JAMA-PDE5i in heart failure

ORIGINAL CONTRIBUTION

ONLINE FIRST

Effect of Phosphodiesterase-5 Inhibition
on Exercise Capacity and Clinical Status
in Heart Failure With Preserved Ejection Fraction
A Randomized Clinical Trial
Margaret M. Redfield, MD
Horng H. Chen, MD
Barry A. Borlaug, MD
Marc J. Semigran, MD
Kerry L. Lee, PhD
Gregory Lewis, MD
Martin M. LeWinter, MD
Jean L. Rouleau, MD
David A. Bull, MD
Douglas L. Mann, MD
Anita Deswal, MD
Lynne W. Stevenson, MD
Michael M. Givertz, MD
Elizabeth O. Ofili, MD
Christopher M. O’Connor, MD
G. Michael Felker, MD
Steven R. Goldsmith, MD
Bradley A. Bart, MD
Steven E. McNulty, MS
Jenny C. Ibarra, MSN
Grace Lin, MD
Jae K. Oh, MD
Manesh R. Patel, MD
Raymond J. Kim, MD
Russell P. Tracy, PhD
Eric J. Velazquez, MD
Kevin J. Anstrom, PhD
Adrian F. Hernandez, MD
Alice M. Mascette, MD
Eugene Braunwald, MD
for the RELAX Trial
1268

Importance Studies in experimental and human heart failure suggest that phosphodiesterase-5 inhibitors may enhance cardiovascular function and thus exercise capacity in heart failure with preserved ejection fraction (HFPEF).
Objective To determine the effect of the phosphodiesterase-5 inhibitor sildenafil compared with placebo on exercise capacity and clinical status in HFPEF.
Design Multicenter, double-blind, placebo-controlled, parallel-group, randomized clinical trial of 216 stable outpatients with HF, ejection fraction Ն50%, elevated N-terminal
brain-type natriuretic peptide or elevated invasively measured filling pressures, and reduced
exercise capacity. Participants were randomized from October 2008 through February 2012
at 26 centers in North America. Follow-up was through August 30, 2012.
Interventions Sildenafil (n=113) or placebo (n=103) administered orally at 20 mg,
3 times daily for 12 weeks, followed by 60 mg, 3 times daily for 12 weeks.
Main Outcome Measures Primary end point was change in peak oxygen consumption after 24 weeks of therapy. Secondary end points included change in 6-minute walk
distance and a hierarchical composite clinical status score (range, 1-n, a higher value indicates better status; expected value with no treatment effect, 95) based on time to death,
time to cardiovascular or cardiorenal hospitalization, and change in quality of life for participants without cardiovascular or cardiorenal hospitalization at 24 weeks.
Results Median age was 69 years, and 48% of patients were women. At baseline,
median peak oxygen consumption (11.7 mL/kg/min) and 6-minute walk distance (308
m) were reduced. The median E/e' (16), left atrial volume index (44 mL/m2), and pulmonary artery systolic pressure (41 mm Hg) were consistent with chronically elevated
left ventricular filling pressures. At 24 weeks, median (IQR) changes in peak oxygen consumption (mL/kg/min) in patients who received placebo (Ϫ0.20 [IQR, Ϫ0.70 to 1.00])
or sildenafil (Ϫ0.20 [IQR, Ϫ1.70 to 1.11]) were not significantly different (P=.90) in
analyses in which patients with missing week-24 data were excluded, and in sensitivity
analysis based on intention to treat with multiple imputation for missing values (mean
between-group difference, 0.01 mL/kg/min, [95% CI, Ϫ0.60 to 0.61]). The mean clinical status rank score was not significantly different at 24 weeks between placebo (95.8)
and sildenafil (94.2) (P=.85). Changes in 6-minute walk distance at 24 weeks in patients who received placebo (15.0 m [IQR, Ϫ26.0 to 45.0]) or sildenafil (5.0 m [IQR,
Ϫ37.0 to 55.0]; P=.92) were also not significantly different. Adverse events occurred in
78 placebo patients (76%) and 90 sildenafil patients (80%). Serious adverse events occurred in 16 placebo patients (16%) and 25 sildenafil patients (22%).
Conclusion and Relevance Among patients with HFPEF, phosphodiesterase-5 inhibition with administration of sildenafil for 24 weeks, compared with placebo, did
not result in significant improvement in exercise capacity or clinical status.
Trial Registration clinicaltrials.gov Identifier: NCT00763867
JAMA. 2013;309(12):1268-1277
Published online March 11, 2013. doi:10.1001/jama.2013.2024
Author Affiliations are listed at the end of this
article.
Corresponding Author: Margaret M. Redfield, MD,

JAMA, March 27, 2013—Vol 309, No. 12

Downloaded From: http://jama.jamanetwork.com/ on 10/24/2013

www.jama.com

Division of Cardiovascular Diseases, Mayo Clinic, 200
First St SW, Guggenheim 9, Rochester, MN 55905
(redfield.margaret@mayo.edu).

©2013 American Medical Association. All rights reserved.

PHOSPHODIESTERASE-5 INHIBITION IN HEART FAILURE

H

EART FAILURE WITH PRE served ejection fraction
(HFPEF) or diastolic heart
failure is a common and
highly morbid condition.1 Clinical trials
of renin-angiotensin system antagonists have not demonstrated improvement in outcomes or clinical status in
HFPEF, and effective therapies are
needed.2 Phosphodiesterase-5 (PDE-5)
metabolizes the nitric oxide and natriuretic peptide systems’ second messenger cyclic guanosine monophosphate and
thus may limit beneficial nitric oxide and
natriuretic peptide actions in the heart,
vasculature, and kidneys.
Preclinical studies suggest that
inhibition of PDE-5 reverses adverse
cardiac structural and functional remodeling and enhances vascular, neuroendocrine, and renal function.3 In clinical studies, PDE-5 inhibitor therapy
improved exercise tolerance and clinical status in patients with idiopathic pulmonary arterial hypertension and in
patients with heart failure and reduced
ejection fraction.4-7 A small, singlecenter study of HFPEF observed
improved hemodynamics, left ventricular diastolic function, right ventricular
systolic function, left ventricular hypertrophy, and lung function with 6 to 12
months of therapy with a PDE-5 inhibitor compared with placebo.8 In aggregate, these studies suggest the potential for PDE-5 inhibition to ameliorate
several key pathophysiological perturbations in HFPEF, and thus improve
exercise capacity and clinical status.
Accordingly, the Phosphodiesterase-5 Inhibition to Improve Clinical Status and Exercise Capacity in Heart Failure with Preserved Ejection Fraction
(RELAX) trial was designed to test the
hypothesis that, compared with placebo, therapy with the PDE-5 inhibitor sildenafil would improve exercise
capacity in HFPEF after 24 weeks of
therapy, assessed by the change in peak
oxygen consumption.

METHODS
Study Oversight

All study participants provided written informed consent prior to screen-

ing. The National Heart, Lung, and
Blood Institute–sponsored Heart Failure Clinical Research Network conceived, designed, and conducted the
RELAX trial. The trial protocol was approved by a National Heart, Lung, and
Blood Institute–appointed protocol review committee and data and safety
monitoring board, and by the institutional review board at each participating site. The Duke Clinical Research Institute served as the data coordinating
center.
Study Design

The rationale for, and design of, the
RELAX trial have been described.3 Patients with normal (Ն50%) ejection
fraction and heart failure with New
York Heart Association functional class
II through IV whose symptoms are
stable while receiving medical therapy
were eligible to participate if they had
objective evidence of heart failure (previous heart failure hospitalization, or
acute heart failure therapy with intravenous diuretic therapy, or chronic loop
diuretic therapy for heart failure with
left atrial enlargement, or invasively
documented elevation in left ventricular filling pressures).
A peak oxygen consumption of 60%
or less of the age- and sex–specific
normal value9 with a respiratory exchange ratio of 1.0 or more at the cardiopulmonary exercise test screening,
and either elevated (Ն400 pg/mL)
N-terminal fragment of the precursor
to brain-type natriuretic peptide (NTproBNP) level or elevation in left ventricular filling pressures at the time of
an NT-proBNP level less than 400
pg/mL were required for study entry.
A complete list of the trial inclusion and
exclusion criteria is provided in the
online supplement (eTable 1, available at http://www.jama.com). As required in federally funded trials, a
self-identification of investigatordefined race or ethnicity option was
recorded.
Participants who met screening criteria underwent baseline studies (history and physical examination, cardiopulmonary exercise test, 6-minute walk



distance, Minnesota Living with Heart
Failure Questionnaire, echocardiography, cardiac magnetic resonance
imaging [CMRI, if sinus rhythm], and
phlebotomy for biomarkers), and were
then randomly assigned, in a 1:1 ratio,
to receive either sildenafil or placebo
with the use of an automated webbased system. A permuted block randomization scheme was used with
stratification according to clinical site
and the presence of atrial fibrillation.
Study drug was administered orally
at 20 mg, 3 times daily for 12 weeks.
At the end of 12 weeks, each patient’s
history and physical examination, cardiopulmonary exercise test, 6-minute
walk distance, and heart failure questionnaire were repeated. Phlebotomy for
sildenafil levels was performed 2 hours
after a scheduled dose of the study drug
were obtained. The dose was then increased to 60 mg, 3 times daily for 12
weeks, after which baseline studies were
repeated, including phlebotomy for
sildenafil levels 2 hours after study drug.
If adverse effects developed, study staff
could recommend discontinuation or
return to a lower or previously tolerated dose of the study drug.
Blinded core laboratories assessed
biomarkers (University of Vermont),
cardiopulmonary exercise tests (Massachusetts General Hospital, Harvard University), CMRIs (Duke University), and
echocardiograms (Mayo Clinic).
Study End Points

The primary end point was the change
in peak oxygen consumption after 24
weeks of therapy. A number of subgroup analyses were prespecified. Secondary end points included a composite hierarchical-rank clinical score in
which patients were ranked (range, 1-n
with data) based on time to death
(tier 1), time to hospitalization for
cardiovascular or cardiorenal causes
(tier 2), and change in heart failure questionnaire from baseline (tier 3) for patients without cardiovascular or cardiorenal hospitalization after 24 weeks of
therapy.10 Because 189 patients had data
for this end point, the anchor value
(mean value in each group indicating no
JAMA, March 27, 2013—Vol 309, No. 12 1269


Figure. Flow of Patients Through the Trial
216 Outpatients with heart
failure randomized a

103 Randomized to receive placebo
103 Received placebo as assigned

113 Randomized to receive sildenaﬁl
113 Received sildenaﬁl as assigned

5 Did not complete 24-week follow-up visit
(withdrew consent) b

12 Did not complete 24-week follow-up visit
9 Withdrew consent c
3 Died

98 Completed 24-week follow-up visit
82 Completed study per protocol
16 Did not complete study per protocol
13 Reduced dosage of study drug
3 Discontinued study drug b

101 Completed 24-week follow-up visit
74 Completed study per protocol
27 Did not complete study per protocol
19 Reduced dosage of study drug
8 Discontinued study drug c

94 Included in the primary analysis
9 Excluded
5 Withdrew consent
2 Unable to perform exercise test
1 Unwilling to perform exercise test
1 Had inadequate peak oxygen consumption data

91 Included in the primary analysis
22 Excluded
9 Withdrew consent
7 Unable to perform exercise test
2 Unwilling to perform exercise test
1 Had inadequate peak oxygen consumption data
3 Died

Some patients discontinued study drug for a specific reason (Ͼ1 may apply) and later withdrew.
a Data regarding the number of patients screened for eligibility was not collected.
b Study drug discontinued due to adverse event (1), physician decision (1), other (2), or for withdrawal of consent (4).
c Study drug discontinued due to adverse event (9), physician decision (1), other (5), or for withdrawal of consent (5).

treatment effect) was 95. Other secondary end points included change in
6-minute walk distance at 24 weeks and
change in peak oxygen consumption and
6-minute walk distance after 12 weeks
of therapy. Peak sildenafil levels at 12
and 24 weeks and coinciding plasma cyclic guanosine monophosphate levels at
24 weeks were assessed. Using other prespecified end points, we assessed the effect of PDE-5 inhibition on left ventricular structure and vascular function by
CMRI, Doppler-estimated diastolic function parameters and pulmonary artery
systolic pressure, and biomarkers that
reflect renal and neuroendocrine function, oxidative stress, and collagen
metabolism.
The percent-predicted peak oxygen
consumption, 6-minute walk distance, and the presence of chronotropic incompetence and left ventricular hypertrophy were calculated using
published criteria (eMethods).11-14
Statistical Analysis

Powercalculationswerebasedonthestandard deviation for change in peak oxygen
1270

consumption and the magnitude of its
change when associated with improvements in other markers of clinical status
(New York Heart Association class,
6-minutewalkdistance,andqualityoflife
scores) in heart failure trials.5,14,15 Based
on these studies, a difference between
treatment groups of 1.2 mL/kg/min in the
change in peak oxygen consumption was
considered clinically significant.
We estimated a 20% rate of incomplete primary end point data due to
death, withdrawal, or incidence of
new factors limiting ability to exercise.
Using a 2-sample t test and a 2-sided ␣
value of .05, a sample size of 190
patients had 85% power to detect a
difference of 1.2 mL/kg/min in change
in peak oxygen consumption, assuming 20% missing data and a standard
deviation of change in peak oxygen
consumption of 2.5 mL/kg/min. As an
early-blinded interim analysis of
aggregated primary end point completeness indicated that the missingness rate approached 20%, the blinded
investigators recommended increasing
the sample size to 215 patients.

JAMA, March 27, 2013—Vol 309, No. 12


The primary analyses were 2-sided,
and patients without 24-week data were
excluded. Sensitivity analysis for the primary end point was based on the intention to treat principle and utilized multiple imputations with 100 imputed data
sets to account for missing 24-week data.
In addition, a prespecified “last observation carried forward” sensitivity analysis used “carry-forward” of 12-week data
if 24-week data were missing.
Data are presented as median (interquartile range [IQR]). For the comparison of treatment groups in the
primary analysis, a multivariable linearregression model was used adjusting for
baseline peak oxygen consumption. A
similar approach was used for the secondary end point of change in 6-minute walk distance, adjusting for baseline 6-minute walk distance. For the
composite hierarchical-rank clinical
score, patients were ranked independent of treatment assignment from 1
(worst outcome—the earliest death) to
n (best outcome—survival with no cardiovascular or cardiorenal hospitalization, and the most favorable improvement in heart failure questionnaire) and
treatments compared using the Wilcoxon rank sum test. For primary and
secondary end points, a P value less
than .05 was considered significant. For
subgroup analyses, a treatment by subgroup interaction P less than .001 was
considered significant.
All analyses were conducted with the
useofSASstatisticalsoftware,version9.2.
RESULTS
Patient Population

A total of 216 patients were enrolled in
the trial from October 13, 2008, through
February21,2012,at26sitesintheUnited
States and Canada (FIGURE). The baseline characteristics were not significantly
different between treatment groups except for a lower prevalence of hypertension in the sildenafil group (TABLE 1).
Patients in this study were older, 248%
women, and obese with controlled blood
pressure and multiple comorbidities includinghypertension,ischemicheartdisease, atrial fibrillation, diabetes, anemia,
and chronic kidney disease.



Table 1. Baseline Characteristics of the Patients
No. (%) of Patients
All (N = 216)
Age, median (IQR), y
Women

Placebo (n = 103)

69 (62-77)

69 (62-77)

68 (62-77)

55 (53)

49 (43)

104 (48)

Self-reported white race

197 (91)

BMI, median (IQR)

32.9 (28.3-39.1)

NYHA functional classification
II

101 (47)

46 (45)

55 (49)

115 (53)

57 (55)

58 (51)

III
MLHFQ total score, median (IQR)
Systolic blood pressure, median (IQR), mm Hg
Heart rate, median (IQR), beats/min
Jugular venous pressure Ն8 cm a
Edema

95 (92)

Sildenafil (n = 113)

32.8 (28.6-38.6)

102 (90)
33.3 (28.2-40.0)

43 (30-62)

43 (29-58)

44 (30-63)

126 (113-138)

127 (113-138)

124 (113-138)

69 (61-78)

68 (62-77)

70 (61-80)

95/209 (45)

48/99 (48)

47/110 (43)

125 (58)

61 (59)

64 (57)

Ն1 Heart failure hospitalizations in previous year

79 (37)

40 (39)

39 (35)

Qualifying ejection fraction, median (IQR), %

60 (55-66)

60 (55-65)

60 (55-66)

Medical history
Hypertension

183 (85)

93 (90)

90 (80)

Ischemic heart disease

84 (39)

37 (36)

47 (42)

Atrial fibrillation or flutter

111 (51)

52 (50)

59 (52)

93 (43)

45 (44)

48 (42)

Diabetes mellitus
COPD

42 (19)

20 (19)

22 (19)

Anemia b

76 (35)

34 (33)

42 (38)

Local laboratory creatinine, median (IQR), mg/dL

1.2 (0.9-1.5)

1.1 (0.9-1.5)

1.3 (1.0-1.5)

Glomerular filtration rate, median (IQR), mL/min/1.73 m2

57 (43-75)

59 (43-76)

57 (43-70)

56 (54)

63 (56)

Stage 3/4 chronic kidney disease
Medications at enrollment
Loop diuretic
Any diuretic

119 (55)
166 (77)

79 (77)

87 (77)

186 (86)

86 (83)

100 (88)

ACE inhibitor or angiotensin II receptor blocker

152 (70)

78 (76)

74 (65)

␤-Blocker

164 (76)

77 (75)

87 (77)

Aldosterone antagonist

23 (11)

9 (9)

14 (12)

Calcium channel blocker

66 (31)

35 (34)

31 (27)

138 (64)

67 (65)

71 (63)

Statin
Core laboratory biomarker data
Cystatin C, median (IQR), mg/L c

1.31 (1.07-1.74)

1.34 (1.08-1.76)

NT-proBNP, median (IQR), pg/mL d

700 (283-1553)

648 (247-1679)

1.29 (1.04,1.70)
757 (318-1465)

Aldosterone, median (IQR), ng/dL c

18.9 (11.8-28.3)

18.0 (11.4-29.4)

20.7 (13.0-27.5)

Endothelin-1, median (IQR), pg/mL c

2.35 (1.85-3.22)

2.36 (1.95-3.20)

2.37 (2.03-3.14)

NT-procollagen III, median (IQR), ␮g/L c

7.7 (6.1-10.0)

7.9 (6.4-11.2)

7.5 (5.5-9.3)

Uric acid, median (IQR), mg/dL e

7.4 (5.8-8.5)

7.1 (5.8-8.5)

7.5 (5.8-8.5)

11.7 (10.2-14.4)

11.9 (10.1-14.4)

11.7 (10.4-14.5)

Peak functional status
Oxygen consumption, median (IQR), mL/kg/min f
Predicted peak oxygen consumption, median (IQR), %

41 (35-49)

41 (36-49)

41 (34-48)

Respiratory exchange ratio, median (IQR) f

1.09 (1.02-1.15)

1.10 (1.03-1.15)

1.09 (1.02-1.16)

Systolic blood pressure, median (IQR), mm Hg g

156 (132-170)

154 (134-168)

156 (130-170)

Chronotropic incompetence f

164 (77)

6-Minute walk distance, median (IQR), m

308 (229-383)

305 (213-372)

308 (241-392)

69 (51-83)

68 (48-83)

70 (54-83)

Predicted 6-minute walk distance, median (IQR), %

79 (78)

85 (76)

Abbreviations: ACE, angiotensin-converting enzyme; BMI, body mass index, calculated as weight in kilograms divided by height in meters squared; COPD, chronic obstructive
pulmonary disease; IQR, interquartile range; MLHFQ, Minnesota Living with Heart Failure Questionnaire; NT-proBNP, N-terminal fragment of the precursor to brain-type natriuretic peptide; NYHA, New York Heart Association.
SI conversion factors: To convert aldosterone from ng/dL to pmol/L, multiply by 27.74; creatinine from mg/dL to ␮mol/L, multiply by 88.4; and uric acid from mg/dL to ␮mol/L,
multiply by 59.485.
a Indicates the number of patients/number of patients with nonmissing data for the variable (percentage).
b Anemia is defined as hemoglobin Ͻ13 g/dL for men, and Ͻ12 g/dL for women.
c Available in 214 (102 placebo and 112 sildenafil-treated) patients.
d Available in 213 (101 placebo and 112 sildenafil-treated) patients.
e Available in 212 (101 placebo and 111 sildenafil-treated) patients.
f Available in 215 (103 placebo and 112 sildenafil-treated) patients due to technical limitation precluding core laboratory measurement of peak oxygen consumption.
g Available in 209 (99 placebo and 110 sildenafil-treated) patients.



JAMA, March 27, 2013—Vol 309, No. 12 1271


The heart failure questionnaire score
was consistent with the distribution of
New York Heart Association functional
class observed in the study population.
Evidence of volume overload (elevated
jugular venous pressure or edema) and
hospitalization for heart failure in the previous year were common. The majority
of patients were taking diuretics, reninangiotensin system antagonists, ␤-blockers, and statins. There was evidence of
neuroendocrine activation and altered
collagen metabolism as levels of NTproBNP, aldosterone, endothelin-1, and
NT-procollagen III exceeded reference
ranges (eTable 2, available at http://www
.jama.com). Both peak oxygen consumption and the 6-minute walk distance were
reduced and chronotropic incompetence was common.
Baseline echocardiographic and
CMRI characteristics of the patients
were not significantly different between treatment groups (TABLE 2). At
echocardiography, ejection fraction and
left ventricular diastolic dimension
(Ͻ5.6 cm) were within normal limits,
while cardiac index was reduced. Nearly
50% of patients had left ventricular hy-

pertrophy or evidence of concentric remodeling or hypertrophy (relative wall
thickness Ն0.42).16 There was Doppler evidence of diastolic dysfunction
and elevated left ventricular filling pressures with reduced early diastolic medial mitral annular tissue velocity (e')
and elevated ratio of early to late transmitral ventricular filling velocity (E/A),
ratio of early transmitral filling velocity to early diastolic medial mitral annular tissue velocity (E/e'), left atrial volume, and pulmonary artery systolic
pressure.
Overall, 132 patients (61%) were eligible for CMRI, and 117 of these underwent CMRI (89%). Ejection fraction and left ventricular end-diastolic
volume index were within normal limits while cardiac index was reduced. In
the CMRI cohort, 25% of patients had
left ventricular hypertrophy. Aortic distensibility was reduced compared with
published values in older participants
without heart failure.17
Primary End Point

At 24 weeks, the median change in peak
oxygen consumption from baseline was

not significantly different in patients
treated with placebo (Ϫ0.20 [IQR,
Ϫ0.70 to 1.00]) and patients treated
with sildenafil (Ϫ0.20 [IQR, Ϫ1.70 to
1.11]), (P=.90) (TABLE 3). Using multiple imputation to account for missing 24-week data, the mean difference
between sildenafil and placebo was 0.01
mL/kg/min (favoring sildenafil, [95%
CI, Ϫ0.60 to 0.61], P=.98). Carrying
forward 12-week peak oxygen consumption when 24-week data were
missing (placebo [n=5] and sildenafil
[n = 9]), median change in peak oxygen consumption (mL/kg/min) from
baseline was Ϫ0.20 (IQR, Ϫ0.83 to
1.10) in patients treated with placebo
and Ϫ0.13 (IQR, Ϫ1.50 to 1.16) in patients treated with sildenafil (P =.98).
In subgroup analyses (eTable 3), the
change in peak oxygen consumption
was not significantly different between treatment groups when analysis was restricted to those patients still
taking the study drug at week 24, in patients with or without left ventricular
hypertrophy by CMRI, with lower or
higher pulmonary artery systolic
pressure, with lower or higher NT-

Table 2. Baseline Core Laboratory Echocardiographic and Cardiac Magnetic Resonance Imaging Characteristics of the Patients
Placebo
No. of Patients
(n = 103)

Sildenafil

Median (IQR)

Doppler echocardiographic data
Ejection fraction, %

103

Cardiac index, L/min/m2

89

2.48 (2.06-2.86)

Left ventricular end-diastolic dimension, cm

76

Left ventricular mass/BSA, g/m2

73

Left ventricular hypertrophy, No. (%)

60 (56-65)

No. of Patients
(n = 113)
110

Median (IQR)
60 (55-65)

All Patients,
Median (IQR)
(N=216)
60 (56-65)

86

2.47 (2.09-2.92)

2.47 (2.07-2.92)

4.6 (4.3-5.1)

88

4.6 (4.2-5.2)

4.6 (4.2-5.2)

77 (63-90)

85

78 (61-97)

77 (62-96)

73

36 (49)

85

40 (47)

76 (48)

Relative wall thickness Ն0.42, No. (%)

73

32 (44)

85

43 (51)

75 (48)

E/A ratio

70

1.3 (1.0-2.0)

72

1.6 (1.0-2.1)

1.5 (1.0-2.0)

Deceleration time, ms

95

190 (156-228)

98

182 (153-212)

185 (155-218)

Medial e', m/s

92

0.06 (0.05-0.07)

105

0.06 (0.05-0.08)

0.06 (0.05-0.08)

Medial E/e'

90

17 (12-24)

98

15 (10-23)

16 (11-23)

Left atrial volume index, mL/m2

74

43 (38-59)

75

44 (33-59)

44 (35-59)

Pulmonary artery systolic pressure, mm Hg

75

41 (34-54)

63

41 (32-51)

41 (33-51)

CMRI data, median (IQR)
Ejection fraction, %

56

66 (59-71)

61

2.38 (1.94-2.80)

Left ventricular end-diastolic volume/BSA, mL/m2

56

58 (49-68)

61

53 (46-66)

56 (47-67)

Left ventricular mass/BSA, g/m2

56

61 (53-73)

61

65 (54-78)

63 (54-77)

Left ventricular hypertrophy, No. (%)

56

14 (25)

61

16 (26)

Aortic distensibility, median (IQR), 10Ϫ3 mm Hg-1

41

45

2.32 (2.11-2.74)

66 (58-70)

56

1.08 (0.63-1.64)

59

66 (55-70)

Cardiac index, L/min/m2

1.18 (0.74-2.07)

2.35 (1.98-2.76)

30 (26)
1.17 (0.67-1.76)

Abbreviations: BSA, body surface area; CMRI, cardiac magnetic resonance imaging; e', early diastolic medial mitral annular tissue velocity; E/A, elevated ratio of early to late transmitral ventricular filling velocity; E/e', ratio of early transmitral filling velocity to early diastolic medial mitral annular tissue velocity; IQR, interquartile range.

1272

JAMA, March 27, 2013—Vol 309, No. 12




proBNP, with or without atrial fibrillation or treated or not with reninangiotensin system antagonists,
␤-blockers, or statins.
Secondary End Points
and Safety Data

There were no significant differences in
the clinical rank score, change in 6-minute walk distance at 24 weeks, or
change in peak oxygen consumption or
6-minute walk distance at 12 weeks between treatment groups (Table 3).
There were no significant differences in
the components of the clinical rank
score at 24 weeks or in the overall incidence of adverse, or serious adverse,
events in the treatment groups.
Adverse events occurred in 78 patients (76%) who received placebo and
90 patients (80%) who received
sildenafil. Serious adverse events occurred in 16 patients (16%) who received placebo and 25 patients (22%)
who received sildenafil. Adverse events
occurring in 5% or more of either study
group are listed in eTable 4 (available at
http://www.jama.com). Patients treated
with sildenafil had a higher incidence of
vascular adverse events, which included (but were not limited to) headache, flushing, and hypotension, although the change in mean arterial
pressure from baseline to 24 weeks was
not significantly different in patients

treated with sildenafil (Ϫ1 [95% CI, Ϫ8
to 6]) and patients treated with placebo
(Ϫ2 [95% CI, Ϫ10 to 7]), (P=.45). All
serious adverse events exclusive of death
or cardiovascular or cardiorenal hospitalization (Table 3) are listed in eTable
5. There were no other notable differences in the incidence of specific serious adverse events between study groups.
Study Drug and Cyclic Guanosine
Monophosphate Levels

Median sildenafil concentrations measured approximately 2 hours after the last
dose were 78 (IQR, 35-130 ng/mL) at 12
weeks and 200 (IQR, 92-330 ng/mL) at
24 weeks. At week 24, there was a weak
correlation between sildenafil dose and
sildenafil level (r = 0.29, P = .008). In
paired analysis, plasma cyclic guanosine monophosphate levels increased significantly from baseline to 24 weeks in
patients randomized to receive sildenafil (mean increase, 8.72 pmol/mL, [95%
CI, 2.56-14.87], P=.006), but not in patients randomized to placebo (mean increase, 1.28 pmol/mL, [95% CI, Ϫ6.27
to 8.83], P=.74), although the change in
cyclic guanosine monophosphate was
not significantly different between groups
(P=.11).
Additional End Points

At CMRI, there was no difference in
change in left ventricular mass or left

ventricular end-diastolic volume index
between treatment groups (TABLE 4).
There was also no difference in change
in Doppler-assessed left ventricular diastolic function parameters or pulmonary artery systolic pressure between
treatment groups. By CMRI, arterial elastance decreased more and systemic vascular resistance tended to decrease more
in patients treated with sildenafil
(P=.09). However, the change in mean
arterial pressure in the entire study
population was not significantly different between groups, as noted above.
More patients had missing data for aortic distensibility at 24 weeks than at baseline, but there was no difference in
change in distensibility between groups.
Patients treated with sildenafil had a
greater increase in creatinine, cystatin C,
NT-proBNP, uric acid, and endothelin-1 than patients treated with placebo, whereas changes in aldosterone
and NT-procollagen III were not significantly different between groups.
COMMENT
To our knowledge, the RELAX trial is
the first multicenter study to investigate the effect of PDE-5 inhibition in
HFPEF. Contrary to our hypothesis,
long-term PDE-5 inhibition in HFPEF
had no effect on maximal or submaximal exercise capacity, clinical status,
quality of life, left ventricular remod-

Table 3. Primary, Secondary, and Safety End Points
Placebo
No. of
Patients
Primary end point
Change in peak oxygen consumption at 24 wk, median (IQR), mL/kg/min

94

Secondary end points
Clinical rank score, mean a

Sildenafil

Variable

No. of
Patients

Variable

P
Value

Ϫ0.20 (Ϫ0.70 to 1.00)

91

Ϫ0.2 (Ϫ1.70 to 1.11)

.90

94.2

.85

94

95.8

95

Change in 6-minute walk distance at 24 wk, median (IQR), m

95

15.0 (Ϫ26.0 to 45.0)

90

5.0 (Ϫ37.0 to 55.0)

Change in peak oxygen consumption at 12 wk, median (IQR), mL/kg/min

96

0.03 (Ϫ1.10 to 0.67)

97

0.01 (Ϫ1.35 to 1.25)

.98

Change in 6-minute walk distance at 12 wk, median (IQR), m

96

18.0 (Ϫ14.5 to 48.0)

99

10.0 (Ϫ25.0 to 36.0)

.13

Components of clinical rank score at 24 wk
Death, No. (%) b
Hospitalization for cardiovascular or renal cause, No. (%)
Change in MLHFQ, median (IQR)
Safety end points, No. (%)
Adverse events
Serious adverse events

103
103
91

0
13 (13)
Ϫ8 (Ϫ21 to 5)

.92

113

3 (3)

113

15 (13)

.89

Ϫ8 (Ϫ19 to 0)

.44

91

.25

103

78 (76)

113

90 (80)

.49

103

16 (16)

113

25 (22)

.22

Abbreviations: IQR, interquartile range; MLHFQ, Minnesota Living with Heart Failure Questionnaire.
a A mean value of 95 in each group is expected under the null hypothesis of no treatment effect.
b Site investigator identified causes of death were sudden death (n=1), progressive cardiorenal failure (n=1), and noncardiovascular (n=1).



JAMA, March 27, 2013—Vol 309, No. 12 1273


eling, diastolic function parameters, or
pulmonary artery systolic pressure. Renal function worsened more and NTproBNP, endothelin-1, and uric acid
levels increased more in patients treated
with sildenafil. Furthermore, there were
more (but not significantly more) patients in the sildenafil group who withdrew consent, died, or were too ill to
perform the cardiopulmonary exercise test, and patients treated with
sildenafil had a higher incidence of vascular adverse events. These findings do
not suggest that therapy with the PDE-5
inhibitor sildenafil provides clinical
benefit in the general HFPEF population.
Given the strong rationale for testing PDE-5 inhibition in HFPEF3 and the
lack of benefit observed, it is important to consider whether the study
population, specifics of the therapeutic intervention, and end points were
appropriate.
The clinical characteristics of patients with HFPEF enrolled in major ongoing or completed clinical trials have
been summarized.18 The RELAX study
population was similar to others in terms

of age, sex distribution, and body size.
Severity of heart failure in our study was
also similar to or greater than other
HFPEF trials, although comorbidity burden (diabetes, atrial fibrillation, kidney
disease) may have been greater in this
study population. Concentric remodeling and hypertrophy were common but
not severe, and Doppler evidence of elevated filling pressures and pulmonary
hypertension were present as was neuroendocrine activation consistent with
the heart failure state.
However, the characteristics of the
study population were notably different from the only other study to our
knowledge that evaluated the effect of
PDE-5 inhibition in HFPEF. In the
study by Guazzi and coauthors, 8
sildenafil had a number of beneficial effects as outlined above, although effect on exercise capacity was not tested.
Importantly, in that study, patients with
HFPEF had fewer comorbidities and
significantly higher blood pressure, left
ventricular mass, and pulmonary artery systolic pressure than the patients in the RELAX study and catheterization-documented pulmonary

arterial hypertension, profound right
ventricular systolic dysfunction, and
right ventricular failure were present.
This profile is somewhat atypical for
HFPEF cohorts.19
It may be that the primary therapeutic effects of PDE-5 inhibitors in heart
failure involve the drugs’ ability to dilate the pulmonary vascular bed, enhance right ventricular contractility,
and reduce ventricular interdependence,4,19-22 and that pulmonary arterial hypertension and right ventricular failure must be significant in order
to observe clinical benefit in HFPEF.
Our subgroup analysis did not show any
trends toward improvement in peak
oxygen consumption in patients with
higher pulmonary artery systolic pressure, but the presence of pulmonary arterial hypertension or right ventricular dysfunction was not assessed in this
study.
Although left ventricular hypertrophy was common in participants in this
study, it was far less severe than among
participants in the study by Guazzi et
al.8 In murine-pressure overload studies, PDE-5 inhibition did not have an-

Table 4. Additional Prespecified End Points
Placebo
No. of
Patients
Change in left ventricular structure by CMRI at 24 wk
Left ventricular mass by CMRI, g
Left ventricular end-diastolic volume by CMRI, mL
Change in diastolic function parameters at 24 wk
Medial e', m/s

Sildenafil

Median (IQR)

No. of
Patients

Median (IQR)

P
Value

47

0.6 (Ϫ5.7 to 7.9)

49

Ϫ1.5 (Ϫ5.9 to 7.1)

.93

47

Ϫ4.3 (Ϫ15.5 to 8.1)

49

3.7 (Ϫ4.9 to 14.5)

.13

83

0.00 (Ϫ0.01 to 0.01)

77

0.00 (Ϫ0.01 to 0.01)

.88

E/e'

80

Ϫ1.6 (Ϫ4.7 to 2.2)

75

0.2 (Ϫ2.4 to 3.1)

.16

PA systolic pressure, mm Hg

58

Change in vascular function by CMRI at 24 wk
Arterial elastance, mm Hg/mL

Ϫ2 (Ϫ8 to 8)

45

2 (Ϫ5 to 7)

.94
.02

45

0.03 (Ϫ0.14 to 0.23)

47

Ϫ0.08 (Ϫ0.41 to 0.12)

Systemic vascular resistance, wood units

45

0.13 (Ϫ0.15 to 0.43)

47

Ϫ0.04 (Ϫ0.38 to 0.21)

.09

Aortic distensibility, 10Ϫ3 mm Hg-1

31

0.13 (Ϫ0.17 to 0.42)

29

0.18 (Ϫ0.12 to 0.67)

.38

Change in core laboratory biomarkers at 24 wk
Creatinine, mg/dL

94

0.01 (Ϫ0.10 to 0.09)

94

0.05 (Ϫ0.04 to 0.15)

.047

Cystatin C, mg/L

95

0.01 (Ϫ0.08 to 0.11)

95

0.05 (Ϫ0.04 to 0.16)

.01

NT-proBNP, pg/mL

94

Ϫ23 (Ϫ198 to 139)

95

15 (Ϫ90 to 372)

.03

Endothelin-1, pg/mL

95

Ϫ0.01 (Ϫ0.48 to 0.47)

95

0.38 (Ϫ0.10 to 0.97)

.046

Aldosterone, ng/dL

95

95

Ϫ1.1 (Ϫ7.7 to 3.0)

NT-procollagen III, ␮g/L

93

Ϫ0.03 (Ϫ1.49 to 1.54)

0 (Ϫ7.0 to 4.8)

95

0.07 (Ϫ1.17 to 1.42)

.85
.77

Uric acid, mg/dL

94

Ϫ0.1 (Ϫ0.7 to 0.7)

94

0.3 (Ϫ0.4 to 1.4)

.02

Abbreviations: CMRI, cardiac magnetic resonance imaging; e', early diastolic medial mitral annular tissue velocity; E/e', ratio of early transmitral filling velocity to early diastolic medial
mitral annular tissue velocity; IQR, interquartile range; NT-proBNP, N-terminal fragment of the precursor to brain-type natriuretic peptide; PA, pulmonary artery.
SI conversions: To convert aldosterone from ng/dL to pmol/L, multiply by 27.74; creatinine from mg/dL to ␮mol/L, multiply by 88.4; and uric acid from mg/dL to ␮mol/L, multiply
by 59.485.

1274

JAMA, March 27, 2013—Vol 309, No. 12




tihypertrophic effects in mice with less
severe pressure overload and compensated left ventricular hypertrophy with
relatively preserved ejection fraction,
whereas dramatic antiremodeling benefits were observed in mice with severe pressure overload, eccentric left
ventricular hypertrophy, reduced ejection fraction, and pulmonary congestion.23,24 Conceivably, activation of
PDE-5 or of cyclic guanosine monophosphate–sensitive downstream pathways in the left ventricle or other organs may occur only in heart failure
associated with advanced left ventricular remodeling.
In a randomized clinical trial of
PDE-5 inhibition in pulmonary arterial hypertension, the effect of sildenafil on exercise capacity was not dose related as improvement in 6-minute walk
distance was seen with 20 mg 3 times
daily after just 4 weeks of therapy with
no further improvement with higher
doses or longer duration of therapy.4
Sildenafil levels at 12 and 24 weeks were
variable, but on average, similar to random levels observed with similar doses
among participants with heart failure
and reduced ejection fraction for whom
improvement in exercise capacity with
long-term sildenafil use was observed
(Gregory Lewis, MD, Massachusetts
General Hospital, written communication, February 2, 2013).5
For patients with heart failure and reduced ejection fraction, the benefit of
PDE-5 inhibition on exercise capacity
has been demonstrated acutely after a
single 50-mg dose,25,26 and with 75 mg,
3 times daily for 12 weeks (dose uptitrated over the course of 6 weeks).5 Although only 73% of patients in the
sildenafil group attending the 24week follow-up were taking the perprotocol dose, 92% of patients were taking at least 20 mg 3 times daily. While
studies in pulmonary arterial hypertension and heart failure and reduced
ejection fraction have observed effects
on exercise capacity with similar doses
and duration of therapy, we cannot exclude the possibility that inadequate
dose or duration of PDE-5 inhibition
contributed to our findings.

Therapeutic sildenafil levels were
associated with minimal increases in
plasma cyclic guanosine monophosphate. Heart failure with preserved
ejection fraction is characterized by
endothelial dysfunction 2 7 and by
lower natriuretic peptide levels than
observed in heart failure and reduced
ejection fraction,28 which may suggest
limited nitric oxide and natriuretic
peptide activity in HFPEF. Inability to
enhance cyclic guanosine monophosphate with PDE-5 inhibition in
HFPEF may have contributed to our
findings.
As previously described,3 change in
peak oxygen consumption was chosen as the primary end point in the
RELAX trial based on previous preclinical and clinical studies and because noncardiovascular comorbidities and motivational factors can
influence measures of submaximal exercise performance in HFPEF. Our trial
was powered to detect a clinically significant difference in the change in peak
oxygen consumption between groups
and the estimate of variability (SD, 2.5
mL/kg/min) in change in peak oxygen
consumption used in the power calculations was consistent with the standard deviation of change in peak oxygen consumption observed in the
patients treated with placebo (2.0 mL/
kg/min). The lack of treatment effect on
submaximal exercise, clinical status,
and physiologic end points supports the
validity of the observed lack of treatment effect on maximal exercise capacity.
The high prevalence of chronotropic incompetence in the study population is noteworthy. Chronotropic incompetence may contribute to exercise
limitation in HFPEF, and may not be
improved by PDE-5 inhibition.
Although numerous studies in animal models of renal dysfunction suggest that PDE-5 inhibition ameliorates progression of renal dysfunction
of various etiologies,3,29-31 in this trial,
modest but statistically significant worsening of renal function was observed
in patients treated with sildenafil and
was associated with concordant in-



creases in NT-proBNP, uric acid, and
endothelin-1, suggesting that the decline in renal function was physiologically significant. Studies in pulmonary arterial hypertension and erectile
dysfunction have not reported worsening of renal function with PDE-5 inhibitor therapy, but little is known of
the effect of PDE-5 inhibition on renal
function in HFPEF.
There were more (but not significantly more) patients who withdrew
consent, died, or were too ill to perform the cardiopulmonary exercise test
in the sildenafil treatment group, potentially accentuating the lack of benefit observed, particularly if those who
withdrew did so due to adverse effects
or poor clinical status.
A modest decrease in arterial elastance was noted in patients treated with
sildenafil in the CMRI cohort. This may
have been related to an effect of
sildenafil on resistance that tended to
decrease more in patients treated with
sildenafil, but in the entire cohort, there
were no differences in change in mean
arterial pressure between treatment
groups.
The findings of this study must be interpreted in the context of other potential limitations. Multicenter trials
using peak oxygen consumption as a
primary end point are challenging, but
rigorous methodologies were used in
the design and execution of the exercise test study protocol.3 Patients were
selected who could perform the test,
and who had significant reduction in
peak oxygen consumption—these entry criteria may have selected for a
unique HFPEF phenotype. The trial was
not powered to address differences in
clinical outcomes.
CONCLUSIONS
Among patients with HFPEF, PDE-5 inhibition with administration of
sildenafil for 24 weeks, compared with
placebo, did not result in significant improvement in exercise capacity or clinical status. Continued efforts to identify key pathophysiologic perturbations
and novel therapeutic targets in HFPEF
are needed.
JAMA, March 27, 2013—Vol 309, No. 12 1275

Published Online: March 11, 2013. doi:10.1001
/jama.2013.2024
Author Affiliations: Department of Medicine, Mayo
Clinic, Rochester, Minnesota (Drs Redfield, Chen,
Borlaug, Lin, and Oh); Department of Medicine, Massachusetts General Hospital, Boston (Drs Semigran and
Lewis); Departments of Biostatistics and Bioinformatics (Drs Lee and Anstrom) and Medicine (Drs Velazquez
and Hernandez), Statistics (Mr McNulty), and Project Management (Ms Ibarra), Duke Clinical Research
Institute, Durham, North Carolina; Departments of
Medicine (Dr LeWinter) and Pathology and Biochemistry (Dr Tracy), University of Vermont, Burlington; Department of Medicine, University of Montreal and
Montreal Heart Institute, Montreal, Canada (Dr Rouleau); Department of Surgery, University of Utah, Salt
Lake City (Dr Bull); Department of Medicine, Barnes
Jewish Hospital, St Louis, Missouri (Dr Mann); Department of Medicine, Michael E. DeBakey VA Medical Center and Baylor College of Medicine, Houston,
Texas (Dr Deswal); Department of Medicine, Brigham
and Women’s Hospital, Boston, Massachusetts (Drs
Stevenson, Givertz, and Braunwald); Department of
Medicine, Morehouse School of Medicine, Atlanta,
Georgia (Dr Ofili); Departments of Medicine (Drs
O’Connor, Felker, Patel, and Kim) and Duke Cardiovascular Magnetic Resonance Center (Dr Kim), Duke
University Medical Center and Duke Heart Center, Durham, North Carolina; Department of Medicine, Hennepin County Medical Center, Minneapolis, Minnesota (Drs Goldsmith and Bart); and Division of
Cardiovascular Sciences, National Heart, Lung, and
Blood Institute, Bethesda, Maryland (Dr Mascette).
For a complete listing of the RELAX site representatives see the online supplement at jama.com.
Author Contributions: Drs Lee and Anstrom had full
access to all the data in the study and take responsibility for the integrity of the data and the accuracy of
the data analysis.
Study concept and design: Chen, Semigran, Lee,
LeWinter, Rouleau, Bull, Deswal, Stevenson, O’Connor,
Felker, Goldsmith, Bart, McNulty, Velazquez,
Hernandez, Braunwald, Redfield.
Acquisition of data: Chen, Borlaug, Semigran, Lee,
Lewis, LeWinter, Rouleau, Bull, Deswal, Stevenson,
Givertz, Ofili, O’Connor, Felker, Bart, McNulty, Ibarra,
Lin, Patel, Kim, Tracy, Velazquez, Hernandez, Redfield.
Analysis and interpretation of data: Borlaug, Semigran,
Lee, Lewis, LeWinter, Bull, Mann, Deswal, Givertz,
Felker, McNulty, Ibarra, Lin, Oh, Patel, Kim, Tracy,
Velazquez, Anstrom, Hernandez, Mascette, Braunwald,
Redfield.
Drafting of the manuscript: Chen, Borlaug, LeWinter,
Bull, O’Connor, McNulty, Velazquez, Redfield.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Lee, Bull, McNulty, Velazquez,
Anstrom.
Obtained funding: Chen, Semigran, Lee, Bull, Deswal,
Ofili, Felker, McNulty, Velazquez, Hernandez,
Mascette, Redfield.
Administrative, technical, or material support: Chen,
Borlaug, Lee, Rouleau, Bull, O’Connor, McNulty, Ibarra,
Oh, Kim, Tracy, Velazquez, Hernandez, Mascette,
Braunwald.
Study supervision: Borlaug, Lewis, Deswal, Givertz,
O’Connor, Felker, Goldsmith, Bart, McNulty, Kim,
Velazquez, Mascette, Braunwald, Redfield.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure
of Potential Conflicts of Interest. Dr Redfield reported receiving financial support from the National
Institutes of Health (NIH) and royalties from Annexon. Dr Chen reported receiving funding for patents that his institution has licensed to Nile Therapeutics and Annexon with other patents pending at the
US Patent and Trademark Office. He also reported that
he has royalties with Nile Therapeutics, Annexon, and
1276

Up To Date. Dr Borlaug reported receiving consulting fees from Medscape, Amgen, and GlaxoSmithKline
and receiving financial support from Atcor Medical and
Gilead. Dr Semigran reported receiving financial support from the NIH for participation in the Heart Failure Network and from Bayer. Dr Lee reported receiving financial support from the NIH for participation
in the Heart Failure Network. Dr LeWinter reported
receiving financial support from the NIH for participation in the Heart Failure Network. Dr Bull reported
receiving financial support from the NIH for participation in the Heart Failure Network. Dr Deswal reported receiving financial support from the NIH for
participation in the Heart Failure Network. Dr Stevenson reported receiving financial support from the NIH
for participation in the Heart Failure Network. Dr Givertz reported receiving financial support from the NIH
for participation in the Heart Failure Network. Dr Ofili
reported receiving financial support from the NIH, receiving consulting fees from Merck, Novartis, and Arbor Pharmaceuticals and lecture fees from Arbor Pharmaceuticals. She is currently employed by the
Morehouse School of Medicine and has grants pending from the National Heart, Lung, and Blood Institute and National Institute on Minority Health and
Health Disparities. Dr O’Connor reported receiving consulting fees from Roche Diagnostics, Amgen, Novartis, Ikaria, Acetlion Pharma, Heartware, ResMed, Pozen,
GE Healthcare, Johnson & Johnson, Gilead, Critical Diagnostics, BG Medicine, Otsuka, Astellas, Novella, Cytokinetics, and Capricor; holding stock options with
Neurotronik/Interventional Autonomics Corporation; editing American College of Cardiology; and coowning Cardiology Consulting Associates. Dr Felker
reported receiving financial support from the NIH; consulting fees from Novartis, Amgen, Trevena, Roche
Diagnostics, Merck, Singulex, and BG Medicine; and
funding from Amgen, Otsuka, and Roche Diagnostics. Dr Bart reported receiving financial support from
the NIH for participation in the Heart Failure Network. Ms Ibarra reported receiving financial support
from the NIH for participation in the Heart Failure Network. Dr Oh received financial support from the NIH
for the Echo Core Lab. Dr Patel has received consulting fees from Genzyme, Baxter, Jensen, and Bayer, and
his financial institution receives financial support from
AstraZeneca, Johnson & Johnson, and Maquet. Dr Kim
reported receiving financial support from the NIH as
the Heart Failure Network Data Coordinating Center
and MRI Core Lab, and funding for a patent with
Northwestern University. His institution has received
a grant from Siemens. Dr Tracy reported receiving financial support from the NIH for participation in the
Heart Failure Network, and consulting fees from Merck,
Tibotec—Johnson & Johnson, and Abbott. He has also
given expert testimony about biomarkers in blood clotting and is the owner of Haematologic Technologies.
Dr Velazquez reported receiving financial support from
the NIH for participation in the Heart Failure Network. Dr Anstrom reported receiving financial support from the NIH for participation in the Heart Failure Network. Dr Hernandez reported receiving financial
support from the NIH for participation in the Heart
Failure Network and consulting fees from AstraZeneca, Corthera, Janseen, and Bristol-Myers Squibb. His
institution has received funding from Amylin, Johnson & Johnson, and Bristol-Myers Squibb. Dr Braunwald reported receiving consulting fees from Merck,
Daiichi Sankyo, Genzyme, Amorcyte, Medicines Co,
MC Communications, Ikaria, CardioRentis, sanofiaventis; funding (or pending funding) from AstraZeneca, Johnson & Johnson, Merck, sanofi-aventis, Daiichi Sankyo, GlaxoSmithKline, Bristol-Myers Squibb,
Beckman Coulter, Roche Diagnostics, and Pfizer; compensation for lectures for Eli Lilly, Merck, CVRx (no
compensation), CV Therapeutics (now Gilead), Daiichi Sankyo, MC Communications, Manarini International, Medscape, and Bayer; and payment for devel-

JAMA, March 27, 2013—Vol 309, No. 12


opment of educational presentations from MC
Communications. No other disclosures were reported.
Funding/Support: This work was supported by grants
U10HL084904 (for the data coordinating center), and
U10HL084861, U10HL084875, U10HL084877,
U10HL084889, U10HL084890, U10HL084891,
U10HL084899, U10HL084907, and U10HL084931
(for the clinical centers) from the National Heart, Lung,
and Blood Institute. This work is also supported by grant
UL1TR000454 from the National Center for Advancing Translational Sciences and grant 8 U54 MD007588
from the National Institute on Minority Health and
Health Disparities. Pfizer provided study drug (sildenafil
and matched placebo).
Role of the Sponsor: Pfizer had no role in the design
and conduct of the study, collection, management,
analysis, or interpretation of the data, or preparation,
review, or approval of the manuscript. National Heart,
Lung, and Blood Institute representatives provided advice on the trial design and appointed the protocol
review committee and data and safety monitoring
board.
Disclaimer: The authors are solely responsible for the
content of this article, which does not necessarily represent the official views of the National Heart, Lung,
and Blood Institute or the National Institutes of Health.
Online-Only Material: The eMethods, eReferences,
eTables 1 through 5, and eAppendices are available
at http://www.jama.com.
Additional Contributions: We thank the patients who
participated in this study, the Heart Failure Network
site investigators and coordinators, the members of
the Heart Failure Network data and safety monitoring board and protocol review committee, and the National Heart, Lung, and Blood Institute representatives. For a complete listing of the Heart Failure
Network members, see Appendix C available at
http://www.jama.com.
REFERENCES
1. Owan TE, Redfield MM. Epidemiology of diastolic heart failure. Prog Cardiovasc Dis. 2005;
47(5):320-332.
2. Shah RV, Desai AS, Givertz MM. The effect of reninangiotensin system inhibitors on mortality and
heart failure hospitalization in patients with heart failure and preserved ejection fraction: a systematic review and meta-analysis. J Card Fail. 2010;16(3):
260-267.
3. Redfield MM, Borlaug BA, Lewis GD, et al; Heart
Failure Clinical Research Network. PhosphdiesteRasE-5
inhibition to improve clinical status and exercise capacity in diastolic heart failure (RELAX) trial: rationale
and design. Circ Heart Fail. 2012;5(5):653-659.
4. Galiè N, Ghofrani HA, Torbicki A, et al; Sildenafil
Use in Pulmonary Arterial Hypertension (SUPER) Study
Group. Sildenafil citrate therapy for pulmonary arterial hypertension. N Engl J Med. 2005;353(20):
2148-2157.
5. Lewis GD, Shah R, Shahzad K, et al. Sildenafil improves exercise capacity and quality of life in patients
with systolic heart failure and secondary pulmonary
hypertension. Circulation. 2007;116(14):15551562.
6. Guazzi M, Vicenzi M, Arena R. Phosphodiesterase 5 inhibition with sildenafil reverses exercise oscillatory breathing in chronic heart failure: a long-term
cardiopulmonary exercise testing placebo-controlled
study. Eur J Heart Fail. 2012;14(1):82-90.
7. Guazzi M, Vicenzi M, Arena R, Guazzi MD. PDE5
inhibition with sildenafil improves left ventricular diastolic function, cardiac geometry, and clinical status
in patients with stable systolic heart failure: results
of a 1-year, prospective, randomized, placebocontrolled study. Circ Heart Fail. 2011;4(1):817.


8. Guazzi M, Vicenzi M, Arena R, Guazzi MD. Pulmonary hypertension in heart failure with preserved
ejection fraction: a target of phosphodiesterase-5 inhibition in a 1-year study. Circulation. 2011;124
(2):164-174.
9. Fletcher GF, Balady G, Froelicher VF, Hartley LH,
Haskell WL, Pollock ML; Writing Group. Exercise standards: a statement for healthcare professionals from
the American Heart Association. Circulation. 1995;
91(2):580-615.
10. Packer M. Proposal for a new clinical end point
to evaluate the efficacy of drugs and devices in the
treatment of chronic heart failure. J Card Fail. 2001;
7(2):176-182.
11. Enright PL, Sherrill DL. Reference equations for
the six-minute walk in healthy adults. Am J Respir Crit
Care Med. 1998;158(5 Pt 1):1384-1387.
12. Khan MN, Pothier CE, Lauer MS. Chronotropic incompetence as a predictor of death among patients with
normal electrograms taking beta blockers (metoprolol or
atenolol). Am J Cardiol. 2005;96(9):1328-1333.
13. Chirinos JA, Segers P, De Buyzere ML, et al. Left
ventricular mass: allometric scaling, normative values, effect of obesity, and prognostic performance.
Hypertension. 2010;56(1):91-98.
14. Feldman AM, Silver MA, Francis GS, et al; PEECH
Investigators. Enhanced external counterpulsation improves exercise tolerance in patients with chronic heart
failure. J Am Coll Cardiol. 2006;48(6):1198-1205.
15. Abraham WT, Fisher WG, Smith AL, et al;
MIRACLE Study Group, Multicenter InSync Randomized Clinical Evaluation. Cardiac resynchronization in
chronic heart failure. N Engl J Med. 2002;346(24):
1845-1853.
16. Lang RM, Bierig M, Devereux RB, et al; Chamber Quantification Writing Group; American Society
of Echocardiography’s Guidelines and Standards
Committee; European Association of Echocardiography.
Recommendations for chamber quantification: a re-

port from the American Society of Echocardiography’s Guidelines and Standards Committee and the
Chamber Quantification Writing Group, developed in
conjunction with the European Association of Echocardiography, a branch of the European Society of
Cardiology. J Am Soc Echocardiogr. 2005;18(12):
1440-1463.
17. Hundley WG, Kitzman DW, Morgan TM, et al.
Cardiac cycle-dependent changes in aortic area and
distensibility are reduced in older patients with isolated diastolic heart failure and correlate with exercise intolerance. J Am Coll Cardiol. 2001;38(3):
796-802.
18. Shah SJ, Heitner JF, Sweitzer NK, et al. Baseline
characteristics of patients in the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist (TOPCAT) trial. Circ Heart Fail.
2012.
19. Forfia PR, Borlaug BA. Letter by Forfia and Borlaug regarding article, “Pulmonary hypertension in
heart failure with preserved ejection fraction: a target of phosphodiesterase-5 inhibition in a 1-year
study.” Circulation. 2012;125(8):e408-e410.
doi:10.1161/CIRCULATIONAHA.111.064584.
20. Nagendran J, Archer SL, Soliman D, et al. Phosphodiesterase type 5 is highly expressed in the hypertrophied human right ventricle, and acute inhibition of phosphodiesterase type 5 improves contractility.
Circulation. 2007;116(3):238-248.
21. Xie YP, Chen B, Sanders P, et al. Sildenafil prevents and reverses transverse-tubule remodeling and
Ca(2ϩ) handling dysfunction in right ventricle failure
induced by pulmonary artery hypertension.
Hypertension. 2012;59(2):355-362.
22. Borgdorff MA, Bartelds B, Dickinson MG, et al.
Sildenafil enhances systolic adaptation, but does not
prevent diastolic dysfunction, in the pressure-loaded
right ventricle. Eur J Heart Fail. 2012;14(9):10671074.



23. Nagayama T, Hsu S, Zhang M, et al. Pressureoverload magnitude-dependence of the antihypertrophic efficacy of PDE5A inhibition. J Mol Cell
Cardiol. 2009;46(4):560-567.
24. Takimoto E, Champion HC, Li M, et al. Chronic
inhibition of cyclic GMP phosphodiesterase 5A prevents and reverses cardiac hypertrophy. Nat Med.
2005;11(2):214-222.
25. Lewis GD, Lachmann J, Camuso J, et al. Sildenafil improves exercise hemodynamics and oxygen uptake in patients with systolic heart failure. Circulation.
2007;115(1):59-66.
26. Guazzi M, Tumminello G, Di Marco F, Fiorentini
C, Guazzi MD. The effects of phosphodiesterase-5 inhibition with sildenafil on pulmonary hemodynamics
and diffusion capacity, exercise ventilatory efficiency, and oxygen uptake kinetics in chronic
heart failure. J Am Coll Cardiol. 2004;44(12):23392348.
27. Borlaug BA, Olson TP, Lam CS, et al. Global cardiovascular reserve dysfunction in heart failure with
preserved ejection fraction. J Am Coll Cardiol. 2010;
56(11):845-854.
28. Bishu K, Deswal A, Chen HH, et al. Biomarkers
in acutely decompensated heart failure with preserved or reduced ejection fraction. Am Heart J. 2012;
164(5):763-770 e763.
29. Tapia E, Sanchez-Lozada LG, Soto V, et al. Sildenafil
treatment prevents glomerular hypertension and hyperfiltration in rats with renal ablation. Kidney Blood
Press Res. 2012;35(4):273-280.
30. Medeiros PJ, Villarim Neto A, Lima FP, Azevedo
IM, Leao LR, Medeiros AC. Effect of sildenafil in re˜
nal ischemia/reperfusion injury in rats. Acta Cir Bras.
2010;25(6):490-495.
31. Choi DE, Jeong JY, Lim BJ, et al. Pretreatment of
sildenafil attenuates ischemia-reperfusion renal injury in rats. Am J Physiol Renal Physiol. 2009;
297(2):F362-F370.

JAMA, March 27, 2013—Vol 309, No. 12 1277

JAMA-PDE5i in heart failure

More Related Content

What's hot (20)

Similar to JAMA-PDE5i in heart failure (20)

More from Institute for the Study of Urologic Diseases (20)

Recently uploaded (20)

JAMA-PDE5i in heart failure