insulin had an 11 percent to 77 percent lower risk of hypoglycemia compared with the combination of metformin plus premixed insulin, with moderate strength of evidence. For the newer medications (SGLT-2 inhibitors, GLP-1 receptor agonists, and DPP-4 inhibitors), we added to the evidence base by showing that SGLT-2 inhibitor monotherapy may be associated with 54 percent lower odds of mild or moderate hypoglycemia compared with metformin monotherapy, although absolute event rates were small across arms. This is consistent with a high-quality systematic review of the SGLT-2 inhibitors that also showed a non- significant lower risk of hypoglycemia in the SGLT-2 inhibitor arms compared with active comparators, although excluding sulfonylureas.282 We found that mild or moderate hypoglycemia was 1.7-fold higher for the combination of metformin plus an SGLT-2 inhibitor compared with metformin monotherapy. We also found an increased risk of hypoglycemia with the combination of metformin plus premixed or basal insulin compared with metformin plus GLP-1 receptor agonists (range in absolute risk differences of 3% to 13%; moderate strength of evidence). Prior systematic reviews of individual classes of newer agents had sparse data on hypoglycemia when compared with active comparators, although the newer classes were generally found to have low rates of hypoglycemia.277, 282, 298 While we found more studies
- 367. 310 comparing metformin combinationswith metformin plus sulfonylurea, there were still few studies on the newer medication classes as monotherapy and in combination with metformin. Figure 98. Pooled odds ratios of mild/moderate hypoglycemia and strength of evidence for monotherapy and metformin-based combination comparisons CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; Met = metformin; OR = odds ratio; PLE = profile likelihood estimate; RD = absolute risk difference; SGLT2 = sodium-glucose co-transporter-2 inhibitors; SU = sulfonylurea; TZD = thiazolidinediones; wks = weeks The width of the horizontal lines represents the 95 percent confidence intervals for each pooled analysis. Drug 1 is the reference group. Gastrointestinal Side Effects Metformin and GLP-1 receptor agonists as monotherapy and in combination had an increased risk of gastrointestinal (GI) adverse events (typically nausea, vomiting, or diarrhea) when compared with most other comparators, with moderate to high strength of evidence (Figure 99). Several medications had similar rates of GI adverse events with moderate or high strength of evidence: thiazolidinedione versus sulfonylurea, metformin plus a sulfonylurea versus metformin plus a DPP-4 inhibitor, metformin plus a thiazolidinedione versus metformin plus a sulfonylurea, metformin plus a sulfonylurea versus metformin plus a SGLT-2 inhibitor, metformin monotherapy versus metformin plus a SGLT-2 inhibitor, and metformin monotherapy versus metformin plus a DPP-4 inhibitor. We confirmed findings of the 2011 report16 showing that metformin had a greater risk of GI adverse events than thiazolidinediones, sulfonylureas, or DPP-4 inhibitors. We also report new findings showing GLP-1 receptor agonists have higher risk of GI adverse events when compared with thiazolidinediones and sulfonylureas, both as monotherapy or when used in combination with metformin. Our data confirm the GLP-1 comparative findings from a prior Cochrane
- 368. 311 systematic review277 and addinformation about specific combination comparisons and specific types of GI adverse events. GLP-1 receptor agonists also had higher risk of nausea and vomiting than metformin but no significant difference in diarrhea. The combinations of metformin plus DPP-4 inhibitors did not have worse GI adverse events than metformin monotherapy or metformin combinations. Lastly, we report new findings that SGLT-2 inhibitors showed no difference in GI adverse events when compared in combination with metformin against metformin plus sulfonylureas or compared with metformin monotherapy. Figure 99. Pooled odds ratios of gastrointestinal adverse events and strength of evidence for monotherapy and metformin-based combination comparisons* AE = adverse event; CI = confidence interval; DPP4 = dipeptidyl peptidase-4 inhibitors; GI = gastrointestinal; Met = metformin; OR = odds ratio; PLE = profile likelihood estimate; RD = absolute risk difference; SGLT2 = sodium-glucose co-transporter-2 inhibitors; SU = sulfonylurea; TZD = thiazolidinediones The width of the horizontal lines represents the 95 percent confidence intervals for each pooled analysis. Drug 1 is the reference group. * All results presented in this graph are based on short-term (less than 52 weeks) studies unless otherwise specified. † Based on studies with 104 weeks of followup. Cancer The evidence about cancer was generally insufficient because of a lack of studies, and the existing evidence was of low strength. Of 25 RCTs reporting on cancer, only eight (32 percent) had at least 2 years of followup. Most published studies for the comparisons did not report on cancer events in all arms, which limited our ability to synthesize the evidence quantitatively and to draw conclusions.
- 369. 312 Reviews and meta-analysespublished since the 2011 report suggest that metformin decreases the risk of many types of cancer309, 310 and suggest that pioglitazone311 increases the risk of bladder cancer slightly, but we did not include many of the studies supporting those conclusions because of our stringent inclusion criteria for observational studies. We excluded the PROspective pioglitAzone Clinical Trial In macroVascular Events (PROactive) study,312 because it did not evaluate a comparison of interest. This trial found a higher rate of bladder cancer in the pioglitazone versus placebo arm, which did not persist in the 5.8-year followup study that included only 74 percent of the original study population, most of whom did not take pioglitazone after the randomized period.313 However, our review adds low strength of evidence for many comparisons of GLP-1 receptor agonists and DPP-4 inhibitors. Evidence on these therapies and cancer outcomes is of particular interest given the preclinical evidence linking incretins (the GLP-1 receptor agonists and DPP-4 inhibitors) to cancer.314 The three large, placebo-controlled RCTs described above, Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus Thrombolysis in Myocardial Infarction (SAVOR TIMI 53), Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care (EXAMINE), and Trial Evaluating Cardiovascular Outcomes with Sitagliptin (TECOS), did not find an increased risk of cancer for DPP-4 inhibitors added to standard therapy.302-304 The FDA prescribing information for the GLP-1 receptor agonists, liraglutide,315 albiglutide,316 exenatide,317 and dulaglutide318 includes a warning regarding the potential for a link between these agents and medullary thyroid cancer, based on data in mice and rats.319 The evidence that we identified on the incretin-based therapies and cancer was of low strength and inconsistent. However, we found low strength of evidence that the combination of metformin plus a sulfonylurea was favored over the combination of metformin plus a DPP-4 inhibitor for longer-term cancer risk. An unpublished study’s results, as well as longer-term followup of one of the included published studies,141 were consistent with this finding and may have increased this evidence to moderate strength. Congestive Heart Failure There was only one long-term 4-year RCT and only a few observational studies of medium quality with 6 to 8 years of followup that could provide a comparative assessment of the safety of diabetes medications on congestive heart failure. We found low strength of evidence of 1.2 to 1.6 increased odds of heart failure with the thiazolidinedione class of medications, when compared with sulfonylureas or metformin. Our strength of evidence on this outcome dropped to low in this update (from moderate in the prior review), because we excluded lower-quality observational studies and also excluded the RECORD trial for this outcome, owing to the active comparator being either sulfonylurea or metformin, instead of a single active comparator. RECORD showed that the combination of thiazolidinediones and another agent (sulfonylurea or metformin) was associated with a significant doubling in the risk of heart failure in comparison with the combination of sulfonylurea and metformin (61/2220 versus 29/2227, risk ratio (RR), 2.1; 95% CI, 1.35 to 3.27).49 These results showing a higher risk of congestive heart failure with thiazolidinediones were also confirmed in two recent meta-analyses.320, 321 Both thiazolidinediones, rosiglitazone and pioglitazone, are contraindicated in patients with serious or severe heart failure (Stage 3 or Stage 4), according to product labels.322, 323 We had low or insufficient strength of evidence for most other medication comparisons for heart failure, including the newer agents. Despite recent concerns about congestive heart failure with DPP-4 inhibitors, we found low or insufficient strength of evidence on the comparative
- 370. 313 safety of thisdrug class for this outcome in mainly short duration studies [five short duration RCTs reporting no events in the DPP-4 inhibitor arms, one short duration RCT with one event in the metformin plus DPP-4 arm and none in the comparator arm, and one RCT which reported few congestive heart failure events in the metformin plus DPP-4 inhibitor arm compared with the metformin plus sulfonylurea arm (three versus six respectively)]. Several large, double-blind, placebo-controlled RCTs evaluating DPP-4 inhibitors on cardiovascular outcomes in adults with moderate to high cardiovascular risk were excluded from our systematic review of head-to-head comparisons but deserve mention due to recent controversy.302, 303, 324 Two of these RCTs (comparing either saxagliptin or alogliptin with placebo) reported a small increased risk of hospitalization for congestive heart failure in adults at moderate to high cardiovascular risk (between-group absolute risk differences of 0.7% and 0.9%).302, 324 The EXAMINE trial with the alogliptin comparison reported these differences solely for the outcome of first hospitalization for heart failure in adults without pre-existing congestive heart failure as part of a posthoc subgroup analysis.324 The third placebo-controlled RCT303 compared sitagliptin with placebo on cardiovascular outcomes and reported no between-group differences in hospitalization for congestive heart failure. It is unclear if differences in these trials were due to differences in drug type, chance, or other causes. Due to these findings, however, the FDA has requested additional labeling for saxagliptin and alogliptin to reflect concerns of the potential increased risk of hospitalization for congestive heart failure.325 Further research directly comparing DPP-4 inhibitors with other active comparators on heart failure outcomes will be useful in determining the comparative safety of these medications on heart failure risk, including results of two RCTs [the Cardiovascular Outcome Study of Linagliptin Versus Glimepiride in Patients with Type 2 Diabetes (CAROLINA) and the Cardiovascular and Renal Microvascular Outcome Study with Linagliptin in Patients with Type 2 Diabetes Mellitus (CARMELINA) studies] of linagliptin, which are in progress.326, 327 Liver Injury Similar to the 2011 report, we found little evidence on liver injury. Compared to the prior report showing no between-group differences in liver injury, we downgraded the available evidence for metformin versus thiazolidinedione monotherapy (from moderate to low) and downgraded the evidence for thiazolidinedione versus sulfonylurea monotherapy (from high to low). Notably, there are FDA warnings of post-marketing cases of hepatic failure for both alogliptin328 and pioglitazone,329 but we found low or insufficient strength of evidence for thiazolidinedione- and DPP-4 inhibitor-based comparisons and liver injury. Lactic Acidosis Prior evidence on the elevated risk of lactic acidosis with phenformin, an earlier biguanide, and case reports of lactic acidosis among metformin users have led to continued concern about an increased risk of lactic acidosis with metformin; however, for most of the ~300 case reports on metformin and lactic acidosis, other factors contributing to lactic acidosis could not be excluded (e.g., acute myocardial infarction330, 331 or acute kidney failure). Consistent with the prior report16 and a Cochrane review on this topic,332 we did not find an increased risk of lactic acidosis with metformin based on the little evidence identified. A more recent systematic review by Inzucchi et al.333 evaluated the risk of lactic acidosis associated with metformin use in adults with mild to moderate chronic kidney disease (CKD; estimated glomerular filtration rates of 30- 60 mL/min per 1.73 m2 ). Using data from 65 studies (mainly observational), they reported an
- 371. 314 overall incidence of3-10 per 100,000 person-years of lactic acidosis in metformin users across studies,333 which is similar to the background prevalence in adults with diabetes not on metformin.332 The FDA is currently reviewing two citizen petitions to expand the use of metformin to adults with diabetes and mild to moderate CKD, with potential dose reductions in metformin to enhance safety in these populations. Pancreatitis Compared to the prior report, we identified many more studies on pancreatitis but found low strength of evidence for most comparisons. The DPP-4 inhibitors and GLP-1 receptor agonists were of most interest for this outcome, given the spontaneous reports to the FDA of pancreatitis associated with these agents. In the three large, placebo-controlled RCTs described above (SAVOR TIMI 53, EXAMINE, TECOS),302-304 more cases of pancreatitis have been observed in the DPP-4 inhibitor than placebo arms with a consistent absolute risk difference of 0.1 percent (number needed to harm of 1,000). In the SAVOR TIMI 53 trial, definite acute pancreatitis occurred in 17 participants (0.2 percent) in the saxagliptin arm and nine participants (0.1 percent) in the placebo arm (P =0.17).302 In EXAMINE, acute pancreatitis occurred in 12 participants (0.4 percent) in the alogliptin arm and in eight participants (0.3 percent) in the placebo arm. Finally, in TECOS, acute pancreatitis occurred in 23 participants (0.3 percent) and in 12 participants (0.2 percent) in the sitagliptin and placebo arms, respectively (HR, 1.93; 95% CI, 0.96 to 3.88; P =0.07). We excluded some of the Liraglutide Effect and Action in Diabetes (LEAD) RCTs from this report, because they did not evaluate comparisons of interest;334 seven participants exposed to liraglutide and one exposed to a sulfonylurea were diagnosed with pancreatitis across the six LEAD trials.335 Systematic reviews of RCTs336 and observational studies337 have reported small, non- statistically significant increases in the relative odds of pancreatitis for incretin-based therapies versus control treatments. However, these meta-analyses are underpowered for this rare outcome, grouped GLP-1 receptor agonists and DPP-4 inhibitors together, grouped comparators (placebo and different active treatments), and included studies of patients without diabetes. Severe Allergic Reactions In the prior report, we did not find evidence on severe allergic reactions. However, the issue of hypersensitivity reactions with diabetes medications has become more prominent with the uptake of DPP-4 inhibitors338 and GLP-1 receptor agonists.339-341 In March, 2012, the FDA added a warning about the risk of hypersensitivity reactions with DPP-4 inhibitors.342 Although still of low strength, we found the strongest evidence (based on four RCTs) that the addition of a DPP-4 inhibitor to metformin increases the risk of hypersensitivity reactions over metformin monotherapy alone. Prior data on the risk of hypersensitivity reactions with DPP-4 inhibitors have been mixed.338 The SAVOR-TIMI 53 trial found similar rates of hypersensitivity reactions for saxagliptin compared with placebo.302 In EXAMINE, angioedema was more frequent for the alogliptin arm (17/2701, 0.6%) than for the placebo arm (13/2679, 0.5%).304 TECOS did not report on this outcome.303 Macular Edema and Decreased Vision We did not find conclusive evidence on outcomes of macular edema and decreased vision. A concern regarding the risks of macular edema with the thiazolidinediones persists and is based primarily on observational studies.343-345 Seven compared with three participants in the
- 372. 315 rosiglitazone and activecomparator arms, respectively, developed macular edema in the RECORD trial (described above), although there were more than 2000 treated in each arm.49 Adverse Events Specific to SGLT-2 Inhibitors Our findings of an increased risk of genital mycotic infections for SGLT-2 inhibitors compared to other agents are consistent with recent reviews of this topic.282, 346 The existing systematic reviews did not evaluate comparative effectiveness but, instead, grouped comparators together for synthesis. In contrast to one of these recent systematic reviews of SGLT-2 monotherapy,282 but consistent with another,346 we did not find an increased risk of urinary tract infection (except for low strength of evidence in women for SGLT-2 inhibitor monotherapy relative to DPP-4 inhibitors) or volume depletion events; of note, all evidence in this report was of low strength (or insufficient) for urinary tract infection and volume depletion. We also evaluated fracture risk and renal insufficiency associated with SGLT-2 inhibitors, given the issues raised about these adverse events in Vasilakou et al. 2013; in that review, the authors did not make conclusions about these outcomes, as their data were limited. We did not identify substantial evidence on these outcomes either. However, on September 10, 2015, the FDA strengthened its warning of an increased risk of fractures with canagliflozin based on pooled data from nine clinical trials.347 The risk of fracture was increased for canagliflozin (1.4/100 patient-years and 1.5/100 patient-years for canagliflozin 100 mg and canagliflozin 300 mg respectively) versus the comparator (1.1/100 patient-years for placebo and active comparators combined) with a mean follow up of 85 weeks across trials. The labeling for canagliflozin notes that clinicians should consider factors that increase fracture when starting canagliflozin.348 The FDA issued a warning regarding the risk of ketoacidosis with SGLT-2 inhibitors on May 15, 2015.325 We did not evaluate this outcome, as it was not a concern at the time that we selected outcomes for this report. The FDA warning stemmed from the observation of 20 cases of ketoacidosis in patients taking SGLT-2 inhibitors recorded in the FDA Adverse Event Reporting System (FAERS) database through June 6, 2014, followed by continued reports of ketoacidosis in patients taking SGLT-2 inhibitors since that time. A recent analysis of 17,596 participants from randomized trials of canagliflozin (mainly placebo-controlled), with 24,000 patient-years of exposure, demonstrated a higher number of patients experiencing ketoacidosis in the canagliflozin versus comparator arms: canagliflozin 100 mg: 4 (0.07 percent); canagliflozin 300 mg: 6 (0.11 percent); and comparator: 2 (0.03 percent).349 The authors noted that six of the 10 patients with ketoacidosis in the canagliflozin arms were found to have type 1 diabetes, latent autoimmune diabetes of adulthood (LADA) or antibodies to GAD65.349 The FDA has not changed labeling of the SGLT-2 inhibitors, at this time.325 Subgroups The limited evidence on outcomes in subgroups was for the outcome of HbA1c and did not show differential effects of the included comparisons on glycemic control by age, sex, race/ethnicity, or body mass index. Otherwise, the evidence on the comparative effectiveness of diabetes medications in subgroups defined by age, sex, race/ethnicity, and body mass index was generally inconclusive. This is especially unfortunate for the age and race/ethnicity sub- populations because of the known disparities in diabetes prevalence and diabetes outcomes for these groups. Older Americans suffer disproportionately from diabetes, with over 25 percent of
- 373. 316 persons 65 yearsof age and older having diabetes compared to 16 percent of persons 45 years of age to 64 years of age, and there is concern about the safety of medications (and polypharmacy) in older adults.1, 8 Also, compared to non-Hispanic white adults in the United States, diabetes is 20 percent more common in Asian Americans, 70 percent more common in Hispanics and non- Hispanic blacks, and twice as common in American Indians/Alaska Natives.1 Racial and ethnic minorities are also more likely to suffer from diabetes complications, including diabetic end- stage renal disease,350 retinopathy,351 amputations, hospitalization for cardiovascular outcomes,352 and diabetes-related mortality.353 Finally, racial and ethnic minorities are less likely to have controlled diabetes (HbA1c <7 percent), but are more likely to be on oral treatment only for diabetes.354 Applicability The applicability of these studies depends largely on the similarity of the study populations to the U.S. population with type 2 diabetes and the similarity of the interventions to usual clinical care (e.g., comparability of the drug interventions including dosing and duration of exposure to drugs). The included studies generally had populations, interventions, outcomes, and settings applicable to U.S. adults with type 2 diabetes, with a few notable exceptions, as described below. Study population differences are the most pronounced threat to applicability. Study participants were mainly middle-aged (mean age in the mid 50s), overweight or obese adults who had diabetes for 3 to 7 years at the start of the studies. This is similar to the U.S. population with type 2 diabetes, which has a mean age of 60.5 years and a mean body mass index of 33 kg/m2 (23.5 percent overweight, 65.3 percent obese).355 However, most studies excluded people older than 75 or 80 years of age and excluded people with significant renal, hepatic, and cardiovascular disease, and other significant co-morbid conditions, making these studies less applicable, given that 52 percent of US adults with diabetes are older than 60 years of age, and just over 25 percent have a history of cardiovascular disease.355 When race was reported in the included studies, most subjects were Caucasian, although about 10 to 20 percent of study participants were of other races. These studies are, therefore, less applicable to people of different races and ethnicities, who make up about 40 percent of the US population with diabetes and, importantly, these groups have a greater diabetes burden than Caucasians (i.e., African Americans, Hispanics, Asian Americans, and American Indians).1, 355 Characteristics of the interventions also impact applicability, and most studies used dosing, frequency, and monitoring comparable to usual care. However, a threat to applicability relates to the duration of drug exposure, especially for glycemic control. The vast majority of RCTs lasted for 2 years or less. In usual care, patients with diabetes are on medications for over 10 years and are on multiple medications that impact adherence and side effects. Also, the glycemic response to medications may degrade over time; retained insulin sensitivity may allow insulin sensitizers (like metformin) to work longer as monotherapy than medications that are not insulin sensitizers. Also, in roughly one-third of the included trials, rescue therapy was used if participants did not meet specific glycemic goals, and participants were often censored from the study at that time. Thus, the results of these studies may not reflect what will occur with the clinical usage of the studied medications. We generally had few concerns regarding applicability of the trial settings to usual care. While many trials did not take place exclusively in the United States, they did occur in similar settings. About half the trials occurred partly or exclusively in the United States, Italy, and/or were multinational; the rest of the trials occurred in developed or newly industrialized countries.
- 374. 317 However, few trials(about 25 percent) reported on the setting of recruitment, such as primary care or specialty care, so we cannot definitively comment on how like this is to usual care. Implications for Clinical and Policy Decisionmaking This update provides additional evidence supporting metformin as the first-line medication therapy to treat type 2 diabetes, when tolerated; the evidence also supports the addition to metformin of a number of treatment options, based on patient preferences. Not only is metformin favored on the intermediate outcomes of HbA1c and weight, and not associated with serious adverse events, we found more conclusive evidence to support that cardiovascular mortality is lower with metformin compared with sulfonylureas. This evidence supports current guidelines, such as the American College of Physicians356 and American Diabetes Association26 guidelines, which recommend metformin as a first-line treatment choice. The American Association of Clinical Endocrinologists19 guideline also lists metformin as one of its first-line choices for treatment of type 2 diabetes, although it allows more flexibility in the choice of first-line therapy. Metformin is currently contraindicated in patients with “renal disease or renal dysfunction,”357 because of concerns for an increased risk of lactic acidosis in this population. However, as described above, this risk is small and may not be higher than the background risk of lactic acidosis for patients with type 2 diabetes.332, 333 Twenty-two percent of patients with type 2 diabetes in the United States are estimated to have at least mild chronic kidney disease, indicating a large group of patients with type 2 diabetes who are not currently candidates for metformin therapy.358 Furthermore, some patients with type 2 diabetes are unable tolerate the side effects of metformin. The selection of initial diabetes therapy is an important clinical question for this relatively large population in which metformin is contraindicated or not tolerated. We evaluated non-metformin-based monotherapy comparisons in this report and demonstrated that, with the exception of DPP-4 inhibitors, which are not as effective in reducing HbA1c as metformin, the other monotherapies generally decrease HbA1c similarly (and comparably) to metformin. As described in detail, the other monotherapies’ effects on weight vary as do their adverse effects, such as congestive heart failure (increased risk for thiazolidinediones), hypoglycemia (highest risk with sulfonylureas, including for severe hypoglycemia for many comparisons), gastrointestinal side effects (nausea and vomiting with GLP-1 receptor agonists), and genital mycotic infections (increased risk for SGLT-2 inhibitors). Most importantly, we do not have conclusive evidence on the relative long-term effects of non- metformin-based monotherapy on all-cause mortality or cardiovascular outcomes and rare, serious adverse events (e.g., pancreatitis risk with GLP-1 receptor agonists). Therefore, the alternative to metformin initial therapy is unclear and, seemingly, must be based on individual patient factors (e.g., HbA1c goal, risk of hypoglycemia) and preferences (e.g., avoidance of weight gain, cost). Similarly, our evaluation of metformin-based combination therapies provides some insight into the selection of add-on therapy to metformin but is not definitive, because of the uncertainty of long-term outcomes and differential effects on weight and side effects. Comparisons of the metformin-based combinations suggested similar HbA1c-lowering for the metformin-based combination therapies (except for DPP-4 inhibitors added to metformin having a smaller HbA1c- lowering effect), differential weight effects, highest hypoglycemia risk with metformin plus a sulfonylurea, highest risk of gastrointestinal side effects with metformin plus GLP-1 receptor agonists, and increased risk of genital mycotic infections with metformin plus an SGLT-2 inhibitor.
- 375. 318 As the newermedications (DPP-4 inhibitors, SGLT-2 inhibitors, and GLP-1 receptor agonists) remain on the market, become available as generics, and have additional data on comparative efficacy (for long-term outcomes) and safety, these newer medications may be preferred by patients. Therefore, the continued emphasis in guidelines about accounting for patient preferences when choosing therapy will be critical.26 In terms of cost, pioglitazone is the newest agent that has a generic. The first patent for Byetta expires in December 2016 and Januvia will have a patent expiry in 2017. If generics are available soon after, this will give patients and clinicians more affordable options for therapy. In summary, we did not find large differences in HbA1c-lowering effects of the diabetes medications studied, except for DPP-4 inhibitors, which are not as effective as metformin. Weight effects of the medications are differential, and there is only evidence on cardiovascular mortality to support metformin over sulfonylureas as monotherapy. Each class of drug has different side effects (e.g., hypoglycemia, GI side effects, congestive heart failure), and the evidence on rare, serious side effects is less strong. Therefore, factors such as patient preferences and costs are likely to continue to drive selection of and adherence to the diabetes medications. Limitations of the Comparative Effectiveness Review Process Several important limitations to our updated systematic review deserve mention. Although this was an update of a comprehensive review published in 2007 and an update in 2011, we focused this update a priori on studies with active control comparators, which are most relevant for clinical practice. Placebo-controlled trials had been included in the original 2007 review but excluded in the 2011 update. In general, the majority of placebo-controlled trials are short. However, the exclusion of placebo-controlled trials has implications for the review, including the inability to evaluate rare outcomes using indirect comparisons. To conclude from an active- control study that one medication is more effective than another requires prior knowledge that the active-control drug has been studied previously and is known to be more effective than placebo. Because the 2007 review had included placebo-controlled trials, we know that many drugs were more effective than placebo for the intermediate outcomes for many drug comparisons. However, this assumption may be less valid for the newer medications, where evidence on comparisons with placebo from other systematic reviews, such as the Cochrane Reviews,277, 283 will also be helpful in making conclusions. In addition, our inclusion criteria required that all studies fit into one or more of the pre- specified comparisons of interest (Table 2), which identified specific drug-drug or two-drug comparisons. For example, studies that included any number of “background medications” were excluded. Our rationale was to avoid attributing outcomes to the medication of interest when it was truly due to the background medication. This was especially important because of our goal of evaluating two-drug combinations. Applying the inclusion criteria, which required pre- specified comparisons of interest, had several implications. This criteria required the exclusion of several large trials,9-12, 312, 359-363 some of which evaluated HbA1c-lowering strategies rather than individual medications, as well as some smaller trials and observational studies. Of note, the RECORD study49 was included for the intermediate outcomes but excluded for the long-term and safety outcomes, because it did not stratify these outcomes by comparisons of interest. While excluding this study for these outcomes lowered our evidence grade for congestive heart failure, it did not change the overall conclusion. Another consequence of requiring direct comparisons of interest was that some of the recent studies of exenatide174, 364-366
- 376. 319 and liraglutide261, 367-369 asadd-on therapy to metformin did not have a specific comparison of interest and were therefore excluded. However, these studies would not have changed our findings as HbA1c were similar to those observed in the included studies, and these studies did not report on mortality, cardiovascular outcomes, pancreatitis, or cancer – outcomes of particular interest for these agents. We also had strict criteria for including only medium- to high-quality observational studies. For instance, we required observational studies to have accounted for confounding by age, gender, race/socioeconomic status, and co-morbid conditions. The article could have used propensity score methodologies or other appropriate methods to account for differences between groups, or could have restricted to one race or socioeconomic status, for examples. By excluding observational studies with a higher risk of bias, we included only observational studies that could provide the most valid results. This resulted in the exclusion of many observational studies of harms, which could have strengthened the evidence base, but this was necessary to reduce confounding by indication in these studies. We selected Key Questions focused on intermediate and long-term clinical outcomes through a topic refinement at the beginning of this process, which involved input from stakeholders on the Technical Expert Panel. Diabetes care is an extensive field, and we note the omission of key outcomes. For example, we did not collect information about patient-reported outcomes, such as medication adherence and barriers to adherence, health-related quality of life, or treatment satisfaction. These outcomes are important, because they may mediate the efficacy of treatment, and also are valuable to patients and clinicians. Future reviews with methodologies designed to capture many different study designs, including qualitative studies, and use of a wide range of measures, are needed to address these outcomes. For microvascular outcomes, we included studies evaluating proximal measures such as change in retinal exam or changes in microalbuminuria which may be less clinically relevant than other microvascular outcomes of blindness and changes in estimated glomerular filtration rate. However, we were unable to conclude anything about the comparative effects on the microvascular outcomes due to lack of sufficient evidence. These distinctions may become more important as more evidence accrues on these different microvascular outcomes. Although we assessed the mean difference in HbA1c between intervention groups in Key Question 1, we did not include the durability of HbA1c changes over time as an outcome, which may best be addressed using long-term well-designed observational studies. In terms of pooling results, we chose to combine similar studies for pooled estimates. For study duration, we chose to combine similar duration studies which were often less than or equal to 52 weeks. While between-group differences might vary between 12, 24 and 52 week studies, we did not find substantial clinical or statistical heterogeneity related to these study duration differences in our pooled analyses. We also chose to combine results by drug class for most comparisons, except where clinically indicated (e.g., separated rosiglitazone and pioglitazone in cardiovascular comparisons). This may have led us to miss small differences within a drug class. However, if there was clinical or statistical heterogeneity noted among the studies, we evaluated for differences by drug type. For instance, we did not combine GLP-1 receptor agonists together in the hemoglobin A1c section (unless all studies used a single drug within the class, such as exenatide) owing to potential differences by drug type in glycemic control. This potential clinical heterogeneity was noted prior to combining the individual medications and was also identified when examining statistical heterogeneity among those studies. Also, in the 2007 report,15 we found that glyburide/glibenclamide had a higher absolute risk difference of mild, moderate, or
- 377. 320 total hypoglycemia comparedwith other sulfonylureas (pooled RD 3%; 95% CI, 0.5% to 5%). In this update, which focused on interclass comparisons, the studies that included glyburide/glibenclamide as the sulfonylurea did not consistently have larger between-group differences in hypoglycemia risk compared to the other sulfonylurea studies. Therefore, these studies were combined with the other sulfonylurea comparisons for hypoglycemia. Limitations of the Evidence Base The major limitation of the evidence base was a lack of evidence supporting conclusions on the comparative long-term (followup at least 2 years) clinical (mortality, cardiovascular outcomes, and microvascular outcomes) and safety outcomes of the medications of interest. Given the low event rates for these outcomes and the timeframe in which they develop, RCTs, while extremely helpful, do not feasibly provide all of the evidence on long-term outcomes. Once we applied selection criteria to account for confounding by key factors, including confounding by indication, an important threat to validity in this setting, we did not identify many observational studies on the long-term and rare outcomes. Given the resources needed, not surprisingly, we did not identify any RCTs designed to evaluate long-term outcomes as the primary outcome. The included RCT evidence was underpowered for these outcomes based on the combination of small sample sizes, low event rates, and short study durations (generally 12 months or less). Substantial losses to followup, often differential across study arms, were another major limitation to the evidence on long-term clinical and safety outcomes. Additional limitations of the evidence base on long-term and safety outcomes included lack of reporting on these outcomes (including lack of reporting across all study arms), lack of active ascertainment of safety outcomes, and lack of an intention-to-treat approach. As expected, the vast majority of included RCTs were industry-sponsored, raising the possibility of publication bias and other forms of bias, such as selective reporting of outcomes. While publication bias and reporting bias generally were not found, publication bias analyses have limited power, owing to the small numbers of studies for any given comparison. Although we cannot conclude that bias was present, we have to be especially concerned about the following issues identified across the included RCTs (which are important regardless of sponsorship): For the long-term clinical and safety outcomes, many studies reported an event in one arm but not in the comparator arm, making it challenging to compare medications. Several studies failed to report the significance of between-group differences and the measures of dispersion, thereby hindering efforts to estimate effect size across trials for intermediate outcomes. Some trials compared medications using dissimilar doses, limiting our ability to draw conclusions about efficacy. Also, many studies had high rates of withdrawals; even if the studies described the rates of withdrawals, they often did not use a valid method for accounting for missing data.370 Finally, authors of the included randomized trials often did not describe their method of randomization and often did not describe double-blinding, making it difficult to appropriately assess risk of bias of individual studies.
- 378. 321 Research Gaps andFuture Research Needs Based on the limitations of the evidence base, using the PICOT framework, we highlight several major gaps in the evidence in Table 127. We report these for all of the Key Questions (Key Questions 1-4; comparative effectiveness and safety) together, to avoid duplicating research gaps that apply to more than one Key Question. We also added a specific “methodologic” category to complement the content-oriented research gaps. We provide recommendations on future research needs corresponding to these research gaps (Table 127). In particular, we want to highlight the importance of future research using high-quality observational studies to determine the comparative effects of diabetes medications on long-term clinical and safety outcomes. Multi-year (or decade) trials are often infeasible. Supplementing the rare RCT that can be conducted for these outcomes with truly high-quality observational studies is paramount. We propose that, at a minimum, such observational studies will need to follow patients over time, analyze similar comparison groups, and account for confounding by indication (including duration of diabetes and co-morbid conditions). Databases with sufficient sample size, followup over time, data on treatments (including doses and duration), and confounders, such as demographics, duration of diabetes, and co-morbid conditions, will be necessary. A recent review by Patorno et al. provided a thorough evaluation of threats to the validity of observational studies of diabetes medications and cardiovascular outcomes and outlined approaches to avoiding these biases.371 Briefly, the following are major methodological pitfalls and strategies to avoiding biases in the conduct of future observational studies of the comparative effectiveness and safety of diabetes medications371 : Confounding by indication: Basic variables which must be considered to reduce confounding by indication include demographics, duration of diabetes, and co-morbid conditions. Many statistical methods may be sufficient (e.g., multivariate regression, restriction, instrumental variables), but propensity score methods may be the strongest to handle confounding by indication. In particular, high-dimensional propensity score algorithms may be the most rigorous, as they can help deal with unmeasured confounders. Immortal time bias: Prospective studies that define cohort entry based on exposure to a drug (versus calendar time or diagnosis of diabetes during a specified window), that have covariate information prior to exposure, and that do not condition cohort entry on events that occur during followup (e.g., initiation of insulin) are most likely to avoid immortal time bias. Time- and cumulative exposure-varying incidence of outcomes: The effects of medications on the outcomes of interest may vary over time and with cumulative drug exposure; study designs evaluating new users of drugs and accounting for exposure time will minimize biases due to time- and cumulative exposure-dependent effects of medications. Reverse causation: Analyses allowing for lag time after exposure can reduce the chance of reverse causality. Informative censoring: Censoring of observations when a drug exposure stops may lead to informative censoring, because reasons related to drug continuation may also be related to the outcome of interest. Analyses accounting for latency of drug effects can address this issue.
- 379. 322 Time-varying drugexposure: Although time-to-event analyses are often preferred to evaluate risk factors for outcomes such as cardiovascular disease, in the case of diabetes, drug changes may be related to the outcome. Sensitivity analyses can be used to support analyses based on time-to-event analyses. Time-dependent confounders: Inclusion of important confounding factors, such as co- morbid conditions that change over time, and the use of statistical methods, such as marginal structural models, may be helpful in handling such confounders.
- 380. 323 Table 127. Evidencegaps and future research needs for the comparative effectiveness and safety of diabetes medications for adults with type 2 diabetes PICOT Category Evidence Gap Future Research Needs Population Lack of study of older adults, racial/ethnic minorities, and persons with co-morbid conditions such as significant renal, cardiovascular and hepatic impairment. Limited evidence on a priori subgroups of interest such as older adults, racial/ethnic minorities, sex, and BMI Studies which include diverse populations Studies with an a priori plan to investigate differences by important subgroups of interest Interventions & Comparators HbA1c, weight, hypoglycemia and GI adverse events Limited information on GLP-1 receptor agonist comparisons as monotherapy and in combination with metformin versus specific diabetes medication comparators. Limited information on metformin plus insulin versus other metformin- based combinations. RCTs evaluating the GLP-1 receptor agonists as monotherapy and in combination with metformin. If adding GLP-1 receptor agonists to different background medications, then RCTs should conduct stratified randomization by background medication and evaluate effects by background medication. RCTs evaluating intermediate outcomes for metformin plus the addition of insulin with other metformin-based combinations, and in particular metformin plus a GLP- 1 receptor agonist would be useful for patients and clinicians contemplating an add-on injectable to metformin. Outcomes All-cause mortality and macrovascular and microvascular outcomes Limited information on macrovascular outcomes and death Existing evidence underpowered Limited number of high-quality observational studies No conclusive evidence on microvascular outcomes No RCTs evaluated these outcomes as a primary outcome Inconsistent outcome definitions, ascertainment, and reporting in each study arm High-quality observational studies for all comparisons Longer duration RCTs for all comparisons evaluating macrovascular and microvascular events as primary outcomes Standardized definitions for macrovascular and especially microvascular outcomes (e.g., incident nephropathy based on eGFR and urine albumin:creatinine ratios) Reporting on outcomes in all arms of RCTs
- 381. 324 Table 127. Evidencegaps and future research needs for the comparative effectiveness and safety of diabetes medications for adults with type 2 diabetes (continued) PICOT Category Evidence Gap Future Research Needs Rare safety outcomes Existing evidence underpowered Lack of high-quality observational studies Inconsistent outcome definitions, ascertainment, and reporting in each study arm, especially for pancreatitis and cancer No conclusive evidence on any of the following adverse events: CHF, pancreatitis, cancer, liver injury, lactic acidosis, severe allergic reactions, macular edema No conclusive evidence on volume depletion for SGLT-2 inhibitor comparisons High-quality observational studies for rare outcomes* Specific safety outcomes and drugs require further study including the following: o CHF – DPP-4 inhibitors o Macular edema – TZDs o Pancreatitis – DPP-4 inhibitors and GLP-1 receptor agonists o Thyroid cancer – GLP-1 receptor agonists o Volume depletion – SGLT-2 inhibitors o Ketoacidosis – SGLT-2 inhibitors RCTs o Active ascertainment of all safety outcomes o Standardized definitions for all safety outcomes o Reporting on safety outcomes in all arms Timing Most evidence is for short-term outcomes as few studies lasted more than 2 years Longer duration studies (>2 years) o To determine durability of short-term comparative effects on HbA1c and weight o To determine long-term clinical effectiveness (e.g., all-cause mortality and cardiovascular outcomes) and safety Methodological High, and often differential, losses to followup in RCTs Lack of reporting on randomization methods (for RCTs) Lack of reporting on allocation concealment, blinding, and withdrawals for all studies Lack of appropriate accounting for confounding in observational studies Lack of reporting on treatments in observational studies Complete or near-complete followup in RCTs (focus on retention) Appropriate methods to account for losses to followup if needed (e.g., multiple imputation) Reporting on methods for randomization, allocation concealment, and blinding in RCTs High-quality observational studies for long-term comparative effectiveness and safety of diabetes medications* BMI = body mass index; CHF = congestive heart failure; DPP-4 = dipeptidyl peptidase-4; eGFR = estimated glomerular filtration rate; GI = gastrointestinal; GLP-1 = glucagon- like peptide-1; HbA1c = hemoglobin A1c; RCT = randomized controlled trial; SGLT-2 sodium-glucose co-transporter-2; TZD = thiazolidinediones *See text above for more detail.
- 382. 325 Conclusions Although the comparativelong-term benefits and harms of most diabetes medications remain unclear, the evidence supports use of metformin as a first-line agent, because of its beneficial effects on HbA1c, weight, and long-term outcomes (cardiovascular mortality benefit for metformin versus sulfonylureas, in particular), and its relative safety. With the exception of DPP- 4 inhibitors, which have smaller effects on HbA1c reduction compared to metformin, the HbA1c-lowering of the other diabetes medications are similar, for monotherapy and metformin- based combination comparisons. The alternative to metformin monotherapy is unclear because of a lack of evidence on long-term effectiveness and safety outcomes on the other monotherapy comparisons. Likewise, the comparative effectiveness of metformin-based combinations for long-term macrovascular, microvascular, and rare safety outcomes is unclear. Monotherapy and metformin-based combination comparisons have differential effects on weight and side effects (e.g., hypoglycemia, GI side effects). In this report, we provide comprehensive information on the relative benefits and harms of diabetes medications to inform personalized treatment choices by patients and their clinicians, as well as to support decisionmaking by payers and regulators.
- 383. 326 References 1. Centers forDisease Control and Prevention. National Diabetes Statistics Report: Estimates of Diabetes and Its Burden in the United States, 2014. Atlanta, GA: U.S. Department of Health and Human Services; 2014; http://www.cdc.gov/diabetes/pubs/statsreport14/n ational-diabetes-report-web.pdf. Accessed 2015 February 26. 2. Lee JW, Brancati FL, Yeh HC. Trends in the prevalence of type 2 diabetes in Asians versus whites: results from the United States National Health Interview Survey, 1997- 2008. Diabetes Care. 2011 Feb;34(2):353-7. PMID: 21216863. 3. Nichols GA, Schroeder EB, Karter AJ, et al. Trends in diabetes incidence among 7 million insured adults, 2006-2011: the SUPREME-DM project. Am J Epidemiol. 2015 Jan 1;181(1):32-9. PMID: 25515167. 4. Maruthur NM. The growing prevalence of type 2 diabetes: increased incidence or improved survival? Curr Diab Rep. 2013 Dec;13(6):786-94. PMID: 24072478. 5. Centers for Disease Control and Prevention. Diabetes Public Health Resource: Incidence ang Age at Diagnosis. 2013; http://www.cdc.gov/diabetes/statistics/incidence_ national.htm. Accessed 2015 January 27. 6. Anon. Economic costs of diabetes in the U.S. In 2007. Diabetes Care. 2008 Mar;31(3):596-615. PMID: 18308683. 7. Anon. Standards of medical care in diabetes- 2014. Diabetes Care. 2014;37(SUPPL.1):S14-S80. PMID: 24357209. 8. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of hyperglycemia in type 2 diabetes: A patient-centered approach: Position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Spectrum. 2012;25(3):154-71. PMID: 22517736 9. Anon. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998 Sep 12;352(9131):854-65. PMID: 9742977. 10. Anon. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998 Sep 12;352(9131):837-53. PMID: 9742976. 11. Duckworth W, Abraira C, Moritz T, et al. Glucose control and vascular complications in veterans with type 2 diabetes. The New England journal of medicine. 2009 Jan 8;360(2):129-39. PMID: 19092145. 12. Gerstein HC, Miller ME, Byington RP, et al. Effects of intensive glucose lowering in type 2 diabetes. The New England journal of medicine. 2008 Jun 12;358(24):2545-59. PMID: 18539917. 13. Papademetriou V, Lovato L, Doumas M, et al. Chronic kidney disease and intensive glycemic control increase cardiovascular risk in patients with type 2 diabetes. Kidney Int. 2014 Sep 17. PMID: 25229335. 14. Saremi A, Schwenke DC, Bahn G, et al. The effect of intensive glucose lowering therapy among major racial/ethnic groups in the Veterans Affairs Diabetes Trial. Metabolism. 2015 Feb;64(2):218-25. PMID: 25456099. 15. Bolen S, Wilson L, Vassy J, et al. Comparative Effectiveness and Safety of Oral Diabetes Medications for Adults with Type 2 Diabetes. Comparative Effectiveness Review No 8 (Prepared by the Johns Hopkins Evidence-based Practice Center under Contract No 290-02-0018). Rockville, MD: Agency for Healthcare Research and Quality; 2007.
- 384. 327 16. Bennett WL,Wilson LM, Bolen S, et al. Oral Diabetes Medications for Adults With Type 2 Diabetes: An Update. Comparative Effectiveness Review No. 27. (Prepared by Johns Hopkins University Evidence-based Practice Center under Contract No. 290-02- 0018.) AHRQ Publication No. 11-EHC038- EF. Rockville, MD: Agency for Healthcare Research and Quality; 2011. 17. Bolen S, Feldman L, Vassy J, et al. Systematic review: comparative effectiveness and safety of oral medications for type 2 diabetes mellitus. Ann Intern Med. 2007 Sep 18;147(6):386-99. PMID: 17638715. 18. Bennett WL, Maruthur NM, Singh S, et al. Comparative effectiveness and safety of medications for type 2 diabetes: an update including new drugs and 2-drug combinations. Ann Intern Med. 2011 May 3;154(9):602-13. PMID: 21403054. 19. Garber A, Abrahamson M, Barzilay J, et al. American association of clinical endocrinologists' comprehensive diabetes management algorithm 2013 consensus statement. Endocrine Practice. 2013;19(SUPPL. 2):1-48. PMID: 23816937. 20. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of Hyperglycemia in Type 2 Diabetes, 2015: A Patient-Centered Approach: Update to a position statement of the american diabetes association and the european association for the study of diabetes. Diabetes Care. 2015;38(1):140-9. PMID: 25538310 21. Type 2 Diabetes: National Clinical Guideline for Management in Primary and Secondary Care (Update). London: Royal College of Physicians of London.; 2008. 22. Colhoun HM, Livingstone SJ, Looker HC, et al. Hospitalised hip fracture risk with rosiglitazone and pioglitazone use compared with other glucose-lowering drugs. Diabetologia. 2012 Nov;55(11):2929-37. PMID: 22945303. 23. Lu CJ, Sun Y, Muo CH, et al. Risk of stroke with thiazolidinediones: a ten-year nationwide population-based cohort study. Cerebrovasc Dis. 2013;36(2):145-51. PMID: 24029780. 24. Mahaffey KW, Hafley G, Dickerson S, et al. Results of a reevaluation of cardiovascular outcomes in the RECORD trial. Am Heart J. 2013 Aug;166(2):240-9 e1. PMID: 23895806. 25. Mamtani R, Haynes K, Bilker WB, et al. Association between longer therapy with thiazolidinediones and risk of bladder cancer: a cohort study. J Natl Cancer Inst. 2012 Sep 19;104(18):1411-21. PMID: 22878886. 26. Anon. (7) Approaches to glycemic treatment. Diabetes Care. 2015 Jan;38 Suppl:S41-8. PMID: 25537707. 27. Turner LW, Nartey D, Stafford RS, et al. Ambulatory treatment of type 2 diabetes in the U.S., 1997-2012. Diabetes Care. 2014 Apr;37(4):985-92. PMID: 24198301. 28. Raebel MA, Xu S, Goodrich GK, et al. Initial antihyperglycemic drug therapy among 241 327 adults with newly identified diabetes from 2005 through 2010: a surveillance, prevention, and management of diabetes mellitus (SUPREME-DM) study. Ann Pharmacother. 2013 Oct;47(10):1280- 91. PMID: 24259692. 29. Stone NJ, Robinson JG, Lichtenstein AH, et al. 2013 ACC/AHA Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014 Jun 24;129(25 Suppl 2):S1-S45. PMID: 24222016. 30. Anon. (8) Cardiovascular disease and risk management. Diabetes Care. 2015 Jan;38 Suppl:S49-57. PMID: 25537708. 31. Tsertsvadze A, Maglione M, Chou R, et al. Updating comparative effectiveness reviews: current efforts in AHRQ's Effective Health Care Program. J Clin Epidemiol. 2011 Nov;64(11):1208-15. PMID: 21684114. 32. Higgins JPT, S. G. Cochrane handbook for systemic reviews of interventions Version 5.1.0. 2011; http://handbook.cochrane.org/. Accessed Oxford, England.
- 385. 328 33. Jadad AR,Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996 Feb;17(1):1-12. PMID: 8721797. 34. Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health. 1998 Jun;52(6):377-84. PMID: 9764259. 35. Institute of Medicine. Finding What Works in Health Care: Standards for Systematic Reviews. Washington, DC: The National Academies Press; 2011; books.nap.edu/openbook.php?record_id=13 059&page=81. Accessed 2014 July 9. 36. Gaede P, Lund-Andersen H, Parving HH, Pedersen O. Effect of a multifactorial intervention on mortality in type 2 diabetes. The New England journal of medicine. 2008 Feb 7;358(6):580-91. PMID: 18256393. 37. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003 Sep 6;327(7414):557-60. PMID: 12958120. 38. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986 Sep;7(3):177-88. PMID: 3802833. 39. Cornell JE, Mulrow CD, Localio R, et al. Random-Effects Meta-analysis of Inconsistent Effects: A Time for Change. Ann Intern Med. 2014 Feb 18;160(4):267- 70. PMID: 24727843. 40. Balshem H, Stevens A, Ansari M, et al. Finding Grey Literature Evidence and Assessing for Outcome and Analysis Reporting Biases When Comparing Medical Interventions: AHRQ and the Effective Health Care Program Methods Guide for Effectiveness and Comparative Effectiveness Reviews. Rockville MD2008. 41. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994 Dec;50(4):1088-101. PMID: 7786990. 42. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997 Sep 13;315(7109):629-34. PMID: 9310563. 43. Owens DK, Lohr KN, Atkins D, et al. AHRQ series paper 5: grading the strength of a body of evidence when comparing medical interventions--agency for healthcare research and quality and the effective health- care program. J Clin Epidemiol. 2010 May;63(5):513-23. PMID: 19595577. 44. Singh N. Diabetes, heart rate, and mortality. J Cardiovasc Pharmacol Ther. 2002 Apr;7(2):117-29. PMID: 12075400. 45. Lenters-Westra E, Schindhelm RK, Bilo HJ, et al. Differences in interpretation of haemoglobin A1c values among diabetes care professionals. Neth J Med. 2014 Nov;72(9):462-6. PMID: 25431391. 46. Dupont WD, Plummer WD, Jr. Power and sample size calculations. A review and computer program. Control Clin Trials. 1990 Apr;11(2):116-28. PMID: 2161310. 47. Tosi F, Muggeo M, Brun E, et al. Combination treatment with metformin and glibenclamide versus single-drug therapies in type 2 diabetes mellitus: a randomized, double-blind, comparative study. Metabolism. 2003 2003 Jul;52(7):862-7. PMID: 12870162 48. Alba M, Ahren B, Inzucchi SE, et al. Sitagliptin and pioglitazone provide complementary effects on postprandial glucose and pancreatic islet cell function. Diabetes Obes Metab. 2013 Dec;15(12):1101-10. PMID: 23782502. 49. Home PD, Pocock SJ, Beck-Nielsen H, et al. Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicentre, randomised, open-label trial. Lancet. 2009 Jun 5;373(9681):2125-35. PMID: 19501900. 50. Kahn SE, Haffner SM, Heise MA, et al. Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy. The New England journal of medicine. 2006 Dec 7;355(23):2427-43. PMID: 17145742. 51. Bergenstal RM, Forti A, Chiasson JL, et al. Efficacy and safety of taspoglutide versus sitagliptin for type 2 diabetes mellitus (T- emerge 4 trial). Diabetes Ther. 2012 Dec;3(1):13. PMID: 23138449.
- 386. 329 52. Tolman KG,Freston JW, Kupfer S, Perez A. Liver safety in patients with type 2 diabetes treated with pioglitazone: results from a 3- year, randomized, comparator-controlled study in the US. Drug Saf. 2009;32(9):787- 800. PMID: 19670918. 53. Gallwitz B, Guzman J, Dotta F, et al. Exenatide twice daily versus glimepiride for prevention of glycaemic deterioration in patients with type 2 diabetes with metformin failure (EUREXA): an open-label, randomised controlled trial. Lancet. 2012 Jun 16;379(9833):2270-8. PMID: 22683137. 54. Del Prato S, Nauck M, Duran-Garcia S, et al. Long-term glycaemic response and tolerability of dapagliflozin versus a sulphonylurea as add-on therapy to metformin in type 2 diabetes patients: 4-year data. Diabetes Obes Metab. 2015 Mar 4. PMID: 25735400. 55. Derosa G, Maffioli P, Salvadeo SA, et al. Direct comparison among oral hypoglycemic agents and their association with insulin resistance evaluated by euglycemic hyperinsulinemic clamp: the 60's study. Metabolism. 2009 2009 Apr 22. PMID: 19394976 56. Gupta AK, Smith SR, Greenway FL, Bray GA. Pioglitazone treatment in type 2 diabetes mellitus when combined with portion control diet modifies the metabolic syndrome. Diabetes Obes Metab. 2009 2009 Apr;11(4):330-7. PMID: 19267711 57. Erdem G, Dogru T, Tasci I, et al. The effects of pioglitazone and metformin on plasma visfatin levels in patients with treatment naive type 2 diabetes mellitus. Diabetes Res Clin Pract. 2008 2008 Nov;82(2):214-8. PMID: 18778865 58. Iliadis F, Kadoglou NP, Hatzitolios A, et al. Metabolic effects of rosiglitazone and metformin in Greek patients with recently diagnosed type 2 diabetes. In Vivo. 2007 2007 Nov-Dec;21(6):1107-14. PMID: 18210765 59. Rosenstock J, Rood J, Cobitz A, et al. Initial treatment with rosiglitazone/metformin fixed-dose combination therapy compared with monotherapy with either rosiglitazone or metformin in patients with uncontrolled type 2 diabetes. Diabetes Obes Metab. 2006 2006 Nov;8(6):650-60. PMID: 17026489 60. Yamanouchi T, Sakai T, Igarashi K, et al. Comparison of metabolic effects of pioglitazone, metformin, and glimepiride over 1 year in Japanese patients with newly diagnosed Type 2 diabetes. Diabet Med. 2005 2005 Aug;22(8):980-5. PMID: 16026361 61. Ramachandran A, Snehalatha C, Salini J, Vijay V. Use of glimepiride and insulin sensitizers in the treatment of type 2 diabetes--a study in Indians. J Assoc Physicians India. 2004 2004 Jun;52:459-63. PMID: 15645955 62. Schernthaner G, Matthews DR, Charbonnel B, et al. Efficacy and safety of pioglitazone versus metformin in patients with type 2 diabetes mellitus: a double-blind, randomized trial. J Clin Endocrinol Metab. 2004 2004 Dec;89(12):6068-76. PMID: 15579760 63. Lawrence JM, Reid J, Taylor GJ, et al. Favorable effects of pioglitazone and metformin compared with gliclazide on lipoprotein subfractions in overweight patients with early type 2 diabetes. Diabetes Care. 2004 2004 Jan;27(1):41-6. PMID: 14693964 64. Pavo I, Jermendy G, Varkonyi TT, et al. Effect of pioglitazone compared with metformin on glycemic control and indicators of insulin sensitivity in recently diagnosed patients with type 2 diabetes. J Clin Endocrinol Metab. 2003 2003 Apr;88(4):1637-45. PMID: 12679450 65. Hallsten K, Virtanen KA, Lonnqvist F, et al. Rosiglitazone but not metformin enhances insulin- and exercise-stimulated skeletal muscle glucose uptake in patients with newly diagnosed type 2 diabetes. Diabetes. 2002 2002 Dec;51(12):3479-85. PMID: 12453903 66. Kiyici S, Ersoy C, Kaderli A, et al. Effect of rosiglitazone, metformin and medical nutrition treatment on arterial stiffness, serum MMP-9 and MCP-1 levels in drug naive type 2 diabetic patients. Diabetes Research and Clinical Practice. [doi: DOI: 10.1016/j.diabres.2009.07.004]. 2009 2009/10//;86(1):44-50. PMID: 19674806
- 387. 330 67. Perez A,Zhao Z, Jacks R, Spanheimer R. Efficacy and safety of pioglitazone/metformin fixed-dose combination therapy compared with pioglitazone and metformin monotherapy in treating patients with T2DM. Curr Med Res Opin. 2009 Dec;25(12):2915-23. PMID: 19827910. 68. Kato T, Sawai Y, Kanayama H, et al. Comparative study of low-dose pioglitazone or metformin treatment in Japanese diabetic patients with metabolic syndrome. Exp Clin Endocrinol Diabetes. 2009 Nov;117(10):593-9. PMID: 19924605. 69. Esteghamati A, Ghasemiesfe M, Mousavizadeh M, et al. Pioglitazone and metformin are equally effective in reduction of chemerin in patients with type 2 diabetes. J Diabetes Investig. 2014 May 4;5(3):327- 32. PMID: 24843782. 70. Erem C, Ozbas HM, Nuhoglu I, et al. Comparison of effects of gliclazide, metformin and pioglitazone monotherapies on glycemic control and cardiovascular risk factors in patients with newly diagnosed uncontrolled type 2 diabetes mellitus. Exp Clin Endocrinol Diabetes. 2014 May;122(5):295-302. PMID: 24710641. 71. Genovese S, De Berardis G, Nicolucci A, et al. Effect of pioglitazone versus metformin on cardiovascular risk markers in type 2 diabetes. Adv Ther. 2013 Feb;30(2):190- 202. PMID: 23359066. 72. Taslimi S, Esteghamati A, Rashidi A, et al. Treatment with pioglitazone is associated with decreased preprandial ghrelin levels: a randomized clinical trial. Peptides. 2013 Feb;40:89-92. PMID: 23276779. 73. Russell-Jones D, Cuddihy RM, Hanefeld M, et al. Efficacy and safety of exenatide once weekly versus metformin, pioglitazone, and sitagliptin used as monotherapy in drug- naive patients with type 2 diabetes (DURATION-4): a 26-week double-blind study. Diabetes Care. 2012 Feb;35(2):252-8. PMID: 22210563. 74. Yoon KH, Shin JA, Kwon HS, et al. Comparison of the efficacy of glimepiride, metformin, and rosiglitazone monotherapy in korean drug-naive type 2 diabetic patients: the practical evidence of antidiabetic monotherapy study. Diabetes Metab J. 2011 Feb;35(1):26-33. PMID: 21537410. 75. Fidan E, Onder Ersoz H, Yilmaz M, et al. The effects of rosiglitazone and metformin on inflammation and endothelial dysfunction in patients with type 2 diabetes mellitus. Acta Diabetol. 2011 Dec;48(4):297-302. PMID: 21424914. 76. Esposito K, Maiorino MI, Di Palo C, et al. Effects of pioglitazone versus metformin on circulating endothelial microparticles and progenitor cells in patients with newly diagnosed type 2 diabetes--a randomized controlled trial. Diabetes Obes Metab. 2011 May;13(5):439-45. PMID: 21255215. 77. Esteghamati A, Azizi R, Ebadi M, et al. The Comparative Effect of Pioglitazone and Metformin on Serum Osteoprotegerin, Adiponectin and Intercellular Adhesion Molecule Concentrations in Patients with Newly Diagnosed Type 2 Diabetes: a Randomized Clinical Trial. Exp Clin Endocrinol Diabetes. 2015 Jan 21. PMID: 25607338. 78. Turkmen Kemal Y, Guvener Demirag N, Yildirir A, et al. Effects of rosiglitazone on plasma brain natriuretic peptide levels and myocardial performance index in patients with type 2 diabetes mellitus. Acta Diabetol. 2007 2007 Sep;44(3):149-56. PMID: 17721754 79. Gupta RK, Rehan HS, Rohatagi A, et al. The effect of glipizide, metformin and rosiglitazone on nontraditional cardiovascular risk factors in newly diagnosed patients with type 2 diabetes mellitus. International journal of diabetes in developing countries [serial on the Internet]. 2010; (3): Available from: http://onlinelibrary.wiley.com/o/cochrane/clcentr al/articles/971/CN-00797971/frame.html.
- 388. 331 80. Goldstein BJ,Feinglos MN, Lunceford JK, et al. Effect of initial combination therapy with sitagliptin, a dipeptidyl peptidase-4 inhibitor, and metformin on glycemic control in patients with type 2 diabetes. Diabetes Care. 2007 2007 Aug;30(8):1979- 87. PMID: 17485570 81. Williams-Herman D, Johnson J, Teng R, et al. Efficacy and safety of initial combination therapy with sitagliptin and metformin in patients with type 2 diabetes: a 54-week study. Curr Med Res Opin. 2009 2009 Mar;25(3):569-83. PMID: 19232032 82. Aschner P, Katzeff HL, Guo H, et al. Efficacy and safety of monotherapy of sitagliptin compared with metformin in patients with type 2 diabetes. Diabetes Obes Metab. 2010 Mar;12(3):252-61. PMID: 20070351. 83. Jadzinsky M, Pfutzner A, Paz-Pacheco E, et al. Saxagliptin given in combination with metformin as initial therapy improves glycaemic control in patients with type 2 diabetes compared with either monotherapy: a randomized controlled trial. Diabetes Obes Metab. 2009 2009 Jun;11(6):611-22. PMID: 19515181 84. Pratley RE, Fleck P, Wilson C. Efficacy and safety of initial combination therapy with alogliptin plus metformin versus either as monotherapy in drug-naive patients with type 2 diabetes: a randomized, double-blind, 6-month study. Diabetes Obes Metab. 2014 Jul;16(7):613-21. PMID: 24400655. 85. Williams-Herman D, Johnson J, Teng R, et al. Efficacy and safety of sitagliptin and metformin as initial combination therapy and as monotherapy over 2 years in patients with type 2 diabetes. Diabetes Obes Metab. 2010 May;12(5):442-51. PMID: 20415693. 86. Haak T, Meinicke T, Jones R, et al. Initial combination of linagliptin and metformin improves glycaemic control in type 2 diabetes: a randomized, double-blind, placebo-controlled study. Diabetes Obes Metab. 2012 Jun;14(6):565-74. PMID: 22356132. 87. Pfutzner A, Paz-Pacheco E, Allen E, et al. Initial combination therapy with saxagliptin and metformin provides sustained glycaemic control and is well tolerated for up to 76 weeks. Diabetes Obes Metab. 2011 Jun;13(6):567-76. PMID: 21342412. 88. Henry RR, Murray AV, Marmolejo MH, et al. Dapagliflozin, metformin XR, or both: initial pharmacotherapy for type 2 diabetes, a randomised controlled trial. Int J Clin Pract. 2012 May;66(5):446-56. PMID: 22413962. 89. List JF, Woo V, Morales E, et al. Sodium- glucose cotransport inhibition with dapagliflozin in type 2 diabetes. Diabetes Care. 2009 Apr;32(4):650-7. PMID: 19114612. 90. Ferrannini E, Berk A, Hantel S, et al. Long- term safety and efficacy of empagliflozin, sitagliptin, and metformin: An active- controlled, parallel-group, randomized, 78- week open-label extension study in patients with type 2 diabetes. Diabetes Care. 2013;36(12):4015-21. PMID: 24186878 91. Umpierrez G, Povedano ST, Manghi FP, et al. Efficacy and Safety of Dulaglutide Monotherapy Versus Metformin in Type 2 Diabetes in a Randomized Controlled Trial (AWARD-3). Diabetes Care. 2014 May 19. PMID: 24842985. 92. Yuan GH, Song WL, Huang YY, et al. Efficacy and tolerability of exenatide monotherapy in obese patients with newly diagnosed type 2 diabetes: a randomized, 26 weeks metformin-controlled, parallel-group study. Chin Med J (Engl). 2012 Aug;125(15):2677-81. PMID: 22931974. 93. Teramoto T, Yamada N, Shirai K, Saito Y. Effects of pioglitazone hydrochloride on Japanese patients with type 2 diabetes mellitus. J Atheroscler Thromb. 2007 2007 Apr;14(2):86-93. PMID: 17485893 94. Hanefeld M, Patwardhan R, Jones NP. A one-year study comparing the efficacy and safety of rosiglitazone and glibenclamide in the treatment of type 2 diabetes. Nutr Metab Cardiovasc Dis. 2007 2007 Jan;17(1):13-23. PMID: 17174222
- 389. 332 95. Jain R,Osei K, Kupfer S, et al. Long-term safety of pioglitazone versus glyburide in patients with recently diagnosed type 2 diabetes mellitus. Pharmacotherapy. 2006 2006 Oct;26(10):1388-95. PMID: 16999648 96. Nakamura T, Matsuda T, Kawagoe Y, et al. Effect of pioglitazone on carotid intima- media thickness and arterial stiffness in type 2 diabetic nephropathy patients. Metabolism. 2004 2004 Oct;53(10):1382-6. PMID: 15375799 97. Nakamura T, Ushiyama C, Shimada N, et al. Comparative effects of pioglitazone, glibenclamide, and voglibose on urinary endothelin-1 and albumin excretion in diabetes patients. J Diabetes Complications. 2000 2000 Sep-Oct;14(5):250-4. PMID: 11113686 98. Bakris G, Viberti G, Weston WM, et al. Rosiglitazone reduces urinary albumin excretion in type II diabetes. J Hum Hypertens. 2003 2003 Jan;17(1):7-12. PMID: 12571611 99. Pfutzner A, Marx N, Lubben G, et al. Improvement of cardiovascular risk markers by pioglitazone is independent from glycemic control: results from the pioneer study. J Am Coll Cardiol. 2005 2005 Jun 21;45(12):1925-31. PMID: 15963388 100. Tan MH, Johns D, Strand J, et al. Sustained effects of pioglitazone vs. glibenclamide on insulin sensitivity, glycaemic control, and lipid profiles in patients with Type 2 diabetes. Diabet Med. 2004 2004 Aug;21(8):859-66. PMID: 15270789 101. Tan M, Johns D, Gonzalez Galvez G, et al. Effects of pioglitazone and glimepiride on glycemic control and insulin sensitivity in Mexican patients with type 2 diabetes mellitus: A multicenter, randomized, double-blind, parallel-group trial. Clin Ther. 2004 2004 May;26(5):680-93. PMID: 15220012 102. Nakamura T, Sugaya T, Kawagoe Y, et al. Effect of pioglitazone on urinary liver-type fatty acid-binding protein concentrations in diabetes patients with microalbuminuria. Diabetes Metab Res Rev. 2006 2006 Sep- Oct;22(5):385-9. PMID: 16506273 103. Shihara N, Kitaoka M, Inagaki N, et al. Randomized controlled trial of single-agent glimepiride and pioglitazone in Japanese patients with type 2 diabetes: A comparative study. J Diabetes Investig. 2011 Oct 7;2(5):391-8. PMID: 24843519. 104. Rosenstock J, Inzucchi SE, Seufert J, et al. Initial combination therapy with alogliptin and pioglitazone in drug-naive patients with type 2 diabetes. Diabetes Care. 2010 Nov;33(11):2406-8. PMID: 20724648. 105. Xu W, Bi Y, Sun Z, et al. Comparison of the effects on glycaemic control and beta-cell function in newly diagnosed type 2 diabetes patients of treatment with exenatide, insulin or pioglitazone: a multicentre randomized parallel-group trial (the CONFIDENCE study). J Intern Med. 2015 Jan;277(1):137- 50. PMID: 25039675. 106. Barnett AH, Patel S, Harper R, et al. Linagliptin monotherapy in type 2 diabetes patients for whom metformin is inappropriate: an 18-week randomized, double-blind, placebo-controlled phase III trial with a 34-week active-controlled extension. Diabetes Obes Metab. 2012 Dec;14(12):1145-54. PMID: 22974280. 107. Arjona Ferreira JC, Marre M, Barzilai N, et al. Efficacy and safety of sitagliptin versus glipizide in patients with type 2 diabetes and moderate-to-severe chronic renal insufficiency. Diabetes Care. 2013 May;36(5):1067-73. PMID: 23248197. 108. Scott R, Wu M, Sanchez M, Stein P. Efficacy and tolerability of the dipeptidyl peptidase-4 inhibitor sitagliptin as monotherapy over 12 weeks in patients with type 2 diabetes. Int J Clin Pract. 2007 2007 Jan;61(1):171-80. PMID: 17156104 109. Seino Y, Rasmussen MF, Nishida T, Kaku K. Efficacy and safety of the once-daily human GLP-1 analogue, liraglutide, vs glibenclamide monotherapy in Japanese patients with type 2 diabetes. Curr Med Res Opin. 2010 May;26(5):1013-22. PMID: 20199137.
- 390. 333 110. Kaku K,Rasmussen MF, Nishida T, Seino Y. Fifty-two-week, randomized, multicenter trial to compare the safety and efficacy of the novel glucagon-like peptide-1 analog liraglutide vs glibenclamide in patients with type 2 diabetes. J Diabetes Investig. 2011 Nov 30;2(6):441-7. PMID: 24843528. 111. Madsbad S, Schmitz O, Ranstam J, et al. Improved glycemic control with no weight increase in patients with type 2 diabetes after once-daily treatment with the long- acting glucagon-like peptide 1 analog liraglutide (NN2211): a 12-week, double- blind, randomized, controlled trial. Diabetes Care. 2004 2004 Jun;27(6):1335-42. PMID: 15161785 112. Garber A, Henry R, Ratner R, et al. Liraglutide versus glimepiride monotherapy for type 2 diabetes (LEAD-3 Mono): a randomised, 52-week, phase III, double- blind, parallel-treatment trial. Lancet. 2009 2009 Feb 7;373(9662):473-81. PMID: 18819705 113. Garber A, Henry RR, Ratner R, et al. Liraglutide, a once-daily human glucagon- like peptide 1 analogue, provides sustained improvements in glycaemic control and weight for 2 years as monotherapy compared with glimepiride in patients with type 2 diabetes. Diabetes Obes Metab. 2011 Apr;13(4):348-56. PMID: 21205128. 114. Roden M, Weng J, Eilbracht J, et al. Empagliflozin monotherapy with sitagliptin as an active comparator in patients with type 2 diabetes: A randomised, double-blind, placebo-controlled, phase 3 trial. The Lancet Diabetes and Endocrinology [serial on the Internet]. 2013; (3): Available from: http://onlinelibrary.wiley.com/o/cochrane/clcentr al/articles/599/CN-00915599/frame.html. 115. Suzuki K, Tanaka S, Aoki C, et al. Greater efficacy and improved endothelial dysfunction in untreated type 2 diabetes with liraglutide versus sitagliptin. Dokkyo Journal of Medical Sciences. 2014;41(3):211-20. PMID: COULD NOT FIND THIS ONE 116. Leiter LA, Harris SB, Chiasson J-L, et al. Efficacy and safety of Rosiglitazone as monotherapy or in combination with metformin in primary care settings. Can J Diabetes. 2005 2005;29(4):384-92. PMID: COULD NOT FIND THIS ONE 117. Kaku K. Efficacy and safety of therapy with metformin plus pioglitazone in the treatment of patients with type 2 diabetes: a double- blind, placebo-controlled, clinical trial. Curr Med Res Opin. 2009 2009 May;25(5):1111- 9. PMID: 19309251 118. Scott R, Loeys T, Davies MJ, Engel SS. Efficacy and safety of sitagliptin when added to ongoing metformin therapy in patients with type 2 diabetes. Diabetes Obes Metab. 2008 2008 Sep;10(10):959-69. PMID: 18201203 119. Weissman P, Goldstein BJ, Rosenstock J, et al. Effects of rosiglitazone added to submaximal doses of metformin compared with dose escalation of metformin in type 2 diabetes: the EMPIRE Study. Curr Med Res Opin. 2005 2005 Dec;21(12):2029-35. PMID: 16368054 120. Bailey CJ, Bagdonas A, Rubes J, et al. Rosiglitazone/metformin fixed-dose combination compared with uptitrated metformin alone in type 2 diabetes mellitus: a 24-week, multicenter, randomized, double- blind, parallel-group study. Clin Ther. 2005 2005 Oct;27(10):1548-61. PMID: 16330291 121. Gomez-Perez FJ, Fanghanel-Salmon G, Antonio Barbosa J, et al. Efficacy and safety of rosiglitazone plus metformin in Mexicans with type 2 diabetes. Diabetes Metab Res Rev. 2002 2002 Mar-Apr;18(2):127-34. PMID: 11994904 122. Einhorn D, Rendell M, Rosenzweig J, et al. Pioglitazone hydrochloride in combination with metformin in the treatment of type 2 diabetes mellitus: a randomized, placebo- controlled study. The Pioglitazone 027 Study Group. Clin Ther. 2000 2000 Dec;22(12):1395-409. PMID: 11192132 123. Fonseca V, Rosenstock J, Patwardhan R, Salzman A. Effect of metformin and rosiglitazone combination therapy in patients with type 2 diabetes mellitus: a randomized controlled trial. JAMA. 2000 2000 Apr 5;283(13):1695-702. PMID: 10755495
- 391. 334 124. Kadoglou NP,Kapelouzou A, Tsanikidis H, et al. Effects of rosiglitazone/metformin fixed-dose combination therapy and metformin monotherapy on serum vaspin, adiponectin and IL-6 levels in drug-naive patients with type 2 diabetes. Exp Clin Endocrinol Diabetes. 2011 Feb;119(2):63-8. PMID: 21031343. 125. Genovese S, Passaro A, Brunetti P, et al. Pioglitazone Randomised Italian Study on Metabolic Syndrome (PRISMA): effect of pioglitazone with metformin on HDL-C levels in Type 2 diabetic patients. J Endocrinol Invest. 2013 Sep;36(8):606-16. PMID: 23511244. 126. DeFronzo RA, Burant CF, Fleck P, et al. Efficacy and tolerability of the DPP-4 inhibitor alogliptin combined with pioglitazone, in metformin-treated patients with type 2 diabetes. J Clin Endocrinol Metab. 2012 May;97(5):1615-22. PMID: 22419732. 127. Borges JL, Bilezikian JP, Jones-Leone AR, et al. A randomized, parallel group, double- blind, multicentre study comparing the efficacy and safety of Avandamet (rosiglitazone/metformin) and metformin on long-term glycaemic control and bone mineral density after 80 weeks of treatment in drug-naive type 2 diabetes mellitus patients. Diabetes Obes Metab. 2011 Nov;13(11):1036-46. PMID: 21682834. 128. Feinglos M, Dailey G, Cefalu W, et al. Effect on glycemic control of the addition of 2.5 mg glipizide GITS to metformin in patients with T2DM. Diabetes Res Clin Pract. 2005 2005 May;68(2):167-75. PMID: 15860246 129. Garber AJ, Donovan DSJ, Dandona P, et al. Efficacy of glyburide/metformin tablets compared with initial monotherapy in type 2 diabetes. J Clin Endocrinol Metab. 2003 2003 Aug;88(8):3598-604. PMID: 12915642 130. Goldstein BJ, Pans M, Rubin CJ. Multicenter, randomized, double-masked, parallel-group assessment of simultaneous glipizide/metformin as second-line pharmacologic treatment for patients with type 2 diabetes mellitus that is inadequately controlled by a sulfonylurea. Clin Ther. 2003 2003 Mar;25(3):890-903. PMID: 12852706 131. Blonde L, Rosenstock J, Mooradian AD, et al. Glyburide/metformin combination product is safe and efficacious in patients with type 2 diabetes failing sulphonylurea therapy. Diabetes Obes Metab. 2002 2002 Nov;4(6):368-75. PMID: 12406033 132. Marre M, Howlett H, Lehert P, Allavoine T. Improved glycaemic control with metformin-glibenclamide combined tablet therapy (Glucovance) in Type 2 diabetic patients inadequately controlled on metformin. Diabet Med. 2002 2002 Aug;19(8):673-80. PMID: 12147149 133. Garber AJ, Larsen J, Schneider SH, et al. Simultaneous glyburide/metformin therapy is superior to component monotherapy as an initial pharmacological treatment for type 2 diabetes. Diabetes Obes Metab. 2002 2002 May;4(3):201-8. PMID: 12047399 134. Hermann LS, Schersten B, Bitzen PO, et al. Therapeutic comparison of metformin and sulfonylurea, alone and in various combinations. A double-blind controlled study. Diabetes Care. 1994 1994 Oct;17(10):1100-9. PMID: 7821128 135. Hermann LS, Bitzen PO, Kjellstrom T, et al. Comparative efficacy of metformin and glibenclamide in patients with non-insulin- dependent diabetes mellitus. Diabete Metab. 1991 1991 May;17(1 Pt 2):201-8. PMID: 1936477 136. Charpentier G, Fleury F, Kabir M, et al. Improved glycaemic control by addition of glimepiride to metformin monotherapy in type 2 diabetic patients. Diabet Med. 2001 2001 Oct;18(10):828-34. PMID: 11678974 137. DeFronzo RA, Goodman AM. Efficacy of metformin in patients with non-insulin- dependent diabetes mellitus. The Multicenter Metformin Study Group. The New England journal of medicine. 1995 1995 Aug 31;333(9):541-9. PMID: 7623902 138. Chien HH, Chang CT, Chu NF, et al. Effect of glyburide-metformin combination tablet in patients with type 2 diabetes. J Chin Med Assoc. 2007 2007 Nov;70(11):473-80. PMID: 18063500
- 392. 335 139. Forst T,Uhlig-Laske B, Ring A, et al. Linagliptin (BI 1356), a potent and selective DPP-4 inhibitor, is safe and efficacious in combination with metformin in patients with inadequately controlled Type 2 diabetes. Diabet Med. 2010 Dec;27(12):1409-19. PMID: 21059094. 140. Kim HS, Kim DM, Cha BS, et al. Efficacy of glimepiride/metformin fixed-dose combination vs metformin uptitration in type 2 diabetic patients inadequately controlled on low-dose metformin monotherapy: A randomized, open label, parallel group, multicenter study in Korea. Journal of Diabetes Investigation. 2014((Kim H.-S.) Department of Internal Medicine Keimyung University School of Medicine Daegu Korea). PMID: 2542271 141. Ahren B, Johnson SL, Stewart M, et al. HARMONY 3: 104-Week Randomized, Double-Blind, Placebo- and Active- Controlled Trial Assessing the Efficacy and Safety of Albiglutide Compared With Placebo, Sitagliptin, and Glimepiride in Patients With Type 2 Diabetes Taking Metformin. Diabetes Care. 2014 Jun 4. PMID: 24898304. 142. Raz I, Chen Y, Wu M, et al. Efficacy and safety of sitagliptin added to ongoing metformin therapy in patients with type 2 diabetes. Curr Med Res Opin. 2008 2008 Feb;24(2):537-50. PMID: 18194595 143. Charbonnel B, Karasik A, Liu J, et al. Efficacy and safety of the dipeptidyl peptidase-4 inhibitor sitagliptin added to ongoing metformin therapy in patients with type 2 diabetes inadequately controlled with metformin alone. Diabetes Care. 2006 2006 Dec;29(12):2638-43. PMID: 17130197 144. DeFronzo RA, Hissa MN, Garber AJ, et al. The efficacy and safety of saxagliptin when added to metformin therapy in patients with inadequately controlled type 2 diabetes with metformin alone. Diabetes Care. 2009 Sep;32(9):1649-55. PMID: 19478198. 145. Reasner C, Olansky L, Seck TL, et al. The effect of initial therapy with the fixed-dose combination of sitagliptin and metformin compared with metformin monotherapy in patients with type 2 diabetes mellitus. Diabetes Obes Metab. 2011 Jul;13(7):644- 52. PMID: 21410627. 146. Yang W, Pan CY, Tou C, et al. Efficacy and safety of saxagliptin added to metformin in Asian people with type 2 diabetes mellitus: a randomized controlled trial. Diabetes Res Clin Pract. 2011 Nov;94(2):217-24. PMID: 21871686. 147. Fonseca V, Zhu T, Karyekar C, Hirshberg B. Adding saxagliptin to extended-release metformin vs. uptitrating metformin dosage. Diabetes Obes Metab. 2012 Apr;14(4):365- 71. PMID: 22192246. 148. Yang W, Guan Y, Shentu Y, et al. The addition of sitagliptin to ongoing metformin therapy significantly improves glycemic control in Chinese patients with type 2 diabetes. J Diabetes. 2012 Sep;4(3):227-37. PMID: 22672586. 149. Kadowaki T, Tajima N, Odawara M, et al. Addition of sitagliptin to ongoing metformin monotherapy improves glycemic control in Japanese patients with type 2 diabetes over 52 weeks. J Diabetes Investig. 2013 Mar 18;4(2):174-81. PMID: 24843649. 150. Derosa G, Carbone A, D'Angelo A, et al. Variations in inflammatory biomarkers following the addition of sitagliptin in patients with type 2 diabetes not controlled with metformin. Intern Med. 2013;52(19):2179-87. PMID: 24088749. 151. White JL, Buchanan P, Li J, Frederich R. A randomized controlled trial of the efficacy and safety of twice-daily saxagliptin plus metformin combination therapy in patients with type 2 diabetes and inadequate glycemic control on metformin monotherapy. BMC Endocr Disord. 2014;14(1):17. PMID: 24565221. 152. Ross SA, Rafeiro E, Meinicke T, et al. Efficacy and safety of linagliptin 2.5?mg twice daily versus 5?mg once daily in patients with type 2 diabetes inadequately controlled on metformin: a randomised, double-blind, placebo-controlled trial. Curr Med Res Opin [serial on the Internet]. 2012; (9): Available from: http://onlinelibrary.wiley.com/o/cochrane/clcentr al/articles/674/CN-00859674/frame.html.
- 393. 336 153. Rosenstock J,Seman LJ, Jelaska A, et al. Efficacy and safety of empagliflozin, a sodium glucose cotransporter 2 (SGLT2) inhibitor, as add-on to metformin in type 2 diabetes with mild hyperglycaemia. Diabetes Obes Metab [serial on the Internet]. 2013; (12): Available from: http://onlinelibrary.wiley.com/o/cochrane/clcentr al/articles/648/CN-00915648/frame.html. 154. Nauck MA, Ellis GC, Fleck PR, et al. Efficacy and safety of adding the dipeptidyl peptidase-4 inhibitor alogliptin to metformin therapy in patients with type 2 diabetes inadequately controlled with metformin monotherapy: a multicentre, randomised, double-blind, placebo-controlled study. Int J Clin Pract. 2009 Jan;63(1):46-55. PMID: 19125992. 155. Taskinen MR, Rosenstock J, Tamminen I, et al. Safety and efficacy of linagliptin as add- on therapy to metformin in patients with type 2 diabetes: a randomized, double-blind, placebo-controlled study. Diabetes Obes Metab. 2011 Jan;13(1):65-74. PMID: 21114605. 156. Rosenstock J, Aggarwal N, Polidori D, et al. Dose-ranging effects of canagliflozin, a sodium-glucose cotransporter 2 inhibitor, as add-on to metformin in subjects with type 2 diabetes. Diabetes Care. 2012 Jun;35(6):1232-8. PMID: 22492586. 157. Seino Y, Miyata Y, Hiroi S, et al. Efficacy and safety of alogliptin added to metformin in Japanese patients with type 2 diabetes: a randomized, double-blind, placebo- controlled trial with an open-label, long- term extension study. Diabetes Obes Metab. 2012 Oct;14(10):927-36. PMID: 22583697. 158. Lavalle-Gonzalez FJ, Januszewicz A, Davidson J, et al. Efficacy and safety of canagliflozin compared with placebo and sitagliptin in patients with type 2 diabetes on background metformin monotherapy: a randomised trial. Diabetologia. 2013 Dec;56(12):2582-92. PMID: 24026211. 159. Nauck M, Weinstock RS, Umpierrez GE, et al. Efficacy and Safety of Dulaglutide Versus Sitagliptin After 52 Weeks in Type 2 Diabetes in a Randomized Controlled Trial (AWARD-5). Diabetes Care. 2014 Apr 17. PMID: 24742660. 160. Wang W, Yang J, Yang G, et al. Efficacy and safety of linagliptin in Asian patients with type 2 diabetes mellitus inadequately controlled by metformin: A multinational 24-week, randomized clinical trial. J Diabetes. 2015 Mar 6. PMID: 25753488. 161. Hermans MP, Delibasi T, Farmer I, et al. Effects of saxagliptin added to sub-maximal doses of metformin compared with uptitration of metformin in type 2 diabetes: the PROMPT study. Curr Med Res Opin. 2012 Oct;28(10):1635-45. PMID: 23020253. 162. Ji L, Zinman B, Patel S, et al. Efficacy and safety of linagliptin co-administered with low-dose metformin once daily versus high- dose metformin twice daily in treatment- naive patients with type 2 diabetes: a double-blind randomized trial. Adv Ther. 2015 Mar;32(3):201-15. PMID: 25805187. 163. Aaboe K, Knop FK, Vilsboll T, et al. Twelve weeks treatment with the DPP-4 inhibitor, sitagliptin, prevents degradation of peptide YY and improves glucose and non- glucose induced insulin secretion in patients with type 2 diabetes mellitus. Diabetes Obes Metab. 2010 Apr;12(4):323-33. PMID: 20380653. 164. Haak T, Meinicke T, Jones R, et al. Initial combination of linagliptin and metformin in patients with type 2 diabetes: efficacy and safety in a randomised, double-blind 1-year extension study. Int J Clin Pract. 2013 Dec;67(12):1283-93. PMID: 24118640. 165. Qiu R, Capuano G, Meininger G. Efficacy and safety of twice-daily treatment with canagliflozin, a sodium glucose co- transporter 2 inhibitor, added on to metformin monotherapy in patients with type 2 diabetes mellitus. Journal of Clinical and Translational Endocrinology. 2014;1(2):54-60. PMID: COULD NOT FIND THIS ONE 166. Haring HU, Merker L, Seewaldt-Becker E, et al. Empaglif lozin as add-on to metformin in patients with type 2 diabetes: A 24-week, randomized, double-blind, placebo- controlled trial. Diabetes Care. 2014;37(6):1650-9. PMID: 24722494
- 394. 337 167. Bailey CJ,Gross JL, Pieters A, et al. Effect of dapagliflozin in patients with type 2 diabetes who have inadequate glycaemic control with metformin: a randomised, double-blind, placebo-controlled trial. Lancet. 2010 Jun 26;375(9733):2223-33. PMID: 20609968. 168. Schumm-Draeger PM, Burgess L, Koranyi L, et al. Twice-daily dapagliflozin co- administered with metformin in type 2 diabetes: a 16-week randomized, placebo- controlled clinical trial. Diabetes Obes Metab. 2015 Jan;17(1):42-51. PMID: 25200570. 169. Bolinder J, Ljunggren O, Kullberg J, et al. Effects of dapagliflozin on body weight, total fat mass, and regional adipose tissue distribution in patients with type 2 diabetes mellitus with inadequate glycemic control on metformin. J Clin Endocrinol Metab. 2012 Mar;97(3):1020-31. PMID: 22238392. 170. Bailey CJ, Gross JL, Hennicken D, et al. Dapagliflozin add-on to metformin in type 2 diabetes inadequately controlled with metformin: a randomized, double-blind, placebo-controlled 102-week trial. BMC Med. 2013;11:43. PMID: 23425012. 171. Derosa G, Cicero AF, Franzetti IG, et al. Effects of exenatide and metformin in combination on some adipocytokine levels: a comparison with metformin monotherapy. Can J Physiol Pharmacol. 2013 Sep;91(9):724-32. PMID: 23984793. 172. Forst T, Michelson G, Ratter F, et al. Addition of liraglutide in patients with Type 2 diabetes well controlled on metformin monotherapy improves several markers of vascular function. Diabet Med. 2012 Sep;29(9):1115-8. PMID: 22288732. 173. Apovian CM, Bergenstal RM, Cuddihy RM, et al. Effects of exenatide combined with lifestyle modification in patients with type 2 diabetes. Am J Med. 2010 May;123(5):468 e9-17. PMID: 20399326. 174. DeFronzo RA, Ratner RE, Han J, et al. Effects of exenatide (exendin-4) on glycemic control and weight over 30 weeks in metformin-treated patients with type 2 diabetes. Diabetes Care. 2005 May;28(5):1092-100. PMID: 15855572. 175. Hamann A, Garcia-Puig J, Paul G, et al. Comparison of fixed-dose rosiglitazone/metformin combination therapy with sulphonylurea plus metformin in overweight individuals with Type 2 diabetes inadequately controlled on metformin alone. Exp Clin Endocrinol Diabetes. 2008 2008 Jan;116(1):6-13. PMID: 18095238 176. Home PD, Jones NP, Pocock SJ, et al. Rosiglitazone RECORD study: glucose control outcomes at 18 months. Diabet Med. 2007 2007 Jun;24(6):626-34. PMID: 17517066 177. Bakris GL, Ruilope LM, McMorn SO, et al. Rosiglitazone reduces microalbuminuria and blood pressure independently of glycemia in type 2 diabetes patients with microalbuminuria. J Hypertens. 2006 2006 Oct;24(10):2047-55. PMID: 16957566 178. Umpierrez G, Issa M, Vlajnic A. Glimepiride versus pioglitazone combination therapy in subjects with type 2 diabetes inadequately controlled on metformin monotherapy: results of a randomized clinical trial. Curr Med Res Opin. 2006 2006 Apr;22(4):751-9. PMID: 16684436 179. Derosa G, Gaddi AV, Piccinni MN, et al. Antithrombotic effects of rosiglitazone- metformin versus glimepiride-metformin combination therapy in patients with type 2 diabetes mellitus and metabolic syndrome. Pharmacotherapy. 2005 2005 May;25(5):637-45. PMID: 15899724 180. Garber A, Klein E, Bruce S, et al. Metformin-glibenclamide versus metformin plus rosiglitazone in patients with type 2 diabetes inadequately controlled on metformin monotherapy. Diabetes Obes Metab. 2006 2006 Mar;8(2):156-63. PMID: 16448519 181. Maffioli P, Fogari E, D'Angelo A, et al. Ultrasonography modifications of visceral and subcutaneous adipose tissue after pioglitazone or glibenclamide therapy combined with rosuvastatin in type 2 diabetic patients not well controlled by metformin. Eur J Gastroenterol Hepatol. 2013 Sep;25(9):1113-22. PMID: 23524525.
- 395. 338 182. Petrica L,Petrica M, Vlad A, et al. Nephro- and neuroprotective effects of rosiglitazone versus glimepiride in normoalbuminuric patients with type 2 diabetes mellitus: a randomized controlled trial. Wien Klin Wochenschr. 2009;121(23-24):765-75. PMID: 20047115. 183. Comaschi M, Demicheli A, Di Pietro C, et al. Effects of pioglitazone in combination with metformin or a sulfonylurea compared to a fixed-dose combination of metformin and glibenclamide in patients with type 2 diabetes. Diabetes Technol Ther. 2007 2007 Aug;9(4):387-98. PMID: 17705695 184. Schondorf T, Musholt PB, Hohberg C, et al. The fixed combination of pioglitazone and metformin improves biomarkers of platelet function and chronic inflammation in type 2 diabetes patients: results from the PIOfix study. J Diabetes Sci Technol. 2011 Mar;5(2):426-32. PMID: 21527115. 185. Pfutzner A, Schondorf T, Tschope D, et al. PIOfix-study: effects of pioglitazone/metformin fixed combination in comparison with a combination of metformin with glimepiride on diabetic dyslipidemia. Diabetes Technol Ther. 2011 Jun;13(6):637-43. PMID: 21457065. 186. Rigby SP, Handelsman Y, Lai YL, et al. Effects of colesevelam, rosiglitazone, or sitagliptin on glycemic control and lipid profile in patients with type 2 diabetes mellitus inadequately controlled by metformin monotherapy. Endocr Pract. 2010 Jan-Feb;16(1):53-63. PMID: 19789153. 187. Chawla S, Kaushik N, Singh NP, et al. Effect of addition of either sitagliptin or pioglitazone in patients with uncontrolled type 2 diabetes mellitus on metformin: A randomized controlled trial. J Pharmacol Pharmacother. 2013 Jan;4(1):27-32. PMID: 23662021. 188. Bergenstal RM, Wysham C, Macconell L, et al. Efficacy and safety of exenatide once weekly versus sitagliptin or pioglitazone as an adjunct to metformin for treatment of type 2 diabetes (DURATION-2): a randomised trial. Lancet. 2010 Aug 7;376(9739):431-9. PMID: 20580422. 189. DeFronzo RA, Triplitt C, Qu Y, et al. Effects of exenatide plus rosiglitazone on beta-cell function and insulin sensitivity in subjects with type 2 diabetes on metformin. Diabetes Care. 2010 May;33(5):951-7. PMID: 20107105. 190. Arechavaleta R, Seck T, Chen Y, et al. Efficacy and safety of treatment with sitagliptin or glimepiride in patients with type 2 diabetes inadequately controlled on metformin monotherapy: a randomized, double-blind, non-inferiority trial. Diabetes Obes Metab. 2011 Feb;13(2):160-8. PMID: 21199268. 191. Forst T, Anastassiadis E, Diessel S, et al. Effect of linagliptin compared to glimepiride on postprandial glucose metabolism, islet cell function, and vascular function parameters in patients with type 2 diabetes mellitus on ongoing metformin treatment. Diabetes Metab Res Rev. 2014 Jan 23. PMID: 24459063. 192. Nauck MA, Meininger G, Sheng D, et al. Efficacy and safety of the dipeptidyl peptidase-4 inhibitor, sitagliptin, compared with the sulfonylurea, glipizide, in patients with type 2 diabetes inadequately controlled on metformin alone: a randomized, double- blind, non-inferiority trial. Diabetes Obes Metab. 2007 2007 Mar;9(2):194-205. PMID: 17300595 193. Schernthaner G, Duran-Garcia S, Hanefeld M, et al. Efficacy and tolerability of saxagliptin compared with glimepiride in elderly patients with type 2 diabetes: a randomized, controlled study (GENERATION). Diabetes Obes Metab. 2015 Jul;17(7):630-8. PMID: 25761977. 194. Gallwitz B, Rosenstock J, Rauch T, et al. 2- year efficacy and safety of linagliptin compared with glimepiride in patients with type 2 diabetes inadequately controlled on metformin: a randomised, double-blind, non-inferiority trial. Lancet. 2012 Aug 4;380(9840):475-83. PMID: 22748821. 195. Goke B, Gallwitz B, Eriksson J, et al. Saxagliptin is non-inferior to glipizide in patients with type 2 diabetes mellitus inadequately controlled on metformin alone: a 52-week randomised controlled trial. Int J Clin Pract. 2010 Nov;64(12):1619-31. PMID: 20846286.
- 396. 339 196. Seck T,Nauck M, Sheng D, et al. Safety and efficacy of treatment with sitagliptin or glipizide in patients with type 2 diabetes inadequately controlled on metformin: a 2- year study. Int J Clin Pract. 2010 Apr;64(5):562-76. PMID: 20456211. 197. Del Prato S, Camisasca R, Wilson C, Fleck P. Durability of the efficacy and safety of alogliptin compared with glipizide in type 2 diabetes mellitus: a 2-year study. Diabetes Obes Metab. 2014 Dec;16(12):1239-46. PMID: 25132212. 198. Cefalu WT, Leiter LA, Yoon KH, et al. Efficacy and safety of canagliflozin versus glimepiride in patients with type 2 diabetes inadequately controlled with metformin (CANTATA-SU): 52 week results from a randomised, double-blind, phase 3 non- inferiority trial. Lancet. 2013 Sep 14;382(9896):941-50. PMID: 23850055. 199. Nauck MA, Del Prato S, Meier JJ, et al. Dapagliflozin versus glipizide as add-on therapy in patients with type 2 diabetes who have inadequate glycemic control with metformin: a randomized, 52-week, double- blind, active-controlled noninferiority trial. Diabetes Care. 2011 Sep;34(9):2015-22. PMID: 21816980. 200. Ridderstrale M, Andersen KR, Zeller C, et al. Comparison of empagliflozin and glimepiride as add-on to metformin in patients with type 2 diabetes: a 104-week randomised, active-controlled, double-blind, phase 3 trial. The Lancet Diabetes and Endocrinology. 2014((Ridderstrale M., [email protected]) Steno Diabetes Center, Gentofte, Denmark). PMID: 24948511 201. Leiter LA, Yoon KH, Arias P, et al. Canagliflozin provides durable glycemic improvements and body weight reduction over 104 weeks versus glimepiride in patients with type 2 diabetes on metformin: a randomized, double-blind, phase 3 study. Diabetes Care. 2015 Mar;38(3):355-64. PMID: 25205142. 202. Zhang H, Zhang X, Hu C, Lu W. Exenatide reduces urinary transforming growth factor- beta1 and type IV collagen excretion in patients with type 2 diabetes and microalbuminuria. Kidney Blood Press Res. 2012;35(6):483-8. PMID: 22687869. 203. Derosa G, Putignano P, Bossi AC, et al. Exenatide or glimepiride added to metformin on metabolic control and on insulin resistance in type 2 diabetic patients. Eur J Pharmacol. 2011 Sep;666(1-3):251-6. PMID: 21645507. 204. Yang W, Chen L, Ji Q, et al. Liraglutide provides similar glycaemic control as glimepiride (both in combination with metformin) and reduces body weight and systolic blood pressure in Asian population with type 2 diabetes from China, South Korea and India: a 16-week, randomized, double-blind, active control trial(*). Diabetes Obes Metab. 2011 Jan;13(1):81-8. PMID: 21114607. 205. Derosa G, Maffioli P, Salvadeo SA, et al. Exenatide versus glibenclamide in patients with diabetes. Diabetes Technol Ther. 2010 Mar;12(3):233-40. PMID: 20151774. 206. Moon JS, Ha KS, Yoon JS, et al. The effect of glargine versus glimepiride on pancreatic beta-cell function in patients with type 2 diabetes uncontrolled on metformin monotherapy: open-label, randomized, controlled study. Acta Diabetol. 2014 Apr;51(2):277-85. PMID: 24445656. 207. Malone JK, Beattie SD, Campaigne BN, et al. Therapy after single oral agent failure: adding a second oral agent or an insulin mixture? Diabetes Res Clin Pract. 2003 2003 Dec;62(3):187-95. PMID: 14625133 208. Kvapil M, Swatko A, Hilberg C, Shestakova M. Biphasic insulin aspart 30 plus metformin: an effective combination in type 2 diabetes. Diabetes Obes Metab. 2006 2006 Jan;8(1):39-48. PMID: 16367881 209. Rosenstock J, Hansen L, Zee P, et al. Dual Add-on Therapy in Type 2 Diabetes Poorly Controlled With Metformin Monotherapy: A Randomized Double-Blind Trial of Saxagliptin Plus Dapagliflozin Addition Versus Single Addition of Saxagliptin or Dapagliflozin to Metformin. Diabetes Care. 2015 Mar;38(3):376-83. PMID: 25352655. 210. Pratley RE, Nauck M, Bailey T, et al. Liraglutide versus sitagliptin for patients with type 2 diabetes who did not have adequate glycaemic control with metformin: a 26-week, randomised, parallel-group, open-label trial. Lancet. 2010 Apr 24;375(9724):1447-56. PMID: 20417856.
- 397. 340 211. Aschner P,Chan J, Owens DR, et al. Insulin glargine versus sitagliptin in insulin-naive patients with type 2 diabetes mellitus uncontrolled on metformin (EASIE): a multicentre, randomised open-label trial. Lancet. 2012 Jun 16;379(9833):2262-9. PMID: 22683131. 212. Diamant M, Van Gaal L, Stranks S, et al. Once weekly exenatide compared with insulin glargine titrated to target in patients with type 2 diabetes (DURATION-3): an open-label randomised trial. Lancet. 2010 Jun 26;375(9733):2234-43. PMID: 20609969. 213. Gallwitz B, Bohmer M, Segiet T, et al. Exenatide twice daily versus premixed insulin aspart 70/30 in metformin-treated patients with type 2 diabetes: a randomized 26-week study on glycemic control and hypoglycemia. Diabetes Care. 2011 Mar;34(3):604-6. PMID: 21285388. 214. Robbins DC, Beisswenger PJ, Ceriello A, et al. Mealtime 50/50 basal + prandial insulin analogue mixture with a basal insulin analogue, both plus metformin, in the achievement of target HbA1c and pre- and postprandial blood glucose levels in patients with type 2 diabetes: a multinational, 24- week, randomized, open-label, parallel- group comparison. Clin Ther. 2007 2007 Nov;29(11):2349-64. PMID: 18158076 215. Raskin PR, Hollander PA, Lewin A, et al. Basal insulin or premix analogue therapy in type 2 diabetes patients. Eur J Intern Med. 2007 2007 Jan;18(1):56-62. PMID: 17223044 216. Davies MJ, Thaware PK, Tringham JR, et al. A randomized controlled trial examining combinations of repaglinide, metformin and NPH insulin. Diabet Med. 2007 2007 Jul;24(7):714-9. PMID: 17403126 217. St John Sutton M, Rendell M, Dandona P, et al. A comparison of the effects of rosiglitazone and glyburide on cardiovascular function and glycemic control in patients with type 2 diabetes. Diabetes Care. 2002 2002 Nov;25(11):2058- 64. PMID: 12401757 218. Derosa G, Cicero AF, Gaddi AV, et al. Long-term effects of glimepiride or rosiglitazone in combination with metformin on blood pressure control in type 2 diabetic patients affected by the metabolic syndrome: a 12-month, double-blind, randomized clinical trial. Clin Ther. 2005 2005 Sep;27(9):1383-91. PMID: 16291411 219. Nauck MA, Del Prato S, Duran-Garcia S, et al. Durability of glycaemic efficacy over 2 years with dapagliflozin versus glipizide as add-on therapies in patients whose type 2 diabetes mellitus was inadequately controlled with metformin. Diabetes Obes Metab. 2014 Jun 12. PMID: 24919526. 220. Adler AI, Stratton IM, Neil HA, et al. Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36): prospective observational study. BMJ. 2000 Aug 12;321(7258):412-9. PMID: 10938049. 221. Holman RR, Paul SK, Bethel MA, et al. Long-term follow-up after tight control of blood pressure in type 2 diabetes. The New England journal of medicine. 2008 Oct 9;359(15):1565-76. PMID: 18784091. 222. Turnbull F, Neal B, Algert C, et al. Effects of different blood pressure-lowering regimens on major cardiovascular events in individuals with and without diabetes mellitus: results of prospectively designed overviews of randomized trials. Arch Intern Med. 2005 Jun 27;165(12):1410-9. PMID: 15983291. 223. Malone JK, Kerr LF, Campaigne BN, et al. Combined therapy with insulin lispro Mix 75/25 plus metformin or insulin glargine plus metformin: a 16-week, randomized, open-label, crossover study in patients with type 2 diabetes beginning insulin therapy. Clin Ther. 2004 2004 Dec;26(12):2034-44. PMID: 15823767 224. Malone JK, Bai S, Campaigne BN, et al. Twice-daily pre-mixed insulin rather than basal insulin therapy alone results in better overall glycaemic control in patients with Type 2 diabetes. Diabet Med. 2005 2005 Apr;22(4):374-81. PMID: 15787659
- 398. 341 225. Andersson C,Olesen JB, Hansen PR, et al. Metformin treatment is associated with a low risk of mortality in diabetic patients with heart failure: a retrospective nationwide cohort study. Diabetologia. 2010 Dec;53(12):2546-53. PMID: 20838985. 226. Horsdal HT, Sondergaard F, Johnsen SP, Rungby J. Antidiabetic treatments and risk of hospitalisation with myocardial infarction: a nationwide case-control study. Pharmacoepidemiol Drug Saf. 2011 Apr;20(4):331-7. PMID: 21442682. 227. Mogensen UM, Andersson C, Fosbol EL, et al. Cardiovascular safety of combination therapies with incretin-based drugs and metformin compared with a combination of metformin and sulphonylurea in type 2 diabetes mellitus - a retrospective nationwide study. Diabetes Obes Metab. 2014 May 14. PMID: 24827939. 228. Scheller NM, Mogensen UM, Andersson C, et al. All-cause mortality and cardiovascular effects associated with the DPP-IV inhibitor sitagliptin compared with metformin, a retrospective cohort study on the Danish population. Diabetes, obesity & metabolism. 2014;16(3):231-6. PMID: 24020750 229. Schramm TK, Gislason GH, Vaag A, et al. Mortality and cardiovascular risk associated with different insulin secretagogues compared with metformin in type 2 diabetes, with or without a previous myocardial infarction: a nationwide study. Eur Heart J. 2011 Aug;32(15):1900-8. PMID: 21471135. 230. Johnson JA, Simpson SH, Toth EL, Majumdar SR. Reduced cardiovascular morbidity and mortality associated with metformin use in subjects with Type 2 diabetes. Diabet Med. 2005 2005 Apr;22(4):497-502. PMID: 15787679 231. Hong J, Zhang Y, Lai S, et al. Effects of metformin versus glipizide on cardiovascular outcomes in patients with type 2 diabetes and coronary artery disease. Diabetes Care. 2013 May;36(5):1304-11. PMID: 23230096. 232. Raskin P, Lewin A, Reinhardt R, Lyness W. Twice-daily dosing of a repaglinide/metformin fixed-dose combination tablet provides glycaemic control comparable to rosiglitazone/metformin tablet. Diabetes Obes Metab. 2009 2009 May 19. PMID: 19476470 233. Pantalone KM, Kattan MW, Yu C, et al. The risk of developing coronary artery disease or congestive heart failure, and overall mortality, in type 2 diabetic patients receiving rosiglitazone, pioglitazone, metformin, or sulfonylureas: a retrospective analysis. Acta Diabetol. 2009 2009 Jun;46(2):145-54. PMID: 19194648 234. Wheeler S, Moore K, Forsberg CW, et al. Mortality among veterans with type 2 diabetes initiating metformin, sulfonylurea or rosiglitazone monotherapy. Diabetologia. 2013 Jun 25. PMID: 23797633. 235. Kahler KH, Rajan M, Rhoads GG, et al. Impact of oral antihyperglycemic therapy on all-cause mortality among patients with diabetes in the Veterans Health Administration. Diabetes Care. 2007 2007 Jul;30(7):1689-93. PMID: 17440170 236. Wang CP, Lorenzo C, Espinoza SE. Frailty Attenuates the Impact of Metformin on Reducing Mortality in Older Adults with Type 2 Diabetes. J Endocrinol Diabetes Obes. 2014;2(2). PMID: 25506599. 237. Pantalone KM, Kattan MW, Yu C, et al. Increase in overall mortality risk in patients with type 2 diabetes receiving glipizide, glyburide or glimepiride monotherapy versus metformin: a retrospective analysis. Diabetes Obes Metab. 2012 Sep;14(9):803- 9. PMID: 22486923. 238. Corrao G, Romio SA, Zambon A, et al. Multiple outcomes associated with the use of metformin and sulphonylureas in type 2 diabetes: a population-based cohort study in Italy. Eur J Clin Pharmacol. 2011 Mar;67(3):289-99. PMID: 21088829. 239. Ferrannini E, Seman L, Seewaldt-Becker E, et al. A Phase IIb, randomized, placebo- controlled study of the SGLT2 inhibitor empagliflozin in patients with type 2 diabetes. Diabetes, obesity & metabolism. 2013;15(8):721-8. PMID: 23398530
- 399. 342 240. Stenlof K,Cefalu WT, Kim KA, et al. Long- term efficacy and safety of canagliflozin monotherapy in patients with type 2 diabetes inadequately controlled with diet and exercise: findings from the 52-week CANTATA-M study. Curr Med Res Opin. 2014 Feb;30(2):163-75. PMID: 24073995. 241. Jones TA, Sautter M, Van Gaal LF, Jones NP. Addition of rosiglitazone to metformin is most effective in obese, insulin-resistant patients with type 2 diabetes. Diabetes Obes Metab. 2003 2003 May;5(3):163-70. PMID: 12681023 242. Prentice JC, Conlin PR, Gellad WF, et al. Capitalizing on prescribing pattern variation to compare medications for type 2 diabetes. Value Health. 2014 Dec;17(8):854-62. PMID: 25498781. 243. Hsiao FY, Huang WF, Wen YW, et al. Thiazolidinediones and cardiovascular events in patients with type 2 diabetes mellitus: a retrospective cohort study of over 473,000 patients using the National Health Insurance database in Taiwan. Drug Saf. 2009 2009;32(8):675-90. PMID: 19591532 244. Brownstein JS, Murphy SN, Goldfine AB, et al. Rapid identification of myocardial infarction risk associated with diabetes medications using electronic medical records. Diabetes Care. 2010 2010 Mar;33(3):526-31. PMID: 20009093 245. Hung YC, Lin CC, Wang TY, et al. Oral hypoglycaemic agents and the development of non-fatal cardiovascular events in patients with type 2 diabetes mellitus. Diabetes Metab Res Rev. 2013 Nov;29(8):673-9. PMID: 23956007. 246. Roumie CL, Hung AM, Greevy RA, et al. Comparative effectiveness of sulfonylurea and metformin monotherapy on cardiovascular events in type 2 diabetes mellitus: a cohort study. Ann Intern Med. 2012 Nov 6;157(9):601-10. PMID: 23128859. 247. Stewart MW, Cirkel DT, Furuseth K, et al. Effect of metformin plus roziglitazone compared with metformin alone on glycaemic control in well-controlled Type 2 diabetes. Diabet Med. 2006 2006 Oct;23(10):1069-78. PMID: 16978370 248. Goke B, Gallwitz B, Eriksson JG, et al. Saxagliptin vs. glipizide as add-on therapy in patients with type 2 diabetes mellitus inadequately controlled on metformin alone: long-term (52-week) extension of a 52-week randomised controlled trial. Int J Clin Pract. 2013 Apr;67(4):307-16. PMID: 23638466. 249. Hung AM, Roumie CL, Greevy RA, et al. Comparative effectiveness of incident oral antidiabetic drugs on kidney function. Kidney Int. 2012 Apr;81(7):698-706. PMID: 22258320. 250. Masica AL, Ewen E, Daoud YA, et al. Comparative effectiveness research using electronic health records: impacts of oral antidiabetic drugs on the development of chronic kidney disease. Pharmacoepidemiol Drug Saf. 2013 Apr;22(4):413-22. PMID: 23436488. 251. Amador-Licona N, Guizar-Mendoza J, Vargas E, et al. The short-term effect of a switch from glibenclamide to metformin on blood pressure and microalbuminuria in patients with type 2 diabetes mellitus. Arch Med Res. 2000 2000 Nov-Dec;31(6):571-5. PMID: 11257323 252. Hung AM, Roumie CL, Greevy RA, et al. Kidney function decline in metformin versus sulfonylurea initiators: assessment of time- dependent contribution of weight, blood pressure, and glycemic control. Pharmacoepidemiol Drug Saf. 2013 Jun;22(6):623-31. PMID: 23592561. 253. Agarwal R, Saha C, Battiwala M, et al. A pilot randomized controlled trial of renal protection with pioglitazone in diabetic nephropathy. Kidney Int. 2005 2005 Jul;68(1):285-92. PMID: 15954919 254. Lehman DM, Lorenzo C, Hernandez J, Wang CP. Statin use as a moderator of metformin effect on risk for prostate cancer among type 2 diabetic patients. Diabetes Care. 2012 May;35(5):1002-7. PMID: 22456867. 255. van Staa TP, Patel D, Gallagher AM, de Bruin ML. Glucose-lowering agents and the patterns of risk for cancer: a study with the General Practice Research Database and secondary care data. Diabetologia. 2012 Mar;55(3):654-65. PMID: 22127412.
- 400. 343 256. Skrivanek Z,Gaydos BL, Chien JY, et al. Dose-finding results in an adaptive, seamless, randomized trial of once-weekly dulaglutide combined with metformin in type 2 diabetes patients (AWARD-5). Diabetes Obes Metab. 2014 Apr 25. PMID: 24762094. 257. Derosa G, Franzetti I, Gadaleta G, et al. Metabolic variations with oral antidiabetic drugs in patients with Type 2 diabetes: comparison between glimepiride and metformin. Diabetes Nutr Metab. 2004 2004 Jun;17(3):143-50. PMID: 15334791 258. Wright AD, Cull CA, Macleod KM, Holman RR. Hypoglycemia in Type 2 diabetic patients randomized to and maintained on monotherapy with diet, sulfonylurea, metformin, or insulin for 6 years from diagnosis: UKPDS73. J Diabetes Complications. 2006 2006 Nov- Dec;20(6):395-401. PMID: 17070446 259. Weir MA, Gomes T, Mamdani M, et al. Impaired renal function modifies the risk of severe hypoglycaemia among users of insulin but not glyburide: a population-based nested case-control study. Nephrol Dial Transplant. 2011 Jun;26(6):1888-94. PMID: 20974644. 260. Gupta A, Ahmad Ansari N, Yadav N. Comparative efficacy and safety of sitagliptin and Glimepiride in patients of newly diagnosed type 2 diabetes mellitus. International Journal of Pharmaceutical Sciences Review and Research. 2013;23(2):137-41. PMID: COULD NOT FIND THIS ONE 261. Nauck M, Frid A, Hermansen K, et al. Efficacy and safety comparison of liraglutide, glimepiride, and placebo, all in combination with metformin, in type 2 diabetes: the LEAD (liraglutide effect and action in diabetes)-2 study. Diabetes Care. 2009 2009 Jan;32(1):84-90. PMID: 18931095. 262. Lee YK, Song SO, Kim KJ, et al. Glycemic Effectiveness of Metformin-Based Dual- Combination Therapies with Sulphonylurea, Pioglitazone, or DPP4-Inhibitor in Drug- Naive Korean Type 2 Diabetic Patients. Diabetes Metab J. 2013 Dec;37(6):465-74. PMID: 24404518. 263. Srivastava S, Saxena GN, Keshwani P, Gupta R. Comparing the efficacy and safety profile of sitagliptin versus glimepiride in patients of type 2 diabetes mellitus inadequately controlled with metformin alone. J Assoc Physicians India. 2012 Mar;60:27-30. PMID: 22799111. 264. Davies M, Heller S, Sreenan S, et al. Once- weekly exenatide versus once- or twice- daily insulin detemir: randomized, open- label, clinical trial of efficacy and safety in patients with type 2 diabetes treated with metformin alone or in combination with sulfonylureas. Diabetes Care. 2013 May;36(5):1368-76. PMID: 23275363. 265. Derosa G, Gaddi AV, Ciccarelli L, et al. Long-term effect of glimepiride and rosiglitazone on non-conventional cardiovascular risk factors in metformin- treated patients affected by metabolic syndrome: a randomized, double-blind clinical trial. J Int Med Res. 2005 2005 May-Jun;33(3):284-94. PMID: 15938589 266. Kowall B, Rathmann W, Kostev K. Are sulfonylurea and insulin therapies associated with a larger risk of cancer than metformin therapy? A retrospective database analysis. Diabetes Care. 2015 Jan;38(1):59-65. PMID: 25336750. 267. Bolinder J, Ljunggren O, Johansson L, et al. Dapagliflozin maintains glycaemic control while reducing weight and body fat mass over 2 years in patients with type 2 diabetes mellitus inadequately controlled on metformin. Diabetes, obesity & metabolism. 2014;16(2):159-69. PMID: 23906445 268. Perez-Monteverde A, Seck T, Xu L, et al. Efficacy and safety of sitagliptin and the fixed-dose combination of sitagliptin and metformin vs. pioglitazone in drug-naive patients with type 2 diabetes. Int J Clin Pract. 2011 Sep;65(9):930-8. PMID: 21849007. 269. Turner RC, Cull CA, Frighi V, Holman RR. Glycemic control with diet, sulfonylurea, metformin, or insulin in patients with type 2 diabetes mellitus: progressive requirement for multiple therapies (UKPDS 49). UK Prospective Diabetes Study (UKPDS) Group. JAMA. 1999 1999 Jun 2;281(21):2005-12. PMID: 10359389
- 401. 344 270. Kawai T,Funae O, Shimada A, et al. Effects of pretreatment with low-dose metformin on metabolic parameters and weight gain by pioglitazone in Japanese patients with type 2 diabetes. Intern Med. 2008 2008;47(13):1181-8. PMID: 18591838 271. United Kingdom Prospective Diabetes Study 24: a 6-year, randomized, controlled trial comparing sulfonylurea, insulin, and metformin therapy in patients with newly diagnosed type 2 diabetes that could not be controlled with diet therapy. United Kingdom Prospective Diabetes Study Group. Ann Intern Med. 1998 1998 Feb 1;128(3):165-75. PMID: 9454524 272. Gallwitz B, Rosenstock J, Patel S, et al. Regardless of the degree of glycaemic control, linagliptin has lower hypoglycaemia risk than all doses of glimepiride, at all time points, over the course of a 2-year trial. Diabetes Obes Metab. 2015 Mar;17(3):276- 84. PMID: 25425502. 273. Kahn SE, Zinman B, Lachin JM, et al. Rosiglitazone-associated fractures in type 2 diabetes: an Analysis from A Diabetes Outcome Progression Trial (ADOPT). Diabetes Care. 2008 2008 May;31(5):845- 51. PMID: 18223031 274. Dormuth CR, Carney G, Carleton B, et al. Thiazolidinediones and fractures in men and women. Arch Intern Med. 2009 2009 Aug 10;169(15):1395-402. PMID: 19667303 275. Schellhase KG, Koepsell TD, Weiss NS. Glycemic control and the risk of multiple microvascular diabetic complications. Fam Med. 2005 Feb;37(2):125-30. PMID: 15690253. 276. Vijan S, Hofer TP, Hayward RA. Estimated benefits of glycemic control in microvascular complications in type 2 diabetes. Ann Intern Med. 1997 Nov 1;127(9):788-95. PMID: 9382399. 277. Shyangdan DS, Royle P, Clar C, et al. Glucagon-like peptide analogues for type 2 diabetes mellitus. Cochrane Database of Systematic Reviews [serial on the Internet]. 2011; (10): Available from: http://onlinelibrary.wiley.com/doi/10.1002/14651 858.CD006423.pub2/abstract. 278. Liu SC, Tu YK, Chien MN, Chien KL. Effect of antidiabetic agents added to metformin on glycaemic control, hypoglycaemia and weight change in patients with type 2 diabetes: A network meta-analysis. Diabetes, obesity & metabolism. 2012 Sep;14(9):810-20. PMID: 22486990. 279. McIntosh B, Cameron C, Singh SR, et al. Second-line therapy in patients with type 2 diabetes inadequately controlled with metformin monotherapy: a systematic review and mixed-treatment comparison meta-analysis. Open Med. 2011;5(1):e35-48. PMID: 22046219. 280. Kahn BB, Flier JS. Obesity and insulin resistance. J Clin Invest. 2000 Aug;106(4):473-81. PMID: 10953022. 281. Purnell TS, Joy S, Little E, et al. Patient preferences for noninsulin diabetes medications: a systematic review. Diabetes Care. 2014 Jul;37(7):2055-62. PMID: 24963113. 282. Vasilakou D, Karagiannis T, Athanasiadou E, et al. Sodium-glucose cotransporter 2 inhibitors for type 2 diabetes: a systematic review and meta-analysis. Ann Intern Med. 2013 Aug 20;159(4):262-74. PMID: 24026259. 283. Richter B, Bandeira-Echtler E, Bergerhoff K, Lerch C. Dipeptidyl peptidase-4 (DPP-4) inhibitors for type 2 diabetes mellitus. Cochrane Database of Systematic Reviews [serial on the Internet]. 2008; (2): Available from: http://onlinelibrary.wiley.com/doi/10.1002/14651 858.CD006739.pub2/abstract. 284. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program (SHEP). SHEP Cooperative Research Group. JAMA. 1991 Jun 26;265(24):3255-64. PMID: 2046107. 285. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. UK Prospective Diabetes Study Group. BMJ. 1998 Sep 12;317(7160):703-13. PMID: 9732337.
- 402. 345 286. Wang JG,Staessen JA, Gong L, Liu L. Chinese trial on isolated systolic hypertension in the elderly. Systolic Hypertension in China (Syst-China) Collaborative Group. Arch Intern Med. 2000 Jan 24;160(2):211-20. PMID: 10647760. 287. Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. DASH-Sodium Collaborative Research Group. The New England journal of medicine. 2001 Jan 4;344(1):3-10. PMID: 11136953. 288. Wu J, Kraja AT, Oberman A, et al. A summary of the effects of antihypertensive medications on measured blood pressure. Am J Hypertens. 2005 Jul;18(7):935-42. PMID: 16053990. 289. Nauman J, Janszky I, Vatten LJ, Wisloff U. Temporal changes in resting heart rate and deaths from ischemic heart disease. JAMA. 2011 Dec 21;306(23):2579-87. PMID: 22187277. 290. Phung OJ, Schwartzman E, Allen RW, et al. Sulphonylureas and risk of cardiovascular disease: systematic review and meta- analysis. Diabet Med. 2013 Oct;30(10):1160-71. PMID: 23663156. 291. Monami M, Genovese S, Mannucci E. Cardiovascular safety of sulfonylureas: A meta-analysis of randomized clinical trials. Diabetes, obesity & metabolism. 2013 Oct;15(10):938-53. PMID: 23594109. 292. Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. The New England journal of medicine. 2007 Jun 14;356(24):2457-71. PMID: 17517853. 293. U.S. Food and Drug Administration. FDA Drug Safety Communication: Updated Risk Evaluation and Mitigation Strategy (REMS) to Restrict Acces to Rosiglitazone- containing Medicines including Avandia, Avandamet, and Avandaryl. 2013; http://www.fda.gov/Drugs/DrugSafety/ucm25500 5.htm. Accessed 2015 February 26. 294. U.S. Food and Drug Administration. FDA significantly restricts access to the diabetes drug Avandia. 2010; http://www.fda.gov/Drugs/DrugSafety/Postmarke tDrugSafetyInformationforPatientsandProviders/ ucm226956.htm. Accessed 2015 February 26. 295. U.S. Food and Drug Administration. FDA requires removal of some prescribing and dispensing restrictions for rosiglitazone- containing diabetes medicines. 2013; http://www.fda.gov/downloads/Drugs/DrugSafet y/UCM381108.pdf. Accessed 2015 February 25. 296. Goossen K, Graber S. Longer term safety of dipeptidyl peptidase-4 inhibitors in patients with type 2 diabetes mellitus: systematic review and meta-analysis. Diabetes Obes Metab. 2012 Dec;14(12):1061-72. PMID: 22519906. 297. Wu D, Li L, Liu C. Efficacy and safety of dipeptidyl peptidase-4 inhibitors and metformin as initial combination therapy and as monotherapy in patients with type 2 diabetes mellitus: a meta-analysis. Diabetes Obes Metab. 2014 Jan;16(1):30-7. PMID: 23803146. 298. Karagiannis T, Paschos P, Paletas K, et al. Dipeptidyl peptidase-4 inhibitors for treatment of type 2 diabetes mellitus in the clinical setting: Systematic review and meta- analysis. BMJ (Online). 2012;344(7850):17. PMID: 22411919 299. Monami M, Ahren B, Dicembrini I, Mannucci E. Dipeptidyl peptidase-4 inhibitors and cardiovascular risk: Ameta- analysis of randomized clinical trials. Diabetes, obesity & metabolism. 2013;15(2):112-20. PMID: 22925682 300. Johansen OE, Neubacher D, von Eynatten M, et al. Cardiovascular safety with linagliptin in patients with type 2 diabetes mellitus: a pre-specified, prospective, and adjudicated meta-analysis of a phase 3 programme. Cardiovasc Diabetol. 2012;11:3. PMID: 22234149. 301. Iqbal N, Parker A, Frederich R, et al. Assessment of the cardiovascular safety of saxagliptin in patients with type 2 diabetes mellitus: pooled analysis of 20 clinical trials. Cardiovasc Diabetol. 2014;13:33. PMID: 24490835.
- 403. 346 302. Scirica BM,Bhatt DL, Braunwald E, et al. Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus. The New England journal of medicine. 2013 Oct 3;369(14):1317-26. PMID: 23992601. 303. Green JB, Bethel MA, Armstrong PW, et al. Effect of Sitagliptin on Cardiovascular Outcomes in Type 2 Diabetes. The New England journal of medicine. 2015 Jul 16;373(3):232-42. PMID: 26052984. 304. White WB, Cannon CP, Heller SR, et al. Alogliptin after acute coronary syndrome in patients with type 2 diabetes. The New England journal of medicine. 2013 Oct 3;369(14):1327-35. PMID: 23992602. 305. Monami M, Dicembrini I, Nardini C, et al. Effects of glucagon-like peptide-1 receptor agonists on cardiovascular risk: A meta- analysis of randomized clinical trials. Diabetes, obesity & metabolism. 2014;16(1):38-47. PMID: 23829656 306. Bonds DE, Miller ME, Bergenstal RM, et al. The association between symptomatic, severe hypoglycaemia and mortality in type 2 diabetes: retrospective epidemiological analysis of the ACCORD study. BMJ. 2010;340:b4909. PMID: 20061358. 307. Holman RR, Farmer AJ, Davies MJ, et al. Three-year efficacy of complex insulin regimens in type 2 diabetes. The New England journal of medicine. 2009 Oct 29;361(18):1736-47. PMID: 19850703. 308. Budnitz DS, Shehab N, Kegler SR, Richards CL. Medication use leading to emergency department visits for adverse drug events in older adults. Ann Intern Med. 2007 Dec 4;147(11):755-65. PMID: 18056659. 309. Franciosi M, Lucisano G, Lapice E, et al. Metformin therapy and risk of cancer in patients with type 2 diabetes: systematic review. PLoS One. 2013;8(8):e71583. PMID: 23936520. 310. Zhang ZJ, Bi Y, Li S, et al. Reduced risk of lung cancer with metformin therapy in diabetic patients: a systematic review and meta-analysis. Am J Epidemiol. 2014 Jul 1;180(1):11-4. PMID: 24920786. 311. Ferwana M, Firwana B, Hasan R, et al. Pioglitazone and risk of bladder cancer: a meta-analysis of controlled studies. Diabet Med. 2013 Sep;30(9):1026-32. PMID: 23350856. 312. Dormandy JA, Charbonnel B, Eckland DJ, et al. Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial In macroVascular Events): a randomised controlled trial. Lancet. 2005 Oct 8;366(9493):1279-89. PMID: 16214598. 313. Erdmann E, Song E, Spanheimer R, et al. Observational follow-up of the PROactive study: a 6-year update. Diabetes Obes Metab. 2014 Jan;16(1):63-74. PMID: 23859428. 314. U.S. Food and Drug Administration. Incretin Mimetic Drugs for Type 2 Diabetes: Early Communication - Reports of Possible Increased Risk of Pancreatitis and Pre- cancerous Findings of the Pancreas. 2013; http://www.fda.gov/Safety/MedWatch/SafetyInfo rmation/SafetyAlertsforHumanMedicalProducts/ ucm343805.htm. Accessed 2015 August 1. 315. U.S. Food and Drug Administration. Highlights of Prescribing Information: Victoza (liraglutide [rDNA origin] injection, solution for subcutaneous use. 2011; http://www.accessdata.fda.gov/drugsatfda_docs/l abel/2011/022341s004lbl.pdf. Accessed 2015 March 2. 316. U.S. Food and Drug Administration. Highlights of Prescribing Information: Tanzeum (albiglutide) for injection, for subcutaneous use. 2014; http://www.accessdata.fda.gov/drugsatfda_docs/l abel/2014/125431s000lbl.pdf. Accessed 2015 March 2. 317. U.S. Food and Drug Administration. Highlights of Prescribing Information: Bydureon (exenatide extended-release) for injectable suspension. 2015; http://www.fda.gov/safety/medwatch/safetyinfor mation/ucm400570.htm. Accessed 2015 August 7. 318. U.S. Food and Drug Administration. Trulicity (dulaglutide) Injection, for Subcutaneous Use. 2015; http://www.fda.gov/safety/medwatch/safetyinfor mation/ucm442202.htm. Accessed 2015 August 7.
- 404. 347 319. U.S. Foodand Drug Administration. Victoza (liraglutide [rDNA origin]) Injection: REMS - Risk of Thyroid C-cell Tumors, Acute Pancreatitis. 2011; http://www.fda.gov/Safety/MedWatch/SafetyInfo rmation/SafetyAlertsforHumanMedicalProducts/ ucm258826.htm; . Accessed 2015 February 25. 320. Lago RM, Singh PP, Nesto RW. Congestive heart failure and cardiovascular death in patients with prediabetes and type 2 diabetes given thiazolidinediones: a meta-analysis of randomised clinical trials. Lancet. 2007 Sep 29;370(9593):1129-36. PMID: 17905165. 321. Singh S, Loke YK, Furberg CD. Long-term risk of cardiovascular events with rosiglitazone: a meta-analysis. JAMA. 2007 Sep 12;298(10):1189-95. PMID: 17848653. 322. GlaxoSmithKline. Highlights of Prescribing Information: Avandia (rosiglitazone maleate) tablets. 2014; http://us.gsk.com/products/assets/us_avandia.pdf. Accessed 2015 March 2. 323. Takeda Pharmaceuticals America. Highlights of Prescribing Information: Actos (pioglitazone) tables for oral use. 2013; http://general.takedapharm.com/content/file/pi.pd f?applicationcode=8a9c4571-a123-4477- 91deb9cafe7d07e3&filetypecode=actospi. Accessed 2015 March 2. 324. Zannad F, Cannon CP, Cushman WC, et al. Heart failure and mortality outcomes in patients with type 2 diabetes taking alogliptin versus placebo in EXAMINE: a multicentre, randomised, double-blind trial. Lancet. 2015 May 23;385(9982):2067-76. PMID: 25765696. 325. U.S. Food and Drug Administration. FDA Panel Wants New DPP-4 Inhibitor Labels - Cardiovascular data warrant new risk information for saxagliptin and alogliptin, advisers say.; http://www.medpagetoday.com/PublicHealthPoli cy/ClinicalTrials/50990. Accessed 2015 July 25, . 326. Clinicaltrials.gov. CAROLINA: Cardiovascular Outcome Study of Linagliptin Versus Glimepiride in Patients With Type 2 Diabetes. 2010; https://clinicaltrials.gov/ct2/show/NCT0124 3424. Accessed 2015 July 30, . 327. Clinicaltrials.gov. Cardiovascular and Renal Microvascular Outcome Study With Linagliptin in Patients With Type 2 Diabetes Mellitus (CARMELINA). 2013; https://clinicaltrials.gov/ct2/show/NCT0189 7532. Accessed 2015 July 30, . 328. NESINA (alogliptin) tablets. 2015; http://general.takedapharm.com/content/file.aspx ?FileTypeCode=NESINAPI&cacheRandomizer= f3b08184-5c02-4cd5-a603-466fafb32324. Accessed 2015 August 7. 329. ACTOS (pioglitazone) tablets for oral use. 2013; http://general.takedapharm.com/content/file.aspx ?filetypecode=actospi&cacheRandomizer=bd24d b50-cfd1-4e08-a5f0-1414e6c0d33a. Accessed 2015 August 7. 330. Brown JB, Pedula K, Barzilay J, et al. Lactic acidosis rates in type 2 diabetes. Diabetes Care. 1998 Oct;21(10):1659-63. PMID: 9773726. 331. Misbin RI, Green L, Stadel BV, et al. Lactic acidosis in patients with diabetes treated with metformin. The New England journal of medicine. 1998 Jan 22;338(4):265-6. PMID: 9441244. 332. Salpeter SR, Greyber E, Pasternak GA, Salpeter EE. Risk of fatal and nonfatal lactic acidosis with metformin use in type 2 diabetes mellitus. Cochrane Database Syst Rev. 2010(4):CD002967. PMID: 20393934. 333. Inzucchi SE, Lipska KJ, Mayo H, et al. Metformin in patients with type 2 diabetes and kidney disease: a systematic review. JAMA. 2014 Dec 24-31;312(24):2668-75. PMID: 25536258. 334. Sethi BK, Viswanathan V, Kumar A, et al. Liraglutide in Clinical Practice: Insights from LEAD Programme. Supplement to JAPI. 2010 June 2010;58:18-22. PMID: COULD NOT FIND THIS ONE 335. Franks AS, Lee PH, George CM. Pancreatitis: a potential complication of liraglutide? Ann Pharmacother. 2012 Nov;46(11):1547-53. PMID: 23136352. 336. Li L, Shen J, Bala MM, et al. Incretin treatment and risk of pancreatitis in patients with type 2 diabetes mellitus: systematic review and meta-analysis of randomised and non-randomised studies. BMJ. 2014;348:g2366. PMID: 24736555.
- 405. 348 337. Wang T,Wang F, Gou Z, et al. Using real- world data to evaluate the association of incretin-based therapies with risk of acute pancreatitis: a meta-analysis of 1,324,515 patients from observational studies. Diabetes Obes Metab. 2015 Jan;17(1):32-41. PMID: 25200423. 338. Karagiannis T, Boura P, Tsapas A. Safety of dipeptidyl peptidase 4 inhibitors: A perspective review. Therapeutic Advances in Drug Safety. 2014;5(3):138-46. PMID: 25083269 339. U.S. Food and Drug Administration. Highlights of Prescribing Information: Byetta (exenatide) Injection. 2014; http://www.accessdata.fda.gov/drugsatfda_docs/l abel/2014/021773s036lbl.pdf Accessed 2015 March 2. 340. U.S. Food and Drug Administration. Highlights of Prescribing Information: Victoza (liraglutide). 2013; http://www.accessdata.fda.gov/drugsatfda_docs/l abel/2013/022341s020lbl.pdf Accessed 2015 March 2. 341. U.S. Food and Drug Administration. Summary Review: 125469Orig1s000. 2013; http://www.accessdata.fda.gov/drugsatfda_docs/n da/2014/125469Orig1s000SumR.pdf Accessed 2015 March 2. 342. U.S. Food and Drug Administration. Janumet (sitagliptin/metformin HCl) tablets. 2013; http://www.fda.gov/Safety/MedWatch/SafetyInfo rmation/ucm196610.htm. Accessed 2015 February 25. 343. Idris I, Warren G, Donnelly R. Association between thiazolidinedione treatment and risk of macular edema among patients with type 2 diabetes. Arch Intern Med. 2012 Jul 9;172(13):1005-11. PMID: 22688528. 344. Silva PS, Cavallerano JD, Sun JK, et al. Effect of systemic medications on onset and progression of diabetic retinopathy. Nature Reviews Endocrinology. 2010;6(9):494-507. PMID: 20664533 345. Singh S, Segal JB. Thiazolidinediones and macular edema. Archives of Internal Medicine. 2012;172(13):1011-3. PMID: 22688825 346. Kawalec P, Mikrut A, Lopuch S. The safety of dipeptidyl peptidase-4 (DPP-4) inhibitors or sodium-glucose cotransporter 2 (SGLT-2) inhibitors added to metformin background therapy in patients with type 2 diabetes mellitus: a systematic review and meta- analysis. Diabetes Metab Res Rev. 2014 May;30(4):269-83. PMID: 24829965. 347. U.S. Food and Drug Administration. FDA Drug Safety Communication: FDA revised label of diabetes drug canagliflozin (Invokana, Invokamet) to include updates on bone fracture risk and new information on decreased bone mineral density. 2015; http://www.fda.gov/Drugs/DrugSafety/ucm46144 9.htm. Accessed 2015 September 16. 348. U.S. Food and Drug Administration. Highlights of Prescribing Information. Invokana (canagliflozin) tablets, for oral use. 2015; http://www.accessdata.fda.gov/drugsatfda_docs/l abel/2015/204042s006lbl.pdf. Accessed 2015 September 16. 349. Erondu N, Desai M, Ways K, Meininger G. Diabetic Ketoacidosis and Related Events in the Canagliflozin Type 2 Diabetes Clinical Program. Diabetes Care. 2015 Jul 22. PMID: 26203064. 350. United States Renal Data System. 2012 Atlas of CKD & ESRD. 2012; http://www.usrds.org/atlas12.aspx. Accessed 2015 March 2. 351. Centers for Disease Control and Prevention. Diabetic Retinopathy. 2015; http://www.cdc.gov/visionhealth/pdf/factsheet.pd f Accessed 2015 March 2. 352. Centers for Disease Control and Prevention. New Diabetes Atlas. 2014; http://www.cdc.gov/diabetes/data/. Accessed 2015 February 27. 353. The Henry J. Kaiser Family Foundation. Number of Diabetes Deaths per 100,000 Population by Race/Ethnicity. 2015; http://kff.org/other/state-indicator/diabetes-death- rate-by-raceethnicity/ Accessed 2015 February 27. 354. Selvin E, Parrinello CM, Sacks DB, Coresh J. Trends in prevalence and control of diabetes in the United States, 1988-1994 and 1999-2010. Ann Intern Med. 2014 Apr 15;160(8):517-25. PMID: 24733192.
- 406. 349 355. Wong HK,Ong KL, Cheung CL, Cheung BM. Utilization of glucose, blood pressure, and lipid lowering medications among people with type II diabetes in the United States, 1999-2010. Ann Epidemiol. 2014 Jul;24(7):516-21 e1. PMID: 24935464. 356. Qaseem A, Humphrey LL, Sweet DE, et al. Oral pharmacologic treatment of type 2 diabetes mellitus: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2012 Feb 7;156(3):218-31. PMID: 22312141. 357. Bristol-Myers Squibb. GLUCOPHAGE® (metformin hydrochloride) Tablets. GLUCOPHAGE® XR (metformin hydrochloride) Extended-Release Tablets. http://packageinserts.bms.com/pi/pi_glucophage_ xr.pdf. Accessed 2015 July 30, . 358. Bailey RA, Wang Y, Zhu V, Rupnow MF. Chronic kidney disease in US adults with type 2 diabetes: an updated national estimate of prevalence based on Kidney Disease: Improving Global Outcomes (KDIGO) staging. BMC Res Notes. 2014;7:415. PMID: 24990184. 359. Turner R, Murchison L, Wright AD, et al. United Kingdom prospective diabetes study 24: A 6-year, randomized, controlled trial comparing sulfonylurea, insulin, and metformin therapy in patients with newly diagnosed type 2 diabetes that could not be controlled with diet therapy. Ann Intern Med. 1998 1998;128(3):165-75. PMID: 9454524. 360. U.K. prospective diabetes study. II. Reduction in HbA1c with basal insulin supplement, sulfonylurea, or biguanide therapy in maturity-onset diabetes. A multicenter study. Diabetes. 1985 Aug;34(8):793-8. PMID: 2862087. 361. United Kingdom Prospective Diabetes Study (UKPDS). 13: Relative efficacy of randomly allocated diet, sulphonylurea, insulin, or metformin in patients with newly diagnosed non-insulin dependent diabetes followed for three years. BMJ. 1995 Jan 14;310(6972):83-8. PMID: 7833731. 362. Patel A, MacMahon S, Chalmers J, et al. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. The New England journal of medicine. 2008 Jun 12;358(24):2560-72. PMID: 18539916. 363. Nissen SE, Nicholls SJ, Wolski K, et al. Comparison of pioglitazone vs glimepiride on progression of coronary atherosclerosis in patients with type 2 diabetes: the PERISCOPE randomized controlled trial. JAMA. 2008 Apr 2;299(13):1561-73. PMID: 18378631. 364. Buse JB, Henry RR, Han J, et al. Effects of exenatide (exendin-4) on glycemic control over 30 weeks in sulfonylurea-treated patients with type 2 diabetes. Diabetes Care. 2004 Nov;27(11):2628-35. PMID: 15504997. 365. Blonde L, Klein EJ, Han J, et al. Interim analysis of the effects of exenatide treatment on A1C, weight and cardiovascular risk factors over 82 weeks in 314 overweight patients with type 2 diabetes. Diabetes Obes Metab. 2006 Jul;8(4):436-47. PMID: 16776751. 366. Zinman B, Hoogwerf BJ, Duran Garcia S, et al. The effect of adding exenatide to a thiazolidinedione in suboptimally controlled type 2 diabetes: a randomized trial. Ann Intern Med. 2007 Apr 3;146(7):477-85. PMID: 17404349. 367. Buse JB, Rosenstock J, Sesti G, et al. Liraglutide once a day versus exenatide twice a day for type 2 diabetes: a 26-week randomised, parallel-group, multinational, open-label trial (LEAD-6). Lancet. 2009 Jul 4;374(9683):39-47. PMID: 19515413. 368. Marre M, Shaw J, Brandle M, et al. Liraglutide, a once-daily human GLP-1 analogue, added to a sulphonylurea over 26 weeks produces greater improvements in glycaemic and weight control compared with adding rosiglitazone or placebo in subjects with Type 2 diabetes (LEAD-1 SU). Diabet Med. 2009 Mar;26(3):268-78. PMID: 19317822. 369. Zinman B, Gerich J, Buse JB, et al. Efficacy and safety of the human glucagon-like peptide-1 analog liraglutide in combination with metformin and thiazolidinedione in patients with type 2 diabetes (LEAD-4 Met+TZD). Diabetes Care. 2009 Jul;32(7):1224-30. PMID: 19289857.
- 407. 350 370. Li T,Hutfless S, Scharfstein DO, et al. Standards should be applied in the prevention and handling of missing data for patient-centered outcomes research: a systematic review and expert consensus. J Clin Epidemiol. 2014 Jan;67(1):15-32. PMID: 24262770. 371. Patorno E, Patrick AR, Garry EM, et al. Observational studies of the association between glucose-lowering medications and cardiovascular outcomes: addressing methodological limitations. Diabetologia. 2014 Nov;57(11):2237-50. PMID: 25212258.
- 408.
- 409. 352 Abbreviations ADOPT = ADiabetes Outcome Progression Trial AHRQ = Agency for Healthcare Research and Quality BMI = body mass index CARMELINA = Cardiovascular and Renal Microvascular Outcome Study with Linagliptin in Patients with Type 2 Diabetes Mellitus CAROLINA = Cardiovascular Outcome Study of Linagliptin Versus Glimepiride in Patients with Type 2 Diabetes CER = comparative effectiveness reviews CI = confidence interval CVD = cardiovascular disease DPP-4 inhibitors = dipeptidyl-peptidase-4 inhibitors EHC = Effective Health Care EXAMINE = Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care; FDA = Food and Drug Administration GI = gastrointestinal GLP-1 receptor agonists = glucagon-like peptide-1 receptor agonists HbA1c = hemoglobin A1c HR = hazard ratio LEAD = Liraglutide Effect and Action in Diabetes MeSH = medical subject headings OR = odds ratio PROactive = PROspective pioglitAzone Clinical Trial In macroVascular Events RCT = randomized controlled trial RECORD = Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycemia in Diabetes RR = risk ratio SAVOR-TIMI = Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus Thrombolysis in Myocardial Infarction SGLT-2 inhibitors = sodium glucose co-transporter 2 inhibitors SOE = Strength of evidence TECOS = Trial Evaluating Cardiovascular Outcomes with Sitagliptin TEP = technical expert panel UKPDS = United Kingdom Prospective Diabetes Study UTI = urinary tract infection
- 410. A-1 Appendix A. DetailedElectronic Database Search Strategies PubMed Strategy Search String #1 (“diabetes mellitus, type 2”[mh] or (diabet*[tiab] and (“non-insulin dependent”[tiab] or type-2[tiab] or “type II”[tiab] or “type 2”[tiab]))) AND (“metformin”[mh] or “thiazolidinediones”[mh] or “glipizide”[mh] or “glyburide”[mh] or “Dipeptidyl-Peptidase IV Inhibitors”[mh] or “Glucagon-Like Peptide 1”[mh] or biguanide*[tiab] or metformin[tiab] or thiazolidinedione*[tiab] or pioglitazone[tiab] or rosiglitazone[tiab] or sulfonylurea*[tiab] or sulphonylurea*[tiab] or glipizide[tiab] or glyburide[tiab] or glimepiride[tiab] or glibenclamide[tiab] or “insulin secretagogues”[tiab] or sitagliptin*[tiab] or saxagliptin*[tiab] or dpp-4[tiab] or dpp-iv[tiab] or liraglutide[tiab] or exenatide[tiab]) NOT (animal[mh] NOT human[mh]) NOT (letter[pt] or comment[pt] or editorial[pt]) AND (("2009/04/01"[edat] : "2014/07/11"[edat])) #2 (“diabetes mellitus, type 2”[mh] or (diabet*[tiab] and (“non-insulin dependent”[tiab] or type-2[tiab] or “type II”[tiab] or “type 2”[tiab]))) AND (linagliptin*[tiab] or alogliptin*[tiab] or albiglutide*[tiab] or dulaglutide*[tiab] or "sodium-glucose co-transporter 2 inhibitors”[tiab] or “sodium-glucose co- transporter 2 inhibitor” [tiab] or “SGLT-2” [tiab] or “canagliflozin”[tiab] or “dapagliflozin”[tiab]) NOT (animal[mh] NOT human[mh]) NOT (letter[pt] or comment[pt] or editorial[pt]) #3 (“diabetes mellitus, type 2”[mh] or (diabet*[tiab] and (“non-insulin dependent”[tiab] or type-2[tiab] or “type II”[tiab] or “type 2”[tiab]))) AND (empagliflozin*[tiab]) NOT (animal[mh] NOT human[mh]) NOT (letter[pt] or comment[pt] or editorial[pt])
- 411. A-2 EMBASE Strategy Search String #1('non insulin dependent diabetes mellitus'/exp OR 'non insulin dependent diabetes mellitus' or (diabet*:ti,ab and (‘non-insulin dependent’:ti,ab or type- 2:ti,ab or ‘type II’:ti,ab or ‘type 2’:ti,ab))) AND ('thiazolidinedione'/exp or 'rosiglitazone'/exp or 'pioglitazone'/exp or 'glipizide'/exp or 'glyburide'/exp or ‘glimepiride’/exp or 'metformin'/exp or ‘sitagliptin’/exp or thiazolidinedione*:ti,ab or pioglitazone:ti,ab or rosiglitazone:ti,ab or sulfonylurea*:ti,ab or sulphonylurea*:ti,ab or glipizide:ti,ab or glyburide:ti,ab or glimepiride:ti,ab or glibenclamide:ti,ab or biguanide*:ti,ab or metformin:ti,ab or ‘insulin secretagogues’:ti,ab or ‘Dipeptidyl-Peptidase IV Inhibitor’/de or saxagliptin/exp or saxagliptin*:ti,ab or sitagliptin/exp or sitagliptin*:ti,ab or dpp-4:ti,ab or dpp-iv:ti,ab or exenatide/exp or exenatide:ti,ab or liraglutide/exp or liraglutide:ti,ab) NOT ([animals]/lim NOT [humans]/lim) NOT (letter:it or comment:it or editorial:it) AND [2009-2014]/py #2 ('non insulin dependent diabetes mellitus'/exp OR 'non insulin dependent diabetes mellitus' or (diabet*:ti,ab and (‘non-insulin dependent’:ti,ab or type- 2:ti,ab or ‘type II’:ti,ab or ‘type 2’:ti,ab))) AND (linagliptin/exp or linagliptin*:ti,ab or alogliptin/exp or alogliptin*:ti,ab or albiglutide/exp or albiglutide*:ti,ab or dulaglutide/exp or dulaglutide*:ti,ab or ‘sodium glucose cotransporter 2 inhibitor’/de or ‘sodium-glucose co-transporter 2 inhibitors’:ti,ab or ‘sodium-glucose co-transporter 2 inhibitor’:ti,ab or ‘sodium glucose cotransporter 2 inhibitors’:ti,ab or ‘sodium glucose cotransporter 2 inhibitor’:ti,ab or ‘SGLT-2”:ti,ab or canagliflozin/exp or canagliflozin:ti,ab or dapagliflozin/exp or dapagliflozin:ti,ab) NOT ([animals]/lim NOT [humans]/lim) NOT (letter:it or comment:it or editorial:it) #3 ('non insulin dependent diabetes mellitus'/exp OR 'non insulin dependent diabetes mellitus' or (diabet*:ti,ab and (‘non-insulin dependent’:ti,ab or type- 2:ti,ab or ‘type II’:ti,ab or ‘type 2’:ti,ab))) AND (empagliflozin/exp or empagliflozin*:ti,ab) NOT ([animals]/lim NOT [humans]/lim) NOT (letter:it or comment:it or editorial:it)
- 412. A-3 Cochrane Strategy Search String #1((diabetes near type-2):ti,ab,kw or (diabet*:ti,ab,kw and (“non-insulin dependent”:ti,ab,kw or type-2:ti,ab,kw or “type II”:ti,ab,kw or “type 2”:ti,ab,kw))) AND (thiazolidinedione*:ti,ab,kw or pioglitazone:ti,ab,kw or rosiglitazone:ti,ab,kw or sulfonylurea*:ti,ab,kw or sulphonylurea*:ti,ab,kw or glipizide:ti,ab,kw or glyburide:ti,ab,kw or glimepiride:ti,ab,kw or glibenclamide:ti,ab,kw or biguanide*:ti,ab,kw or metformin:ti,ab,kw or “insulin secretagogues”:ti,ab,kw or “Dipeptidyl-Peptidase IV Inhibitors”:ti,ab,kw or saxagliptin*:ti,ab,kw or sitagliptin*:ti,ab,kw or liraglutide:ti,ab,kw or exenatide:ti,ab,kw) Publication Year from 2009 to 2014 #2 ((diabetes near type-2):ti,ab,kw or (diabet*:ti,ab,kw and (“non-insulin dependent”:ti,ab,kw or type-2:ti,ab,kw or “type II”:ti,ab,kw or “type 2”:ti,ab,kw))) AND (linagliptin*:ti,ab,kw or alogliptin*:ti,ab,kw or albiglutide*:ti,ab,kw or dulaglutide*:ti,ab,kw or ‘sodium-glucose co-transporter 2 inhibitors’:ti,ab,kw or ‘sodium-glucose co-transporter 2 inhibitor’:ti,ab,kw or ‘sodium glucose cotransporter 2 inhibitors’:ti,ab,kw or ‘sodium glucose cotransporter 2 inhibitor’:ti,ab,kw or ‘SGLT-2’:ti,ab,kw or canagliflozin:ti,ab,kw or dapagliflozin:ti,ab,kw) #3 ((diabetes near type-2):ti,ab,kw or (diabet*:ti,ab,kw and (“non-insulin dependent”:ti,ab,kw or type-2:ti,ab,kw or “type II”:ti,ab,kw or “type 2”:ti,ab,kw))) AND (empagliflozin*:ti,ab,kw)
- 413.
- 414.
- 415.
- 416.
- 417.
- 418.
- 419.
- 420.
- 421.
- 422.
- 423.
- 424.
- 425.
- 426.
- 427.
- 428.
- 429.
- 430. B-18 Study Quality Formfor Randomized Controlled Trials
- 431. B-19 Study Quality Formfor Nonrandomized Study
- 432.
- 433. C-1 Appendix C. Listof Excluded Studies . Insulin vs. Sulfonylureas as add-on to metformin. Drug and Therapeutics Bulletin. 2014;52(10):112-3. Meeting abstract . Risk of acute pancreatitis with 'gliptins'. Drug Ther Bull. 2012;50(12):134. No original data .A therapeutic option for the management of type 2 diabetes. 2013. No original data . Correction To Comparison Of Empaglifl Ozin And Glimepiride As Add-On To Metformin In Patients With Type 2 Diabetes: A 104-Week Randomised, Active- Controlled, Double-Blind, Phase 3 Trial Lancet Diabetes Endocrinol 2014; 2: 691- 700. The Lancet Diabetes and Endocrinology. 2015;3(3):e2. Erratum; No original data . ERRATUM: Valentine V, Hinnen D. Clinical Implications of Canagliflozin Treatment in Patients With Type 2 Diabetes. Clinical Diabetes 2014;33: 5-13 (DOI: 10.2337/diaclin.33.2.96). Clin Diabetes. 2015 Apr;33(2):96. PMID: 25896636. No original data Aaboe K, Knop FK, Vilsboll T, et al. Twelve weeks treatment with the DPP-4 inhibitor sitagliptin improves glycaemic control, but does not improve GLP-1 secretion, in patients with type 2 diabetes - A randomised trial. Diabetologia. 2009;52(S1):S294. No original data Aaboe K, Knop FK, Vilsboll T, et al. Twelve weeks treatment with the DPP-4 inhibitor, Sitagliptin, Reduces total PYY and PYY3-36 and increases PYY1-36 but has no effect on intact GLP-2 in subjects with type 2 diabetes Mellitus-A randomized trial. Diabetes. 2009;58((Aaboe K.; Knop F.K.; Vilsboll T.; Deacon C.F.; Holst J.J.; Madsbad S.; Krarup T.)). Meeting abstract Aaboe K, Vilsboll T, Knop FK, et al. Twelve weeks treatment with the DPP-4 Inhibitor, Sitagliptin, Improves the Insulin- Secreting capacity of the P-Cells in subjects with type 2 diabetes Mellitus-A randomized trial. Diabetes. 2009;58((Aaboe K.; Vilsboll T.; Knop F.K.; Deacon C.F.; Holst J.J.; Madsbad S.; Krarup T.)). Meeting abstract Abbatecola AM, Lattanzio F, Molinari AM, et al. Rosiglitazone and cognitive stability in older individuals with type 2 diabetes and mild cognitive impairment. Diabetes Care. 2010 Aug;33(8):1706-11. PMID: 20435794. No outcome of interest; Does not meet study design criteria Abbatecola AM, Paolisso G. Rosiglitazone and cognitive stability in older persons with type 2 diabetes and mild cognitive impairment. Diabetologia. 2009;52(S1):S67. Meeting abstract Abdulkadir AA, Thanoon IA. Comparative Effects of Glibenclamide and Metformin on C-Reactive Protein and Oxidant/Antioxidant Status in Patients with Type II Diabetes Mellitus. Sultan Qaboos Univ Med J. 2012 Feb;12(1):55-61. PMID: 22375259. Followup less than 3 months Abe M, Okada K, Maruyama T, et al. Clinical effectiveness and safety evaluation of long-term pioglitazone treatment for erythropoietin responsiveness and insulin resistance in type 2 diabetic patients on
- 434. C-2 hemodialysis. Expert OpinPharmacother. 2010 Jul;11(10):1611-20. PMID: 20540652. Background medications Adetunji O, Skrivanek Z, Tahbaz A, et al. A post-hoc pooled analysis of two placebo controlled phase 3 trials, Assessment of Weekly AdministRation of LY2189265 in Diabetes-1 and-5 (AWARD-1 and AWARD-5): Dulaglutide compared with exenatide, sitagliptin, and placebo. Diabetologie und Stoffwechsel. 2014;9((Adetunji O.; Tahbaz A.) Eli Lilly and Company, Medical Affairs, Basingstoke, United Kingdom). Meeting abstract Adetunji O, Skrivanek Z, Tahbaz A, et al. A posthoc pooled analysis of two placebo- controlled phase 3 trials, Assessment of Weekly Administration of LY2189265 in Diabetes 1 and 5 (AWARD-1 and AWARD- 5): Dulaglutide compared with exenatide, sitagliptin and placebo. Diabetic Medicine. 2014;31((Adetunji O.; Tahbaz A.) Medical Department, Eli Lilly and Company, Basingstoke, United Kingdom):50-1. Meeting abstract Agarwala A, Givens E, McGuire DK, et al. Rosiglitazone increases cholesterol efflux capacity in patients with type 2 diabetes. Journal of Investigative Medicine. 2014;62(2):510-1. Meeting abstract Agrawal A, Pradeep. To study the pattern of use and efficacy of anti-diabetic drugs in controlling adequate glycemic levels in diabetic patients in Navi Mumbai. Australasian Medical Journal. 2012;5(1):88- 9. Meeting abstract Ajdi F, Khabbal Y, Safi S. ADR of oral antidiabetic. Drug Safety. 2009;32(10):949- 50. Meeting abstract Al Sifri S, Basiounny A, Echtay A, et al. The incidence of hypoglycaemia in Muslim patients with type 2 diabetes treated with sitagliptin or a sulphonylurea during Ramadan: a randomised trial. Int J Clin Pract. 2011 Nov;65(11):1132-40. PMID: 21951832. No drug comparison of interest Alba M, Ahren B, Inzucchi SE, et al. Initial combination therapy with sitagliptin and pioglitazone: Complementary effects on postprandial glucose and islet cell function. Canadian Journal of Diabetes. 2009;33(3):319-20. Meeting abstract Alexanderson-Rosas E, de Jesus Martinez A, Ochoa-Lopez JM, et al. [Effects of the combined treatment with Metformin/Glimepiride on endothelial function of patients with type 2 diabetes mellitus. A positron emission tomography (PET) evaluation study]. Arch Cardiol Mex. 2009 Oct-Dec;79(4):249-56. PMID: 20191984. Follow-up less than three months Alkharfy KM, Al-Daghri NM, Sabico SB, et al. Vitamin D supplementation in patients with diabetes mellitus type 2 on different therapeutic regimens: a one-year prospective study. Cardiovasc Diabetol. 2013 Aug 7;12(1):113. PMID: 23924389. No drug comparison of interest Allen E, Berglind N. Saxagliptin vs glipizide as add-on therapy to metformin in patients with type 2 diabetes: A 2-year assessment of HbA1c, hypoglycaemia, and weight gain in a randomised, double-blind study.
- 435. C-3 Diabetologia. 2011;54((Allen E.;Berglind N.) Bristol-Myers Squibb, Princeton, United States):S337. Meeting abstract Allen E, Donovan M, Berglind N, et al. Efficacy of saxagliptin according to patient baseline characteristics: A pooled analysis of three add-on pivotal randomised phase 3 clinical trials. Diabetologia. 2010;53((Allen E.; Donovan M.; Berglind N.) Bristol-Myers Squibb, Princeton, United States):S328. Meeting abstract Allen E, Karyekar C, Ohman P. Safety profile of saxagliptin (SAXA) in combination with 2 other agents: Data from dual-therapy trials in patients receiving rescue treatment. Diabetes. 2011;60((Allen E.; Karyekar C.; Ohman P.) Princeton, United States):A619-A20. Meeting abstract Allen E, Slater J, Bryzinski B, et al. Efficacy and safety of saxagliptin (SAXA) in patients with type 2 diabetes stratified by cardiovascular risk factors. Diabetologia. 2012;55((Allen E.; Slater J.) Medical Affairs, Bristol-Myers Squibb, Princeton, United States):S346-S7. Meeting abstract Alvarez-Guisasola F, Yin DD, Nocea G, et al. Association of hypoglycemic symptoms with patients' rating of their health-related quality of life state: a cross sectional study. Health Qual Life Outcomes. 2010;8:86. PMID: 20723229. Does not apply; No drug comparison of interest Ambrosius WT, Danis RP, Goff DC, Jr., et al. Lack of association between thiazolidinediones and macular edema in type 2 diabetes: the ACCORD eye substudy. Arch Ophthalmol. 2010 Mar;128(3):312-8. PMID: 20212201. Background medications Ametov AS, Gusenbekova DG. [Effect of dipeptidyl peptidase-4 inhibitors on lipid metabolism in patients with type 2 diabetes mellitus]. Ter Arkh. 2014;86(8):85-9. PMID: 25306750. Non-English Language Araki A, Iimuro S, Sakurai T, et al. Long- term multiple risk factor interventions in Japanese elderly diabetic patients: The Japanese Elderly Diabetes Intervention Trial - study design, baseline characteristics and effects of intervention. Geriatrics and Gerontology International. 2012;12(SUPPL.1):7-17. Does not apply Aravind SR, Ismail SB, Balamurugan R, et al. Hypoglycemia in patients with type 2 diabetes from India and Malaysia treated with sitagliptin or a sulfonylurea during Ramadan: a randomized, pragmatic study. Curr Med Res Opin. 2012 Aug;28(8):1289- 96. PMID: 22738801. Follow-up less than three months; Background medications Arcidiacono B, Capula C, Chiefari E, et al. Glycemic efficacy of liraglutide is linked to gender in italian type 2 diabetic patients. Diabetes. 2014;63((Arcidiacono B.; Capula C.; Chiefari E.; Vero A.; Oliverio R.; Puccio L.; Liguori R.; Pullano V.; Tirinato D.; Foti D.; Vero R.; Brunetti A.) Catanzaro, Italy, Soverato, Italy):A288. Meeting abstract Ardawi MS, Akbar D, Al-Shaik A, et al. Circulating sclerostin, bone turnover markers and BMD in type-2 diabetic women treated with metformin or pioglitazone. Journal of Bone and Mineral Research.
- 436. C-4 2013;28((Ardawi M.-S.; RouziA.) Center of Excellence for Osteoporosis Research, Faculty of Medicine, Saudi Arabia). Meeting abstract Ardawi MS, Akbar D, Alshaikh A, et al. Circulating sclerostin, bone turnover markers and BMD in type 2 diabetic women treated with metformin or pioglitazone. Osteoporosis International. 2013;24(1):S132-S3. Meeting abstract Arjona Ferreira JC, Corry D, Mogensen CE, et al. Efficacy and safety of sitagliptin in patients with type 2 diabetes and ESRD receiving dialysis: a 54-week randomized trial. Am J Kidney Dis. 2013 Apr;61(4):579- 87. PMID: 23352379. Comorbidity Arjona Ferreira JC, Corry D, Mogensen CE, et al. Efficacy and safety of sitagliptin vs. glipizide in patients with type 2 diabetes mellitus and end-stage renal disease on dialysis: A 54-week randomised trial. Diabetes, Stoffwechsel und Herz. 2011;20(6):430. Meeting abstract Arjona Ferreira JC, Marre M, Rabelink TJ, et al. Efficacy and safety of sitagliptin versus glipizide in patients with type 2 diabetes and moderate to severe chronic renal insufficiency. Diabetes, Stoffwechsel und Herz. 2011;20(6):419. Meeting abstract Armstrong M, Falahati A, Houlihan DD, et al. Effects of two years of liraglutide treatment on fatty liver disease in patients with type 2 diabetes: Analysis of the liraglutide effect and action in diabetes-2 extension trial. Gut. 2010;59((Armstrong M.; Falahati A.; Elbrand B.; Schmidt W.E.; Gough S.; Newsome P.N.) Centre for Liver Research, University of Birmingham, United Kingdom):A1-A2. Meeting abstract Armstrong M, Houlihan D, Schmidt W, et al. Effects of once-daily liraglutide on fatty liver disease in patients with type 2 diabetes (T2D) after 2 years' treatment: Retrospective-analysis of the lead-2 extension trial. Journal of Diabetes. 2011;3((Armstrong M.; Houlihan D.; Newsome P.) Centre for Liver Research, University of Birmingham, Birmingham, United Kingdom):11. Meeting abstract Armstrong MJ, Falahati A, Houlihan D, et al. Effects of two years of liraglutide treatment on fatty liver disease in patients with type 2 diabetes: Analysis of the lead-2 extension trial. Hepatology. 2010;52((Armstrong M.J.; Houlihan D.; Newsome P.N.) Centre for Liver Research, University of Birmingham, Birmingham, United Kingdom):620A. Meeting abstract Armstrong MJ, Houlihan DD, Rowe IA, et al. Safety and efficacy of liraglutide in patients with type 2 diabetes and elevated liver enzymes: individual patient data meta- analysis of the LEAD program. Aliment Pharmacol Ther. 2013 Jan;37(2):234-42. PMID: 23163663. Handsearch Armstrong MJ, Houlihan DD, Rowe IA, et al. Safety and efficacy of liraglutide in patients with type 2 diabetes with elevated liver enzymes: Individual patient data meta- analysis of the LEAD programme. The Lancet. 2013;381((Armstrong M.J., [email protected]; Houlihan D.D.; Rowe I.A.; Tomlinson J.W.) Centre for Liver Research, Institute of Biomedical Research, University of Birmingham,
- 437. C-5 Edgbaston, Birmingham, United Kingdom):S20. Meetingabstract Arnolds S, Sawicki PT. Liraglutide and the preservation of pancreatic (beta)-cell function in early type 2 diabetes: The libra trial. Diabetes care 2014;37:3270-3278. Diabetes Care. 2015;38(2):e25. Meeting abstract Arulanandham A, Raju A, Pradeep Rajkumar LA, et al. Prevalence of clinically significant macular edema [CSME] among glitazone users and non- users of type-2 DM patients with diabetic retinopathy. International Journal of Drug Development and Research. 2012;4(2):132-7. No drug comparison of interest Asanuma H, Kitakaze M. [Prospective pioglitazone clinical trial in macrovascular events]. Nihon Rinsho. 2012 May;70 Suppl 3:301-8. PMID: 22768537. No original data; No drug comparison of interest Aschner P, Sethi B, Gomez-Peralta F, et al. Glargine vs. premixed insulin for management of type 2 diabetes patients failing oral antidiabetic drugs: The GALAPAGOS study. Diabetes. 2013;62((Aschner P.; Sethi B.; Gomez- Peralta F.; Landgraf W.; Dain M.-P.; Pilorget V.; Comlekci A.) Bogota, Colombia, Hyderabad, India, Segovia, Spain, Frankfurt, Germany, Paris, France, Izmir, Turkey):A241-A2. Meeting abstract Aschner P, Sethi B, Gomez-Peralta F, et al. Insulin glargine compared with premixed insulin for management of insulin-naive type 2 diabetes patients uncontrolled on oral antidiabetic drugs: the open-label, randomized GALAPAGOS study. J Diabetes Complications. 2015 Aug;29(6):838-45. PMID: 25981123. Background medications Aso Y, Takebayashi K, Inukai T, et al. Pioglitazone and cardiovascular events in type 2 diabetes: Effects of pioglitazone on cardiovascular outcomes in Japanese patients with type 2 diabetes in higashi- saitama (EPOCH Trial). Diabetes. 2011;60((Aso Y.; Takebayashi K.; Inukai T.; Katsumori K.; Owada K.; Nakamura T.; Naito T.; Itabashi H.; Morita K.; Sekine M.; Takahashi K.; Miyano H.; Takai T.) Koshigaya, Japan):A557. Meeting abstract Atchison L, Steinke EL. Relationship between social and economic factors in diabetes medication-prescribing patterns. Journal of the American Pharmacists Association. 2011;51(2):228. Meeting abstract Aubert RE, Herrera V, Chen W, et al. Rosiglitazone and pioglitazone increase fracture risk in women and men with type 2 diabetes. Diabetes Obes Metab. 2010 Aug;12(8):716-21. PMID: 20590749. Aydin Y, Erden M, Ermis F, et al. Oral antidiabetics and insulins do not increase cancer risk. Acta Medica Mediterranea. 2013;29(4):859-67. Background medications Azar S, El-Mollayess GM, Al Shaar L, et al. Impact of thiazolidinediones on macular thickness and volume in diabetic eyes. Can J Ophthalmol. 2013 Aug;48(4):312-6. PMID: 23931472. No drug comparison of interest; Does not account for confounding Azar ST, Malha LP, Zantout MS, et al. Management and control of patients with type 2 diabetes mellitus in Lebanon: results
- 438. C-6 from the InternationalDiabetes Management Practices Study (IDMPS). J Med Liban. 2013 Jul-Sep;61(3):127-31. PMID: 24422361. Background medications Azoulay L, Dell'Aniello S, Gagnon B, et al. Metformin and the incidence of prostate cancer in patients with type 2 diabetes. Cancer Epidemiol Biomarkers Prev. 2011 Feb;20(2):337-44. PMID: 21148757. No drug comparison of interest Azoulay L, Schneider-Lindner V, Dell'aniello S, et al. Combination therapy with sulfonylureas and metformin and the prevention of death in type 2 diabetes: a nested case-control study. Pharmacoepidemiol Drug Saf. 2010 Apr;19(4):335-42. PMID: 20052677. Background medications Azoulay L, Schneider-Lindner V, Dell'aniello S, et al. Thiazolidinediones and the risk of incident strokes in patients with type 2 diabetes: a nested case-control study. Pharmacoepidemiol Drug Saf. 2010 Apr;19(4):343-50. PMID: 19998318. No drug comparison of interest Azoulay L, Yin H, Filion KB, et al. The use of pioglitazone and the risk of bladder cancer in patients with type 2 diabetes. Pharmacoepidemiology and Drug Safety. 2012;21((Azoulay L.; Yin H.; Filion K.B.; Assayag J.; Suissa S.) Centre for Clinical Epidemiolog, Jewish General Hospital, Montreal, Canada):271. Meeting abstract Azoulay L, Yin H, Filion KB, et al. The use of pioglitazone and the risk of bladder cancer in people with type 2 diabetes: nested case-control study. BMJ. 2012;344:e3645. PMID: 22653981 Does not account for confounding Bach RG, Brooks MM, Lombardero M, et al. Rosiglitazone and outcomes for patients with diabetes mellitus and coronary artery disease in the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial. Circulation. 2013 Aug 20;128(8):785-94. PMID: 23857320. No drug comparison of interest Bailey CJ, Day C, Campbell IW, et al. Glycaemic control and cardiovascular outcome trials in type 2 diabetes. British Journal of Diabetes and Vascular Disease. 2012;12(4):161-4. No original data Bailey CJ, Gross JL, Bastone L, et al. Dapagliflozin as an add-on to metformin lowers hyperglycaemia in type 2 diabetes patients inadequately controlled with metformin alone. Diabetologia. 2009;52(S1):S76. Meeting abstract Bailey CJ, Gross JL, Hennicken D, et al. Correction to Dapagliflozin add-on to metformin in type 2 diabetes inadequately controlled with metformin: A randomized, double-blind, placebo-controlled 102-week trial [BMC Medicine, 11, 193, (2013)]. BMC Med. 2013;11(1). Meeting abstract Bailey CJ, Gross JL, Yadav M, et al. Sustained efficacy of dapagliflozin when added to metformin in type 2 diabetes inadequately controlled by metformin monotherapy. Diabetologia. 2011;54((Bailey C.J.) Aston University, Birmingham, United Kingdom):S67. Meeting abstract Bailey CJ, Iqbal N, T'Joen C, et al. Dapagliflozin monotherapy in drug-naive patients with diabetes: a randomized-
- 439. C-7 controlled trial oflow-dose range. Diabetes Obes Metab. 2012 Oct;14(10):951-9. PMID: 22776824. No drug comparison of interest Bailey CJ, Morales Villegas EC, Woo V, et al. Efficacy and safety of dapagliflozin monotherapy in people with Type 2 diabetes: a randomized double-blind placebo-controlled 102-week trial. Diabet Med. 2014 Nov 8PMID: 25381876. No drug comparison of interest Bailey CJ, Wilding J, Nauck MA, et al. Sustained reductions in weight and HbA1c with dapagliflozin: Long-term results from phase III clinical studies in type 2 diabetes. Diabetologia. 2012;55((Bailey C.J.) Aston University, School of Life and Health Sciences, Birmingham, United Kingdom):S295. Meeting abstract Bailey RA, Damaraju CV, Martin SC, et al. Attainment of diabetes-related quality measures with canagliflozin versus sitagliptin. Am J Manag Care. 2014 Jan;20(1 Suppl):s16-24. PMID: 24512193. No drug comparison of interest Bailey RA, Vijapurkar U, Martin S, et al. Composite quality measure (CQM) attainment in overweight/obese patients with type 2 diabetes mellitus treated with canagliflozin 300 mg (CANA) or sitagliptin 100 mg (SITA). Value in Health. 2014;17(3):A239. Meeting abstract Bailey RA, Vijapurkar U, Meininger GE, et al. Diabetes-related quality measure attainment: canagliflozin versus sitagliptin based on a pooled analysis of 2 clinical trials. Am J Manag Care. 2014 Dec;20(13 Suppl):s296-305. PMID: 25734218. No drug comparison of interest Bailey T, Pratley R, Buse J, et al. Liraglutide produces greater reductions in HbA1c levels compared with sitagliptin or exenatide across five baseline HbA1c categories. Diabetologia. 2011;54((Bailey T.) AMCR Institute, Escondido, United States):S321. Meeting abstract Bain SC, Skrivanek Z, Tahbaz A, et al. Efficacy of long-acting once weekly dulaglutide compared with short-acting twice daily (bid) exenatide in patients with Type 2 diabetes: A posthoc analysis to determine the influence of baseline HbA1c in the Assessment of Weekly Administration of LY2189265 in Diabetes-1 (AWARD-1) trial. Diabetic Medicine. 2014;31((Bain S.C.) Institute of Life Science, Swansea University, Abertawe Bro Morgannwg University Health Board, Swansea, United Kingdom):50. Meeting abstract Bain SC, Stella P, Cao A. Significantly reduced body mass index with liraglutide 1.2 mg treatment versus glimepiride may have an impact on cardiovascular risk in patients with type 2 diabetes. Diabetic Medicine. 2010;27(2):79. Meeting abstract Balena R, Wintle M, Meloni A, et al. Exenatide once weekly, glycaemic goals, and selected cardiovascular risk factors in patients with T2DM: A retrospective analysis of pooled clinical trial data. Diabetes, Stoffwechsel und Herz. 2011;20(6):430-1. Meeting abstract Bannister CA, Holden SE, Jenkins-Jones S, et al. Can people with type 2 diabetes live longer than those without? A comparison of mortality in people initiated with metformin or sulphonylurea monotherapy and matched,
- 440. C-8 non-diabetic controls. DiabetesObes Metab. 2014 Nov;16(11):1165-73. PMID: 25041462. Does not account for confounding Barnett A, Huisman H, Jones R, et al. Efficacy and safety of linagliptin in elderly patients ((greater-than or equal to) 70 years) with type 2 diabetes. Diabetes. 2012;61((Barnett A.; Huisman H.; Jones R.; Von Eynatten M.; Patel S.; Woerle H.-J.) Birmingham, United Kingdom):A260-A1. Meeting abstract Barnett AH, Charbonnel B, Donovan M, et al. Effect of saxagliptin as add-on therapy in patients with poorly controlled type 2 diabetes on insulin alone or insulin combined with metformin. Curr Med Res Opin. 2012 Apr;28(4):513-23. PMID: 22313154. No drug comparison of interest Barnett AH, Harper R, Toorawa R, et al. Linagliptin monotherapy improves glycaemic control in type 2 diabetes patients for whom metformin therapy is inappropriate. Diabetologia. 2010;53((Barnett A.H.) University of Birmingham, Heart of England NHS Foundation Trust, United Kingdom):S327. Meeting abstract Barnett AH, Mithal A, Manassie J, et al. A phase III trial of empagliflozin in patients with Type 2 diabetes with stage 2 or 3 chronic kidney disease (EMPA-REG RENAL(trademark)). Diabetic Medicine. 2014;31((Barnett A.H.) Diabetes Centre, Heart of England NHS Foundation Trust, University of Birmingham, Birmingham, United Kingdom):63. Meeting abstract Barnett AH, Mithal A, Manassie J, et al. Efficacy and safety of empagliflozin added to existing antidiabetes treatment in patients with type 2 diabetes and chronic kidney disease: A randomised, double-blind, placebo-controlled trial. The Lancet Diabetes and Endocrinology. 2014;2(5):369- 84. Background medications; No drug comparison of interest Barnett AH, Mithal A, Manassie J, et al. Empagliflozin in patients with type 2 diabetes mellitus (T2DM) and renal impairment (RI). Diabetes. 2013;62((Barnett A.H.; Mithal A.; Manassie J.; Jones R.; Rattunde H.; Woerle H.J.; Broedl U.C.) Birmingham, United Kingdom, Delhi, India, Berkshire, United Kingdom, Ingelheim, Germany):A286. Meeting abstract Barnett AH, Tahrani AA, Von Eynatten M, et al. The novel DPP-4 inhibitor linagliptin is associated with a very low risk of hypoglycemia: Results from a large phase III program. Diabetes. 2011;60((Barnett A.H.; Tahrani A.A.; Von Eynatten M.; Emser A.; Patel S.; Woerle H.J.) Birmingham, United Kingdom):A623. Meeting abstract Barrington P, Chien JY, Showalter HDH, et al. A 5-week study of the pharmacokinetics and pharmacodynamics of LY2189265, a novel, long-acting glucagon-like peptide-1 analogue, in patients with type 2 diabetes. Diabetes, Obesity and Metabolism. 2011;13(5):426-33. Background medications; No drug comparison of interest Barthelemy M, Boullu Sanchis S, Moreau F, et al. Substitution of insulin by exenatide in bad controlled type 2 diabetic patients: efficacy and predictive factors. Minerva Endocrinol. 2014 Jul 8PMID: 25003223.
- 441. C-9 Does not apply;Does not meet study design criteria Barzilai N, Guo H, Mahoney EM, et al. Efficacy and tolerability of sitagliptin monotherapy in elderly patients with type 2 diabetes: a randomized, double-blind, placebo-controlled trial. Curr Med Res Opin. 2011 May;27(5):1049-58. PMID: 21428727. Placebo-controlled trial Baser O, Tangirala K, Wei W, et al. Real- world outcomes of initiating insulin glargine-based treatment versus premixed analog insulins among US patients with type 2 diabetes failing oral antidiabetic drugs. ClinicoEconomics and Outcomes Research. 2013;5(1):497-505. No drug comparison of interest Baser O, Wei W, Baser E, et al. Clinical and economic outcomes in patients with type 2 diabetes initiating insulin glargine disposable pen versus exenatide BID. J Med Econ. 2011;14(6):673-80. PMID: 21892858. No drug comparison of interest; Background medications Baser O, Wei W, Baser E. Clinical and economic outcomes in patients with type 2 diabetes initiating insulin glargine using disposable pen versus exenatide. Journal of Managed Care Pharmacy. 2011;17(3):237-8. Meeting abstract Basile J, Ptaszynska A, Ying L, et al. The effects of dapagliflozin on cardiovascular risk factors in patients with type 2 diabetes mellitus. Circulation: Cardiovascular Quality and Outcomes. 2012;5(3). Meeting abstract Bayraktar S, Hernadez-Aya LF, Lei X, et al. Effect of metformin on survival outcomes in diabetic patients with triple receptor- negative breast cancer. Cancer. 2012 Mar 1;118(5):1202-11. PMID: 21800293. No drug comparison of interest Bazelier M, Gallagher A, Vestergaard P, et al. Use of thiazolidinediones and risk of osteoporotic fracture: Disease or drugs? Osteoporosis International. 2012;23((Bazelier M.; Gallagher A.; De Vries F.) Utrecht University, Utrecht, Netherlands):S556. Meeting abstract Bazelier MT, Gallagher AM, van Staa TP, et al. Use of thiazolidinediones and risk of osteoporotic fracture: disease or drugs? Pharmacoepidemiol Drug Saf. 2012 May;21(5):507-14. PMID: 22392882. No drug comparison of interest Belhadj M, Dahaoui A, Jamoussi H, et al. Exploring insulin analogue safety and effectiveness in a Maghrebian cohort with type 2 diabetes: Results from the A1chieve study. Diabetes Research and Clinical Practice. 2013;101(SUPPL.1):S4-S14. Does not meet study design criteria Bell K, Hardy E, De Bruin T, et al. The effect of dapagliflozin on hedis performance measures of hba1c in patients with type 2 diabetes mellitus. Value in Health. 2012;15(4):A171-A2. Meeting abstract Bellary S. For type 2 diabetes poorly controlled by metformin monotherapy, the addition of any non-insulin antidiabetic drug reduces HbA1c to a similar extent, but with differing effects on weight and hypoglycaemic risk. Evidence-Based Medicine. 2011;16(2):39-40. No original data Bellows B, McAdam-Marx C, Unni S, et al. 12-month A1c and weight outcomes by drug
- 442. C-10 class in treatmentnaive patients with type 2 diabetes. Journal of Managed Care Pharmacy. 2012;18(2):157. Meeting abstract Bellows BK, Ye X, Unni S, et al. Impact of anti-diabetic drug selection on weight change and hba1c outcomes in treatment naive patients with type 2 diabetes. Diabetes. 2012;61((Bellows B.K.; Ye X.; Unni S.; Mukherjee J.; Iloeje U.H.; McAdam-Marx C.) Salt Lake City, United States):A236-A7. Meeting abstract Bensimon L, Yin H, Suissa S, et al. The use of metformin in patients with prostate cancer and the risk of death. Cancer Epidemiol Biomarkers Prev. 2014 Oct;23(10):2111-8. PMID: 25017246. No drug comparison of interest Berard E, Bongard V, Arveiler D, et al. 14- year risk of all-cause mortality according to hypoglycemic drug exposure in a general population. European Heart Journal. 2011;32((Berard E.; Bongard V.; Ruidavets J.B.) University Hospital of Toulouse, Department of Epidemiology, Inserm UMR1027, Toulouse, France):973-4. Meeting abstract Berberoglu Z, Yazici AC, Bayraktar N, et al. Rosiglitazone decreases fasting plasma peptide YY3-36 in type 2 diabetic women: a possible role in weight gain? Acta Diabetol. 2012 Dec;49 Suppl 1:S115-22. PMID: 22101910. No drug comparison of interest Berberoglu Z, Yazici AC, Demirag NG. Effects of rosiglitazone on bone mineral density and remodelling parameters in Postmenopausal diabetic women: a 2-year follow-up study. Clin Endocrinol (Oxf). 2010 Sep;73(3):305-12. PMID: 20148906. No drug comparison of interest; Placebo- controlled trial Bergenstal R, Wysham C, Yan P, et al. Duration-2: Exenatide once weekly demonstrated significant glycaemic control and weight reduction compared to sitagliptin or pioglitazone after 26 weeks of treatment. Diabetic Medicine. 2010;27(2):5. Meeting abstract Bergenstal R, Wysham C, Yan P, et al. DurAtion-2: Exenatide once weekly demonstrated superior glycemic control and weight reduction compared to sitagliptin or pioglitazone after 26 weeks of treatment. Diabetes. 2009;58((Bergenstal R.; Wysham C.; Yan P.; Macconell L.; Malloy J.; Porter L.)):LB2-LB3. Meeting abstract Bergenstal RM, Li Y, Porter TK, et al. Exenatide once weekly improved glycaemic control, cardiometabolic risk factors and a composite index of an HbA1c < 7%, without weight gain or hypoglycaemia, over 52 weeks. Diabetes Obes Metab. 2013 Mar;15(3):264-71. PMID: 23078638. No drug comparison of interest; Does not meet study design criteria Bergenstal RM, Rosenstock J, Arakaki RF, et al. A randomized, controlled study of once-daily LY2605541, a novel long-acting basal insulin, versus insulin glargine in basal insulin-treated patients with type 2 diabetes. Diabetes Care. 2012 Nov;35(11):2140-7. PMID: 22787177. No drug comparison of interest Berria R, Rosenstock J, Silberman C, et al. Weight loss and associated changes in glycaemic control and cardiovascular biomarkers in patients with type 2 diabetes mellitus receiving incretin therapies in a
- 443. C-11 large cohort database.Diabetologia. 2009;52(S1):S297. Meeting abstract Berstein LM, Boyarkina MP, Teslenko SY. Familial diabetes is associated with reduced risk of cancer in diabetic patients: a possible role for metformin. Med Oncol. 2012 Jun;29(2):1308-13. PMID: 21298495. Does not apply Berthet S, Olivier P, Montastruc JL, et al. Drug safety of rosiglitazone and pioglitazone in France: a study using the French PharmacoVigilance database. BMC Clin Pharmacol. 2011;11:5. PMID: 21609444. Background medications; No drug comparison of interest Bertrand OF, Poirier P, Rodes-Cabau J, et al. Cardiometabolic effects of rosiglitazone in patients with type 2 diabetes and coronary artery bypass grafts: A randomized placebo- controlled clinical trial. Atherosclerosis. 2010 Aug;211(2):565-73. PMID: 20594555. Background medications Best JD, Drury PL, Davis TM, et al. Glycemic control over 5 years in 4,900 people with type 2 diabetes: real-world diabetes therapy in a clinical trial cohort. Diabetes Care. 2012 May;35(5):1165-70. PMID: 22432105. Does not account for confounding; No drug comparison of interest Best JH, Hoogwerf BJ, Herman WH, et al. Risk of cardiovascular disease events in patients with type 2 diabetes prescribed the glucagon-like peptide 1 (GLP-1) receptor agonist exenatide twice daily or other glucose-lowering therapies: a retrospective analysis of the LifeLink database. Diabetes Care. 2011 Jan;34(1):90-5. PMID: 20929995. No drug comparison of interest Best JH, Hoogwerf BJ, Pelletier EM, et al. Risk of cardiovascular events in patients with type 2 diabetes treated with exenatide or other glucose-lowering therapies: A retrospective analysis of the LifeLinkTM database. Diabetologia. 2010;53((Best J.H.; Wenten M.) Medical Research, Amylin Pharmaceuticals, San Diego, United States):S333. Meeting abstract Best JH, Little W, Chiquette E, et al. The risk of heart failure among patients receiving exenatide versus other glucose-lowering medications for type 2 diabetes: A matched retrospective cohort analysis of the GE healthcare electronic medical record database. Diabetes. 2011;60((Best J.H.; Little W.; Chiquette E.; Saunders W.B.) San Diego, United States):A311. Meeting abstract Best JH, Pelletier E, Hoogwerf BJ, et al. Risk of cardiovascular events in patients with diabetes treated with exenatide or thiazolidinediones: A retrospective analysis of the pharmetrics database. Value in Health. 2010;13(3):A56. Meeting abstract Best JH, Rubin RR, Peyrot M, et al. Weight- related quality of life, health utility, psychological well-being, and satisfaction with exenatide once weekly compared with sitagliptin or pioglitazone after 26 weeks of treatment. Diabetes Care. 2011 Feb;34(2):314-9. PMID: 21270189. No outcome of interest Best JH, Wintle M, Saunders WB, et al. Glycemic outcomes among patients receiving exenatide bid or liraglutide for type 2 diabetes in clinical practice: A
- 444. C-12 retrospective analysis ofthe ge centricity emr data. Value in Health. 2011;14(7):A473. Meeting abstract Best JH, Yan P, Malloy J. DURATION 2: Weight-related quality of life, psychological well-being, and satisfaction with exenatide once weekly compared to sitagliptin or piaglitazone after 26 weeks of treatment. Diabetologia. 2009;52(S1):S292-S3. Meeting abstract Bethel MA, Green JB, Milton J, et al. Regional, age and sex differences in baseline characteristics of patients enrolled in the Trial Evaluating Cardiovascular Outcomes with Sitagliptin (TECOS). Diabetes Obes Metab. 2015 Jan 20PMID: 25600421. Background medications Bhushan R, Elkind-Hirsch KE, Bhushan M, et al. Improved glycemic control and reduction of cardiometabolic risk factors in subjects with type 2 diabetes and metabolic syndrome treated with exenatide in a clinical practice setting. Diabetes Technol Ther. 2009 Jun;11(6):353-9. PMID: 19459763. Background medications; No drug comparison of interest Bi Y, Yang H, Zhu D, et al. Effects of exenatide, insulin, and pioglitazone on liver fat content and body fat distributions in newly diagnosed subjects with type 2 diabetes. Diabetes. 2014;63((Bi Y.; Yang H.; Zhu D.; Weng J.) Nanjing, China, Guangzhou, China):A459. Meeting abstract Bilezikian J, Kravitz B, Lewiecki EM, et al. Effects of rosiglitazone on bone: Assessing qct parameters in a mechanistic study in postmenopausal women with type 2 diabetes mellitus. Journal of Bone and Mineral Research. 2010;25((Bilezikian J.) Columbia University, College of Physicians and Surgeons, United States):S143. Meeting abstract Blak BT, Rigney U, Ycas J, et al. Baseline characteristics, weight and glycaemic change among patients in the United Kingdom with type 2 diabetes mellitus (T2DM) prescribed a new antidiabetic treatment class in a real world setting. Value in Health. 2013;16(7):A432. Meeting abstract Blickensderfer A, Pencek R, Li Y, et al. Exenatide once weekly: A retrospective analysis of pooled exenatide clinical trial efficacy data stratified by race, age, duration of diabetes, BMI and gender. Diabetologia. 2011;54((Blickensderfer A.; Pencek R.; Li Y.; Brunell S.) Amylin Pharmaceuticals, Inc., San Diego, United States):S315. Meeting abstract Blin P, Lassalle R, Dureau C, et al. Insulin glargine and risk of cancer: A cohort study in the french national healthcare insurance database. Pharmacoepidemiology and Drug Safety. 2012;21((Blin P.; Lassalle R.; Dureau C.; Ambrosino B.; Bernard M.-A.; Abouelfath A.; Gin H.; Pariente A.; Droz- Perroteau C.; Moore N.) Univ. Bordeaux, Bordeaux, France):8. Meeting abstract Blin P, Lassalle R, Dureau-Pournin C, et al. Insulin glargine and risk of cancer: a cohort study in the French National Healthcare Insurance Database. Diabetologia. 2012 Mar;55(3):644-53. PMID: 22222504. No drug comparison of interest Blonde L, Pencek R, MacConell L. Association among weight change, glycemic control, and markers of cardiovascular risk with exenatide once weekly: a pooled analysis of patients with type 2 diabetes.
- 445. C-13 Cardiovasc Diabetol. 2015Feb 3;14(1):12. PMID: 25645567. No drug comparison of interest Blonde L, Rosenstock J, Sesti G, et al. Liraglutide provides superior glycaemic control vs exenatide when added to metformin and/or sulphonylurea (SU) in type 2 diabetes (T2DM). Journal of Diabetes. 2009;1((Blonde L.) Oschner Diabetes Research Unit, New Orleans, United States):A24. Meeting abstract Blonde L, Woo V, Mathieu C, et al. Achievement of diabetes-related treatment goals with canagliflozin (CANA) in patients with type 2 diabetes mellitus (T2DM). Diabetes. 2014;63((Blonde L.; Woo V.; Mathieu C.; Yee J.; Vijapurkar U.; Meininger G.) New Orleans, LA, Winnipeg, MB, Canada, Leuven, Belgium, Raritan, NJ):A284. Meeting abstract Bloomgarden ZT. Type 2 diabetes: Uses of thiazolidinediones and insulin. Diabetes Care. 2011;34(2):e11-e6. No original data Bloomgren G, Dore D, Patterson R, et al. Incidence of acute pancreatitis in exenatide initiators compared to other antidiabetic drug initiators: A retrospective, cohort study. Diabetes. 2009;58((Bloomgren G.; Dore D.; Patterson R.; Noel R.; Braun D.; Seeger J.)). Meeting abstract Bo S, Ciccone G, Rosato R, et al. Cancer mortality reduction and metformin: a retrospective cohort study in type 2 diabetic patients. Diabetes Obes Metab. 2012 Jan;14(1):23-9. PMID: 21812892. No drug comparison of interest; Background medications Boardman MK, Hanefeld M, Kumar A, et al. DURATION-4: Improvements in glucose control and cardiovascular risk factors in patients with type 2 diabetes treated with exenatide once weekly, metformin, pioglitazone, or sitagliptin. Diabetologia. 2011;54((Boardman M.K.; Northrup J.; Chan M.) Eli Lilly and Company, Indianapolis, United States):S314. Meeting abstract Bode B, DeFronzo R, Bergenstal R, et al. Effect of liraglutide 3.0/1.8 mg on body weight and cardiometabolic risk factors in overweight/obese adults with type 2 diabetes: SCALE diabetes randomised, double-blind, 56-week trial. Diabetologia. 2014;57(1):S83. Meeting abstract Bode B, Stenlof K, Harris S, et al. Long- term efficacy and safety of canagliflozin (CANA) in older patients with type 2 diabetes mellitus (T2DM) over 104 weeks. Diabetes. 2014;63((Bode B.; Stenlof K.; Harris S.; Sullivan D.; Fung A.; Usiskin K.) Atlanta, GA, Gothenburg, Sweden, London, ON, Canada, Raritan, NJ):A71. Meeting abstract Bode B, Stenlof K, Harris S, et al. Long- term efficacy and safety of canagliflozin over 104 weeks in patients aged 55-80 years with type 2 diabetes. Diabetes Obes Metab. 2015 Mar;17(3):294-303. PMID: 25495720. Placebo-controlled trial; No drug comparison of interest Bode B, Stenlof K, Sullivan D, et al. Efficacy and safety of canagliflozin (CANA), a sodium glucose cotransporter 2 inhibitor (SGLT2), in older subjects with type 2 diabetes mellitus. Diabetologia. 2012;55((Bode B.) Atlanta Diabetes Associates, Atlanta, United States):S315.
- 446. C-14 Meeting abstract Bode B,Stenlof K, Sullivan D, et al. Efficacy and safety of canagliflozin treatment in older subjects with type 2 diabetes mellitus: a randomized trial. Hosp Pract (1995). 2013;41(2):72-84. Background medications; No drug comparison of interest Bode BW, Brett J, Falahati A, et al. Comparison of the efficacy and tolerability profile of liraglutide, a once-daily human GLP-1 analog, in patients with type 2 diabetes >/=65 and <65 years of age: a pooled analysis from phase III studies. Am J Geriatr Pharmacother. 2011 Dec;9(6):423- 33. PMID: 22055210. No drug comparison of interest; Background medications Bode BW, Testa MA, Magwire M, et al. Patient-reported outcomes following treatment with the human GLP-1 analogue liraglutide or glimepiride in monotherapy: results from a randomized controlled trial in patients with type 2 diabetes. Diabetes Obes Metab. 2010 Jul;12(7):604-12. PMID: 20590735. No outcome of interest Bodmer M, Meier C, Krahenbuhl S, et al. Long-term metformin use is associated with decreased risk of breast cancer. Diabetes Care. 2010 Jun;33(6):1304-8. PMID: 20299480. No drug comparison of interest Bolinder J, Ljunggren O, Johansson L, et al. Dapagliflozin produces long-term reductions in body weight, waist circumference and total fat mass in patients with type 2 diabetes inadequately controlled on metformin. Diabetologia. 2012;55((Bolinder J.) Karolinska Institute, Stockholm, Sweden):S308. Meeting abstract Bolliger D, Seeberger MD, Lurati Buse G, et al. The influence of pre-admission hypoglycaemic therapy on cardiac morbidity and mortality in type 2 diabetic patients undergoing major non-cardiac surgery: A prospective observational study. Anaesthesia. 2012;67(2):149-57. Background medications; No drug comparison of interest Bonds DE, Miller ME, Dudl J, et al. Severe hypoglycemia symptoms, antecedent behaviors, immediate consequences and association with glycemia medication usage: Secondary analysis of the ACCORD clinical trial data. BMC Endocrine Disorders. 2012;12((Bonds D.E., [email protected]) National Heart Lung and Blood Institute, National Institute of Health, Bethesda, MD, United States). No drug comparison of interest Bonora E, Minervini G, Cook W, et al. Saxagliptin reduces A1C and is well tolerated in patients with type 2 diabetes and high framingham cardiovascular risk or albuminuria. Endocrine Practice. 2014;20(1):34A-5A. Meeting abstract Bosco JL, Antonsen S, Sorensen HT, et al. Metformin and incident breast cancer among diabetic women: a population-based case- control study in Denmark. Cancer Epidemiol Biomarkers Prev. 2011 Jan;20(1):101-11. PMID: 21119073. No drug comparison of interest Bosi E, Ellis G, Moneuse P, et al. Addition of alogliptin vs uptitration of pioglitazone dose in type 2 diabetes mellitus patients on metformin plus pioglitazone therapy. Diabetologia. 2010;53((Bosi E.) Istituto
- 447. C-15 Scientifico San Raffaele,Milano, Italy):S328-S9. Meeting abstract Boule NG, Kenny GP, Larose J, et al. Does metformin modify the effect on glycaemic control of aerobic exercise, resistance exercise or both? Diabetologia. 2013 Nov;56(11):2378-82. PMID: 23975325. Does not apply; No drug comparison of interest Bouzamondo H, Karyekar C, Berglind N, et al. Saxagliptin (SAXA) vs Glipizide (GLIP) as add-on therapy to metformin (MET) for type 2 diabetes (T2D): Assessment of HbA1c, hypoglycemia, and weight gain. Diabetes. 2011;60((Bouzamondo H.; Karyekar C.; Berglind N.; Allen E.) Princeton, United States):A305. Meeting abstract Bowker S, Yasui Y, Veugelers P, et al. Effect of glitazones on cancer mortality in type 2 diabetes. Canadian Journal of Diabetes. 2009;33(3):264. Meeting abstract Bowker SL, Yasui Y, Veugelers P, et al. Glucose-lowering agents and cancer mortality rates in type 2 diabetes: assessing effects of time-varying exposure. Diabetologia. 2010 Aug;53(8):1631-7. PMID: 20407744. Background medications; No drug comparison of interest Boyko EJ, Wheeler S, Moore K, et al. Mortality among veterans with type 2 diabetes initiating metformin, sulfonylurea, or rosiglitazone monotherapy. Diabetes. 2013;62((Boyko E.J.; Wheeler S.; Moore K.; Forsberg C.W.; Riley K.; Floyd J.S.; Smith N.L.) Seattle, United States):A405. Meeting abstract Boyle J, Fisher M. The addition of insulin to metformin and sulphonylureas: Results of the 4-T study. Practical Diabetes International. 2010;27(1):5-6. No original data Brady EM, Davies MJ, Gray LJ, et al. A randomized controlled trial comparing the GLP-1 receptor agonist liraglutide to a sulphonylurea as add on to metformin in patients with established type 2 diabetes during Ramadan: the Treat 4 Ramadan Trial. Diabetes Obes Metab. 2014 Jun;16(6):527- 36. PMID: 24373063. Background medications; No drug comparison of interest Brady EM, Davies MJ, Gray LJ, et al. Treat 4 Ramadan trial: A randomised control trial comparing liraglutide vs a sulphonylurea as add-on to metformin in patients with established type 2 diabetes. Diabetologia. 2013;56((Brady E.M.) Leicester Diabetes Centre, University of Leicester, Leicester, United Kingdom):S371. Meeting abstract Brake JA, Hopkins M, Greenwood A, et al. First three months data from an observational study in people with diabetes commenced on exenatide within a large diabetes centre in both insulin treated and insulin naive patients. Diabetic Medicine. 2009;26((Brake J.A.; Hopkins M.; Greenwood A.; Spelman S.; Brame C.) Diabetes Centre, Royal Liverpool and Broadgreen University Hospital Trust, Liverpool, United Kingdom):142. Meeting abstract Bravis V, Hui E, Gohel B, et al. Impact of the READ (Ramadan focused Education and Awareness in Diabetes) programme on HbA1c, weight and hypoglycaemia. Diabetologia. 2009;52(S1):S388-S9. Meeting abstract
- 448. C-16 Bray GA, SmithSR, Banerji M, et al. Effect of pioglitazone on body fat and bone mineral content in the act now trial. Diabetes. 2012;61((Bray G.A.; Smith S.R.; Banerji M.; Tripathy D.; Buchanan T.; Kitabchi A.E.; Henry R.; Stentz F.B.; Musi N.; Schwenke D.C.; Reaven P.; Defronzo R.A.) Baton Rouge, United States):A236. Meeting abstract Breunig IM, Shaya FT, McPherson ML, et al. Development of heart failure in medicaid patients with type 2 diabetes treated with pioglitazone, rosiglitazone, or metformin. J Manag Care Pharm. 2014 Sep;20(9):895- 903. PMID: 25166288. Background medications Brixner D, McAdam-Marx C, Ye X, et al. 18 Month A1C and weight outcomes of exenatide therapy in patients with type-2 diabetes in a real-world study. Value in Health. 2009;12(3):A97. Meeting abstract Brodovicz KG, Kou TD, Alexander CM, et al. Recent trends in the characteristics of patients prescribed sitagliptin and other oral antihyperglycaemic agents in a large U.S. claims database. Int J Clin Pract. 2013 May;67(5):449-54. PMID: 23574104. Does not apply Bron M, Chen K, Cheng D, et al. Comparison of clinical and economic outcomes associated with dpp4 inhibitors (DPP4I) versus sulfonylurea (SU) in combination with metformin (MET) or pioglitazone (PIO) for the treatment of type 2 diabetes mellitus (T2DM). Value in Health. 2012;15(7):A661-A2. Meeting abstract Bron M, Marynchenko M, Yang H, et al. Hypoglycaemia in adult vs. elderly type 2 diabetes mellitus patients: Risks, costs, and impact on treatment persistence in a U.S. population. Diabetologia. 2011;54((Bron M.; Yang Y.) Global Health Economics and Outcomes Research, Takeda Pharmaceuticals International, Inc., Deerfield, United States):S266. Meeting abstract Bron M, Marynchenko M, Yang H, et al. Hypoglycemia in adult vs elderly type 2 diabetes mellitus patients: Risks, costs, and impact on treatment persistence. Diabetes. 2011;60((Bron M.; Marynchenko M.; Yang H.; Yang Y.; Wu E.; Peng A.) Deerfield, United States):A323. Meeting abstract Bron M, Marynchenko M, Yang H, et al. Hypoglycemia, treatment discontinuation, and costs in patients with type 2 diabetes mellitus on oral antidiabetic drugs. Postgrad Med. 2012 Jan;124(1):124-32. PMID: 22314122. Background medications Brouwer ES, West SL, Kluckman M, et al. Initial and subsequent therapy for newly diagnosed type 2 diabetes patients treated in primary care using data from a vendor-based electronic health record. Pharmacoepidemiol Drug Saf. 2012 Sep;21(9):920-8. PMID: 22250059. No outcome of interest Brown B, Sharp P. Predictive factors in glycaemic response to exenitide and sitagliptin treatment. Diabetic Medicine. 2009;26((Brown B.; Sharp P.) Department of Diabetes, Southampton General Hospital, Southampton, United Kingdom):139. Meeting abstract Bruderer SG, Bodmer M, Jick SS, et al. Incidence of and risk factors for severe hypoglycaemia in treated type 2 diabetes
- 449. C-17 mellitus patients inthe UK - a nested case- control analysis. Diabetes Obes Metab. 2014 Feb 25PMID: 24612200. No drug comparison of interest; Does not meet study design criteria Bruderer SG, Jick SS, Bader G, et al. Incidence of and risk factors for severe hypoglycemia in treated type 2 diabetes mellitus patients in the United Kingdom. International Journal of Clinical Pharmacy. 2013;35(6):1333-4. Meeting abstract Bruhn C. Treatment of type 2 diabetes: Exenatide for once weekly application. Deutsche Apotheker Zeitung. 2011;151(26):39-40. No original data Brunell SC, Pencek R, Li Y, et al. Exenatide once weekly was associated with improved glycemic control regardless of baseline body weight. Diabetes. 2012;61((Brunell S.C.; Pencek R.; Li Y.; Hoogwerf B.J.) San Diego, United States):A297. Meeting abstract Bryzinski B, Allen E, Cook W, et al. Saxagliptin efficacy and safety in patients with type 2 diabetes receiving concomitant statin therapy. J Diabetes Complications. 2014 Nov-Dec;28(6):887-93. PMID: 25168266. Placebo-controlled trial Bunck MC, Cornér A, Eliasson B, et al. Effects of exenatide on measures of ?-cell function after 3 years in metformin-treated patients with type 2 diabetes. Diabetes care; 2011. p. 2041-7. Received more than the FDA-approved dose of exenatide Bunck MC, Corner A, Eliasson B, et al. Effects of exenatide on measures of beta-cell function after 3 years in metformin-treated patients with type 2 diabetes. Diabetes Care. 2011 Sep;34(9):2041-7. PMID: 21868779. Received more than the FDA-approved dose of exenatide Bunck MC, Corner A, Eliasson B, et al. Extended, 3-year, exenatide therapy shows sustainable effects on beta cell disposition index in metformin treated patients with type 2 diabetes. Diabetologia. 2010;53((Bunck M.C.; Heine R.J.; Diamant M.) Department of Endocrinology/Diabetes Center, VU university medical center, Amsterdam, Netherlands):S338. Meeting abstract Bunck MC, Diamant M, Corner A, et al. One-year treatment with exenatide improves beta-cell function, compared with insulin glargine, in metformin-treated type 2 diabetic patients: a randomized, controlled trial. Diabetes Care. 2009 2009 May;32(5):762-8. Received more than the FDA-approved dose of exenatide Bunck MC, Corner A, Eliasson B, et al. One year exenatide therapy, compared with insulin glargine, reduces postprandial oxidative stress in metformin-treated patients with type 2 diabetes. Diabetes. 2009;58((Bunck M.C.; Corner A.; Eliasson B.; Wu Y.; Shaginian R.M.; Yan P.; Heine R.J.; Smith U.; Yki-Jarvinen H.; Taskinen M.-R.; Diamant M.)). Meeting abstract Bunck MC, Diamant M, Eliasson B, et al. Beneficial changes on body composition and circulating adiponectin and hsCRP Levels following one year of exenatide therapy, compared with insulin glargine, in metformin-treated patients with type 2 diabetes. Diabetes. 2009;58((Bunck M.C.; Diamant M.; Eliasson B.; Corner A.; Wu Y.;
- 450. C-18 Shaginian R.M.; YanP.; Taskinen M.-R.; Heine R.J.; Yki-Jarvinen H.; Smith U.)). Meeting abstract Burant C, Fleck P, Wilson C, et al. Effect of alogliptin combined with pioglitazone on beta cell function and insulin resistance in metformin-treated patients with type 2 diabetes. Diabetologia. 2009;52(S1):S314- S5. Meeting abstract Burant CF, Viswanathan P, Marcinak J, et al. TAK-875 versus placebo or glimepiride in type 2 diabetes mellitus: a phase 2, randomised, double-blind, placebo- controlled trial. Lancet. 2012 Apr 14;379(9824):1403-11. PMID: 22374408. Meeting abstract Burr N, Talboys R, Savva S, et al. Type 2 diabetes as a positive risk factor in the aetiology of cholangiocarcinoma: A case- control study in two UK centres. Gut. 2012;61((Burr N.; Rushbrook S.; Phillips M.; Hart A.) Department of Gastroenterology, Norfolk and Norwich University Hospital, Norwich, United Kingdom):A220-A1. Meeting abstract Buse J, Sesti G, Schmidt WE, et al. Glycaemic control improves in type 2 diabetes patients when switching from twice-daily exenatide to once-daily liraglutide. Canadian Journal of Diabetes. 2009;33(3):290. Meeting abstract Buse JB, Garber A, Rosenstock J, et al. Liraglutide treatment is associated with a low frequency and magnitude of antibody formation with no apparent impact on glycemic response or increased frequency of adverse events: results from the Liraglutide Effect and Action in Diabetes (LEAD) trials. J Clin Endocrinol Metab. 2011 Jun;96(6):1695-702. PMID: 21450987. Does not apply; Background medications Buse JB, Nauck M, Forst T, et al. Exenatide