Articles

Same-admission versus interval cholecystectomy for mild

gallstone pancreatitis (PONCHO): a multicentre randomised
controlled trial
David W da Costa*, Stefan A Bouwense*, Nicolien J Schepers, Marc G Besselink, Hjalmar C van Santvoort, Sandra van Brunschot, Olaf J Bakker,
Thomas L Bollen, Cornelis H Dejong, Harry van Goor, Marja A Boermeester, Marco J Bruno, Casper H van Eijck, Robin Timmer, Bas L Weusten,
Esther C Consten, Menno A Brink, B W Marcel Spanier, Ernst Jan Spillenaar Bilgen, Vincent B Nieuwenhuijs, H Sijbrand Hofker, Camiel Rosman,
Annet M Voorburg, Koop Bosscha, Peter van Duijvendijk, Jos J Gerritsen, Joos Heisterkamp, Ignace H de Hingh, Ben J Witteman, Philip M Kruyt,
Joris J Scheepers, I Quintus Molenaar, Alexander F Schaapherder, Eric R Manusama, Laurens A van der Waaij, Jacco van Unen, Marcel G Dijkgraaf,
Bert van Ramshorst, Hein G Gooszen, Djamila Boerma, for the Dutch Pancreatitis Study Group

Summary
Background In patients with mild gallstone pancreatitis, cholecystectomy during the same hospital admission might Lancet 2015; 386: 1261–68
reduce the risk of recurrent gallstone-related complications, compared with the more commonly used strategy of See Editorial page 1212
interval cholecystectomy. However, evidence to support same-admission cholecystectomy is poor, and concerns exist See Comment page 1218
about an increased risk of cholecystectomy-related complications with this approach. In this study, we aimed to *Joint first authors
compare same-admission and interval cholecystectomy, with the hypothesis that same-admission cholecystectomy Department of Surgery
would reduce the risk of recurrent gallstone-related complications without increasing the difficulty of surgery. (D W da Costa MD,
B van Ramshorst PhD,
D Boerma PhD), Department of
Methods For this multicentre, parallel-group, assessor-masked, randomised controlled superiority trial, inpatients Gastroenterology
recovering from mild gallstone pancreatitis at 23 hospitals in the Netherlands (with hospital discharge foreseen within (N J Schepers MD,
48 h) were assessed for eligibility. Adult patients (aged ≥18 years) were eligible for randomisation if they had a serum Prof B L Wuesten PhD,
C-reactive protein concentration less than 100 mg/L, no need for opioid analgesics, and could tolerate a normal oral R Timmer PhD), and Department
of Radiology (T L Bollen MD), St
diet. Patients with American Society of Anesthesiologists (ASA) class III physical status who were older than 75 years Antonius Hospital, Nieuwegein,
of age, all ASA class IV patients, those with chronic pancreatitis, and those with ongoing alcohol misuse were excluded. Netherlands; Department of
A central study coordinator randomly assigned eligible patients (1:1) by computer-based randomisation, with varying Operating Room/Evidence-
block sizes of two and four patients, to cholecystectomy within 3 days of randomisation (same-admission Based Surgery
(S A Bouwense MD,
cholecystectomy) or to discharge and cholecystectomy 25–30 days after randomisation (interval cholecystectomy). Prof H G Gooszen PhD) and
Randomisation was stratified by centre and by whether or not endoscopic sphincterotomy had been done. Neither Department of Surgery
investigators nor participants were masked to group assignment. The primary endpoint was a composite of readmission (Prof H van Goor PhD), Radboud
for recurrent gallstone-related complications (pancreatitis, cholangitis, cholecystitis, choledocholithiasis needing University Medical Center,
Nijmegen, Netherlands;
endoscopic intervention, or gallstone colic) or mortality within 6 months after randomisation, analysed by intention to Department of
treat. The trial was designed to reduce the incidence of the primary endpoint from 8% in the interval group to 1% in Gastroenterology and
the same-admission group. Safety endpoints included bile duct leakage and other complications necessitating Hepatology (N J Schepers,
Prof M J Bruno PhD) and
re-intervention. This trial is registered with Current Controlled Trials, number ISRCTN72764151, and is complete.
Department of Surgery
(Prof C H van Eijck PhD), Erasmus
Findings Between Dec 22, 2010, and Aug 19, 2013, 266 inpatients from 23 hospitals in the Netherlands were randomly Medical Center, Rotterdam,
assigned to interval cholecystectomy (n=137) or same-admission cholecystectomy (n=129). One patient from each Netherlands; Department of
Surgery (M G Besselink PhD,
group was excluded from the final analyses, because of an incorrect diagnosis of pancreatitis in one patient (in the
H C van Santvoort PhD,
interval group) and discontinued follow-up in the other (in the same-admission group). The primary endpoint occurred Prof M A Boermeester PhD),
in 23 (17%) of 136 patients in the interval group and in six (5%) of 128 patients in the same-admission group (risk ratio Department of
0·28, 95% CI 0·12–0·66; p=0·002). Safety endpoints occurred in four patients: one case of bile duct leakage and Gastroenterology
(S van Brunschot MD), and the
one case of postoperative bleeding in each group. All of these were serious adverse events and were judged to be
Clinical Research Unit
treatment related, but none led to death. (M G Dijkgraaf PhD), Academic
Medical Center, Amsterdam,
Interpretation Compared with interval cholecystectomy, same-admission cholecystectomy reduced the rate of recurrent Netherlands; Department of
Surgery, Elisabeth Hospital,
gallstone-related complications in patients with mild gallstone pancreatitis, with a very low risk of cholecystectomy-
Tilburg, Netherlands
related complications. (J Heisterkamp PhD); Department
of Surgery, University Medical
Funding Dutch Digestive Disease Foundation. Center Utrecht, Netherlands
(O J Bakker PhD,
I Q Molenaar PhD); Department
Introduction Cholecystectomy is indicated in these patients to reduce of Surgery, Maastricht
Acute pancreatitis is a common gastrointestinal the risk of recurrent gallstone-related complications University Medical Center,
disorder, mostly caused by gallstones or biliary sludge.1,2 such as pancreatitis, cholecystitis, cholangitis, or Maastricht, Netherlands
(Prof C H Dejong PhD);
Around 80% of affected patients have mild pancreatitis.3 gallstone colics.4,5

[Link] Vol 386 September 26, 2015 1261

Downloaded for Ramesh Ardhanari (drrameshardhanari@[Link]) at Bangalore Baptist Hospital from [Link] by Elsevier on November 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
 Articles

Department of Surgery
(E C Consten PhD) and Research in context
Department of
Gastroenterology Evidence before this study biliary pancreatitis: same-admission cholecystectomy, as
(M A Brink PhD), Meander Our study group published a systematic review regarding the recommended by existing guidelines, and interval
Medical Center, Amersfoort, timing of cholecystectomy after mild gallstone pancreatitis in cholecystectomy, which is the more widely used strategy
Netherlands; Department of
Gastroenterology
2012 (van Baal et al). When we were preparing this systematic according to international audits. Our results show that,
(B W M Spanier PhD) and review, we searched the MEDLINE, Embase, and Cochrane compared with interval cholecystectomy, same-admission
Department of Surgery databases for studies published between January, 1992, and cholecystectomy reduces the risk of readmission for recurrent
(E J Spillenaar Bilgen PhD), July, 2010, using the search terms “cholecystectomy” and gallstone-related complications. This strategy also averted
Rijnstate Hospital, Arnhem,
Netherlands; Department of
“pancreatitis”. Since publication of this systematic review, we painful gallstone colics that occurred in more than half of all
Surgery, University Medical continued this search strategy in the MEDLINE and Cochrane patients in the interval cholecystectomy group. Hardly any
Center Groningen, Netherlands databases up until April 1, 2015. We identified one randomised cholecystectomy-related complications occurred in
(V B Nieuwenhuijs PhD, trial on the topic, in which cholecystectomy within 48 h after either group.
H S Hofker PhD); Department of
Surgery, Canisius-Wilhelmina
admission (immediate cholecystectomy) was compared with
Implications of all the available evidence
Hospital, Nijmegen, cholecystectomy before discharge (same-admission
Our study provides the first high-quality evidence for
Netherlands (C Rosman PhD); cholecystectomy) in patients with mild gallstone pancreatitis
Department of same-admission cholecystectomy. Patients admitted to the
(Aboulian et al, 2010). Immediate cholecystectomy reduced
Gastroenterology, hospital with mild gallstone pancreatitis should undergo
Diakonessenhuis, Utrecht,
median length of hospital stay by 1 day, prompting the
cholecystectomy before discharge, unless their medical history
Netherlands (A M Voorburg MD); investigators to terminate the study after an interim analysis
precludes this strategy. As shown in our study, same-admission
Department of Surgery, Jeroen halfway through the trial. However, the study was not aimed at
Bosch Hospital, cholecystectomy significantly reduces the number of
nor powered for differences in clinically relevant outcomes such
‘s-Hertogenbosch, Netherlands readmissions for recurrent gallstone disease, as compared with
(K Bosscha PhD); Department of as recurrence rates or surgical complications. Moreover, the
interval cholecystectomy. Neither our study nor any of the
Surgery, Gelre Hospital, strategy of immediate cholecystectomy is controversial, since a
previous studies found through our systematic search have
Apeldoorn, Netherlands substantial proportion of patients might still develop severe
(P van Duijvendijk PhD); shown increased difficulty or surgery-related complications in
pancreatitis during this phase, which is a generally
Department of Surgery, patients undergoing same-admission cholecystectomy for mild
Medisch Spectrum Twente, acknowledged contraindication for early surgery.
gallstone pancreatitis. In conclusion, same-admission
Enschede, Netherlands
(J J Gerritsen PhD); Department
Added value of this study cholecystectomy reduces morbidity without any evidence for
of Surgery, Catharina Hospital, This trial was designed to compare the two most common impaired safety and should therefore be considered the
Eindhoven, Netherlands strategies for long-term risk management in patients with mild optimum strategy in patients with mild pancreatitis.
(I H de Hingh PhD); Department
of Gastroenterology
(Prof B J Witteman PhD) and
Department of Surgery Several nationwide audits from both Europe and the guidelines from both the International Association of
(P M Kruyt MD), Gelderse Vallei
Hospital, Ede, Netherlands;
USA have shown that laparoscopic cholecystectomy is Pancreatology–American Pancreatic Association and the
Department of Surgery, Reinier usually done around 6 weeks after discharge from American Gastroenterology Association recommend that
de Graaf Hospital, Delft, hospital for mild gallstone pancreatitis.6–11 Recent studies cholecystectomy is done during the same hospital
Netherlands (J J Scheepers PhD); from the UK have reported that up to a third of all patients admission.4,5 However, no randomised studies have
Department of Surgery, Leiden
University Medical Center,
do not receive any definitive treatment within 1 year after compared same-admission cholecystectomy to the
Leiden, Netherlands discharge from hospital.9,12 This finding conflicts with the existing practice of interval cholecystectomy.16 This
(A F Schaapherder PhD); recommendation of cholecystectomy during the same absence of high-quality evidence might also contribute to
Department of Surgery, Medical admission or at least within 2 weeks after discharge, as the reported low adherence to guidelines.7–9,12,18
Center Leeuwarden,
Netherlands
proposed by the British Society of Gastroenterology.13 The We did a nationwide randomised study to investigate
(E R Manusama PhD); main reason for this delay in cholecystectomy is a whether or not same-admission cholecystectomy, as
Department of perceived danger of perioperative complications in early compared with interval cholecystectomy, reduces
Gastroenterology, Martini cholecystectomy after acute pancreatitis.7,14 Inflammation recurrent gallstone-related complications in patients with
Hospital, Groningen,
Netherlands
and oedema are believed to distort biliary tract anatomy, mild gallstone pancreatitis.
(L A van der Waaij PhD); and thereby complicating dissection with an increased risk
Department of Surgery, of conversion to open cholecystectomy and surgical Methods
Laurentius Hospital, Roermond, complications such as bile duct injury.12,15 A delayed Study design and participants
Netherlands (J van Unen MD)
approach also helps surgical scheduling, since emergency The PONCHO (Pancreatitis of biliary origin: Optimal
Correspondence to:
Dr Djamila Boerma, Dutch
theatre capacity is often scarce.14 timiNg of CHOlecystectomy) study was designed as
Pancreatitis Study Group, The drawback of cholecystectomy being postponed a randomised, controlled, parallel-group, superiority
St Antonius Hospital, until several weeks after discharge is that during this multicentre trial. The rationale and design of the
PO Box 2500, 3430 EM period patients are at risk of developing recurrent PONCHO trial have previously been described in
Nieuwegein, Netherlands,
[Link]@
gallstone-related complications. For example, recurrent detail.19 The study was done at 23 study sites in the
[Link] pancreatitis reportedly occurs in up to 33% of patients in Netherlands, including seven university medical centres
See Online for appendix observational studies.16,17 As a result, the recently revised and 16 teaching hospitals (appendix p 2).

1262 [Link] Vol 386 September 26, 2015

Downloaded for Ramesh Ardhanari (drrameshardhanari@[Link]) at Bangalore Baptist Hospital from [Link] by Elsevier on November 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
 Articles

All adult patients (aged ≥18 years) admitted to these Procedures

centres between Dec 7, 2010, and Aug 14, 2013, diagnosed In the interval cholecystectomy group, patients
with a first episode of gallstone pancreatitis were assessed were discharged and cholecystectomy was electively
for eligibility. The diagnosis of pancreatitis needed at least scheduled 25–30 days after randomisation. This time
two of the following three features: epigastric pain, serum interval is in line with the maximum waiting period
amylase or lipase levels at least three-times the upper recommended by the American and Dutch treatment
limit of normal, and, if done, characteristic findings of guidelines at the time of the design of the trial.22,23
acute pancreatitis on cross-sectional abdominal imaging. Same-admission cholecystectomy was done within
Mild pancreatitis was defined by absence of persistent 3 days after randomisation. All cholecystectomies were
organ failure (ie, >48 h), and local complications such as done by, or under the direct supervision of, a surgeon
pancreatic necrosis or peripancreatic fluid collections on who had undertaken at least 100 cholecystectomies in
CT.20 A biliary cause was defined by gallstones, biliary the past 5 years. Intraoperative cholangiography was
sludge, or a dilated common bile duct on imaging, or not mandatory because only about 3% of Dutch
based on biochemical signs of cholestasis (for details, see surgeons routinely do this procedure.24 The strategy
appendix pp 3, 8). of preoperative stone extraction through endoscopic
Patients were enrolled by the local physicians at retrograde cholangiopancreatography is much more
each hospital and were randomly assigned to the prevalent than intraoperative cholangiography because
two treatment groups once discharge from hospital was of the excellent widespread availability of this procedure
foreseen within 48 h. Additional eligibility criteria were in the Netherlands.
a serum C-reactive protein (CRP) concentration less Data were collected on case record forms by the local
than 100 mg/L, no need for opioid analgesics, and physicians in the 23 participating study sites. All data
tolerance of a normal oral diet, all at the time of for primary and secondary endpoints were checked for
randomisation. Patients with American Society of completeness by the study coordinators with source data
Anesthesiologists (ASA) class III physical status who at each participating centre.
were older than 75 years of age and all ASA class IV Patients were instructed to record all episodes of
patients (ie, irrespective of age) were excluded because gallstone colic (ie, irrespective of readmission) that
of their inherently high risk of complications from occurred during the 6-month follow-up period in a study
anaesthesia or surgery.21 Other exclusion criteria were diary, with reminders via telephone calls from the study
chronic pancreatitis and ongoing alcohol misuse. After research nurse (see appendix pp 5, 6, and 10 for details).
initiation of the trial, pregnancy was added as an The central study coordinators (SAB and DWdC) drafted
exclusion criterion in January, 2012, both for ethical reports for all potential primary and safety endpoints
reasons and because of the paucity of evidence about using the primary clinical and biochemical data as
cholecystectomy in this subgroup. collected by the study nurse. An adjudication committee
The study was investigator initiated and was undertaken of five gastrointestinal surgeons (DB, MGB, HCvS, HvG,
following the principles of the Declaration of Helsinki and CHD) who were masked to treatment allocation then
(originally adopted in 1964, with the last amendment individually assessed primary and safety endpoints using
before this trial in October, 2008) and the Dutch Medical all available data. Any disagreements were resolved in a
Research Involving Human Subjects Act (1998; last consensus meeting.
revised in 2006). The central committee for research in
humans in Nijmegen, the Netherlands (CMO), approved Outcomes
the study protocol. A data safety monitoring committee of The primary endpoint was a composite of gallstone-
four independent, non-participating physicians assessed related complications or mortality occurring within
all serious adverse events after inclusion of every 6 months after randomisation, before or after
50 patients in an unmasked fashion. All patients provided cholecystectomy, analysed by intention to treat. Gallstone-
written informed consent. related complications were defined as acute readmission
for recurrent pancreatitis, cholecystitis, cholangitis,
Randomisation and masking obstructive choledocholithiasis needing endoscopic
Randomisation was done by the central study coordinator retrograde cholangiopancreatography, or gallstone colic).25
using a web-based randomisation module. Randomisation Secondary endpoints were the individual components of
was stratified according to centre and by whether or not the primary endpoint, difficulty of cholecystectomy as
endoscopic sphincterotomy had been done. Computer- assessed by the most experienced surgeon on a 0–10 visual
generated permutated block randomisation with a 1:1 analogue scale, conversion to open cholecystectomy,
allocation ratio and concealed varying permuted block health-care use such as total length of hospital stay after
sizes of two and four patients was used. Owing to the randomisation (including readmission), and the number
invasive nature of the intervention and the logistics of patient-reported colics irrespective of readmission.
involved to do the procedures, neither the trial participants Predefined safety endpoints were cholecystectomy-
nor the investigators could be masked to group allocation. related complications such as bile duct injury and

[Link] Vol 386 September 26, 2015 1263

Downloaded for Ramesh Ardhanari (drrameshardhanari@[Link]) at Bangalore Baptist Hospital from [Link] by Elsevier on November 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
 Articles

bleeding; the need for additional surgical, endoscopic, or ≥75 years) and endoscopic sphincterotomy (yes vs no)
radiological intervention; and other complications such as before randomisation. We chose these subgroups
pneumonia, bacteraemia, and new-onset organ failure.26 because we postulated that elderly patients would be
The appendix provides definitions for the primary and more prone to complications (ie, both gallstone-related
secondary outcomes. and non-gallstone-related) than younger patients, and to
assess a potential protective effect of sphincterotomy on
Statistical analysis the occurrence of gallstone-related complications.16 We
The sample size calculation was based on an expected used logistic regression to test for interactions between
reduction of the primary endpoint from 8% within 4 weeks subgroups.
after discharge in the interval cholecystectomy group to 1% An interim analysis of the primary endpoint was done
in the same-admission cholecystectomy group, as reported by an independent statistician after 50% of the patients
in a recent nationwide retrospective study.6 To show this had completed the 6-month follow-up period, which used
effect with 80% power, a two-sided α level of 5%, and 0·5% the Peto approach with symmetric stopping boundaries at
loss to follow-up, 266 patients were needed. a p value of less than 0·001.27 A futility rule was not used,
An intention-to-treat-analysis was done. We tested since this study is the first randomised trial on this topic
differences in dichotomous data between the groups and we felt strongly that, irrespective of the outcome, the
using the χ² test or Fisher’s exact test (eg, the data for the results of the trial would be informative. The central study
primary outcome and need for intensive care unit coordinator and steering committee were informed that
admission), and we used the Mann-Whitney U test to the Peto criteria were not met and that the trial could
assess differences in continuous data (eg, length of stay continue as planned.
after randomisation and duration of surgery). Predefined For the final analyses, a two-sided p value of less than
subgroup analyses were done based on age (<75 years vs 0·05 was judged to be statistically significant. We did not
adjust p values for multiple testing.
IBM SPSS Statistics version 22 was used for statistical
713 patients with gallstone pancreatitis assessed for eligibility analyses.
This trial is registered with Current Controlled Trials,
number ISRCTN72764151.
447 excluded
291 did not meet inclusion criteria
183 pancreatic necrosis or fluid collection Role of the funding source
21 persisting organ failure
72 recurrent disease or previous The funder of the study had no role in the study design,
cholecystectomy execution, data analysis, or publication. The joint first
15 no informed consent possible authors (DdC and SAB), second author (NJS), statistical
61 met exclusion criteria
25 aged >75 years and ASA class III or and methodological expert (MGD), and last author (DB)
ASA class IV/V had full access to all the data. DB had final responsibility
23 CRP≥100 mg/L or opioid use at time
of discharge for the decision to submit for publication.
11 local or systemic complication
2 pregnancy
6 other reasons (various medical
Results
conditions) Between Dec 5, 2010, and Aug 19, 2013, 713 patients
89 eligible patients excluded with gallstone pancreatitis were assessed for eligibility
26 eligible patients missed
63 refused to participate (figure). After 447 patients were excluded, 266 eligible
patients were enrolled and randomly assigned: 137 to the
interval cholecystectomy group and 129 patients to the
266 patients randomly assigned same-admission cholecystectomy group. The masked
adjudication committee excluded one patient in the
interval group from the final analysis because of an
137 allocated to interval cholecystectomy 129 allocated to same-admission incorrect diagnosis of acute pancreatitis, since the serum
136 received allocated intervention cholecystectomy amylase level did not exceed three times the upper limit
1 refused surgery 129 received allocated intervention of normal. One patient in the same-admission group was
lost to follow-up at 3 months after randomisation. Before
1 excluded because of 1 lost to follow-up at
randomisation, 42 (31%) of 136 patients in the interval
incorrect diagnosis of 3 months group and 35 (27%) of 128 in the same-admission group
pancreatitis
had undergone endoscopic sphincterotomy (p=0·53). In
both groups, sphincterotomy was done a median of 1 day
136 patients included in analysis 128 patients included in analysis (IQR 0–2 days in interval group and 0–1 day in
same-admission group) after admission. The baseline
Figure: Trial profile characteristics of the participants did not differ
ASA=American Society of Anesthesiologists. CRP=C-reactive protein. significantly between the two treatment groups (table 1).

1264 [Link] Vol 386 September 26, 2015

Downloaded for Ramesh Ardhanari (drrameshardhanari@[Link]) at Bangalore Baptist Hospital from [Link] by Elsevier on November 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
 Articles

In the interval group, median time to cholecystectomy

Interval Same-admission
after randomisation was 27 days (IQR 26–29) and cholecystectomy cholecystectomy
100 (74%) of the 136 patients underwent surgery within (n=136) (n=128)
the designated 25–30 days. Recurrent gallstone-related Age, years 54 (41–68) 53 (38–66)
complications necessitated emergency or earlier Female sex 84 (62%) 76 (59%)
cholecystectomy in 13 (10%) patients. One patient in Body-mass index, kg/m2 28 (25–31) 27 (24–32)
the interval group ultimately refused cholecystectomy. Medical history
In the same-admission group, median time to Upper abdominal surgery 6 (4%) 8 (6%)
cholecystectomy following randomisation was 1 day History of gallstone colics 35 (26%) 38 (30%)
(IQR 1–2), and 119 (93%) of the 128 patients underwent
History of cholecystitis 2 (1%) 3 (2%)
surgery within the designated 3 days. The experience
Diabetes 7 (5%) 11 (9%)
of the surgeons doing the cholecystectomy procedure
American Society of Anaesthesiologists class
did not differ between groups (see appendix p 6 for
I: healthy status 51 (38%) 43 (34%)
further details).
II: mild systemic disease 74 (54%) 72 (56%)
The composite primary endpoint of acute readmission
III: severe systemic disease 11 (8%) 13 (10%)
for a gallstone-related complication or mortality occurred
Endoscopic sphincterotomy before randomisation 42 (31%) 35 (27%)
in 23 (17%) of 136 patients in the interval cholecystectomy
C-reactive protein level on the day of randomisation, mg/L 36 (15–69) 31 (11–66)
group, as compared with six (5%) of 128 patients in the
Days of admission before randomisation 5 (3–9) 5 (3–8)
same-admission cholecystectomy group (risk ratio [RR]
Days between randomisation and cholecystectomy 27 (26–29) 1 (1–2)
0·28 [95% CI 0·12–0·66], p=0·002; table 2). In the
interval group, 21 (91%) of the 23 primary endpoints Data are median (IQR) or n (%).
occurred before cholecystectomy, with a median time
Table 1: Baseline characteristics
from discharge to readmission of 15 days (IQR 8–21). In
the same-admission group, all six primary endpoint
events occurred after cholecystectomy and within the first same-admission group. No differences in the number
3 weeks after discharge (median 12 days [IQR 5–18]). of other complications that needed treatment were
Recurrent gallstone pancreatitis occurred in 12 (9%) recorded (table 2).
of 136 patients in the interval group compared with In a subgroup analysis, formal statistical tests showed no
three (2%) of 128 patients in the same-admission group interaction between the different subgroups and the effect
(RR 0·27 [95% CI 0·08–0·92], p=0·03). These of same-admission cholecystectomy in the occurrence of
15 patients were readmitted for a median of 6 days the primary endpoint (p>0·05 for all). In the subgroup of
(IQR 4–10) and did not develop pancreatic necrosis patients who had undergone endoscopic sphincterotomy,
or organ failure. In the same-admission group, a the primary endpoint occurred in seven (17%) of
75-year-old patient with a recent carotid endarterectomy 42 patients, compared with one (3%) of 35 in the
died at home 1 week after cholecystectomy because of same-admission group (p=0·07; appendix pp 12–13). In
ischaemic stroke. the interval group, one patient developed recurrent
In the interval group, 62 (51%) of 121 responding pancreatitis, two cholecystitis, one choledocholithiasis, and
patients reported one or more gallstone colics before three were readmitted for gallstone colic. One patient
cholecystectomy, irrespective of the need for re- in the same-admission group was readmitted for
admission, compared with three (3%) of 93 responding choledocholithiasis.
patients in the same-admission group (risk ratio 0·06
[95% CI 0·02–0·19], p<0·0001). In the interval group, Discussion
this symptom was reported as “severe pain” by 39 (63%) This study shows that in patients with mild gallstone
of 62 patients, compared with two (2%) of 93 in the pancreatitis, same-admission cholecystectomy reduces
same-admission group (appendix pp 5, 6, and 10). the risk of recurrent gallstone-related complications,
Length of hospital stay after randomisation did not including pancreatitis. The very low incidence of
differ between the groups (table 2). Difficulty of cholecystectomy-related complications suggests that
cholecystectomy, the number of conversions, or cholecystectomy can be done safely during the same
health-care use did not differ between the groups (details hospital admission.
about the secondary endpoints and cholecystectomies Several observational and mostly retrospective studies
are provided in appendix pp 6–7). also showed a reduced risk of gallstone-related compli-
In each group, one patient developed a cystic duct cations following same-admission cholecystectomy in
leakage, which was treated by endoscopic sphincterotomy mild gallstone pancreatitis.12,16,18 However, because of
in one patient and by percutaneous catheter drainage in their non-randomised design, these studies are prone to
the other. A haematoma was evacuated by percutaneous selection bias. For example, elderly patients, patients
drainage in one patient in the interval group, and with substantial comorbidity, or those with a severe
by laparoscopy drainage in another patient in the course of pancreatitis might have undergone interval

[Link] Vol 386 September 26, 2015 1265

Downloaded for Ramesh Ardhanari (drrameshardhanari@[Link]) at Bangalore Baptist Hospital from [Link] by Elsevier on November 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
 Articles

<100 mg/L) might have unnecessarily increased length

Interval Same-admission Risk ratio p value
cholecystectomy cholecystectomy (95% CI) of stay in some patients. Therefore, although our study
(n=136) (n=128) has shown the benefit of performing cholecystectomy
Primary endpoint before discharge, future studies should be directed at
Mortality or readmission for 23 (17%) 6 (5%) 0·28 (0·12–0·66) 0·002 exploring the optimum timing of cholecystectomy during
gallstone-related complications a hospital stay.
Secondary endpoints Although existing guidelines recommend conservative
Readmission for gallstone-related complications management in cases of mild gallstone pancreatitis
Recurrent pancreatitis 12 (9%) 3 (2%) 0·27 (0·08–0·92) 0·03 without cholangitis, quite a large percentage of patients
Cholecystitis 2 (2%) 0 0·50 in our study population underwent endoscopic
Choledocholithiasis needing 2 (2%) 1 (1%) 0·53 (0·05–5·79) 1·00 sphincterotomy. However, these rates are similar to those
ERCP reported in large, nationwide studies from the UK and
Gallstone colic 7 (5%) 2 (2%) 0·30 (0·06–1·43) 0·17 USA.7,12,18 In view of the protective effect of sphincterotomy
Mortality 0 1 (1%) 0·48 on the recurrence of pancreatitis, this might have
Patients reporting colics during 62 (51%) 3 (3%) 0·06 (0·02–0·19) <0·0001 moderated the contrast in primary endpoints between
waiting period* the groups in favour of interval cholecystectomy.18 More
Difficulty of cholecystectomy 6 (4–7) 6 (4–7) 0·70 importantly, our results showed that these patients
(assessed on a 0–10 visual
analogue scale)
remained at risk for recurrent gallstone-related
Conversion to open 4 (3%) 5 (4%) 1·31 (0·36–4·77) 0·74
complications even after sphincterotomy. This finding
cholecystectomy† differs from the results of previous retrospective studies
Operating time (min) 60 (44–78) 58 (44–70) 0·47 that suggested that patients after sphincterotomy do not
Total length of stay after 3 (2–5) 3 (2–4) 0·94 need to undergo early cholecystectomy.31 Although
randomisation (days) sphincterotomy might reduce the risk of recurrent
Need for intensive care unit 1 (1%) 1 (1%) 1·00 pancreatitis, it evidently does not provide adequate
admission protection from other events such as cholecystitis
Safety endpoints and colic to warrant interval cholecystectomy.10,17,18
Cystic duct leakage 1 (1%) 1 (1%) 1·00 The findings of our study are in line with a recent
Bleeding needing reoperation or 1 (1%) 1 (1%) 1·00 meta-analysis on prophylactic cholecystectomy after
transfusion
sphincterotomy for gallstone-related complications other
Need for additional intervention
than pancreatitis.32 Some investigators have advocated
Surgical 0 1 (1%) 0·48 the use of endoscopic sphincterotomy as a bridge
Endoscopic 0 1 (1%) 0·48 to cholecystectomy in patients with more severe
Radiological 2 (2%) 0 0·50 pancreatitis, complicated by local complications such
Pneumonia 0 2 (2%) 0·23 as parenchymal necrosis or peripancreatic fluid
Pulmonary embolism 1 (1%) 0 1·00 collections.31,33 This issue has not been addressed in
Data are n (%) or median (IQR) unless otherwise indicated. ERCP=endoscopic retrograde cholangiopancreatography.
prospective trials and needs further study.
*Interval cholecystectomy n=121, same-admission cholecystectomy n=93. †Four patients in the interval group and Although our study was not powered to detect
two in the same-admission group not included in analysis because of primary open cholecystectomy (so n=132 in significant differences in cholecystectomy-related comp-
interval group and n=125 in same-admission group).
lications (eg, bile duct leakage), the overall low incidence
Table 2: Primary, secondary, and safety endpoints of these complications challenges the notion that
cholecystectomy in the early phase after recovery from
cholecystectomy. Only one small randomised study has acute pancreatitis is not safe.15,30 This hypothesis is
been done on timing of cholecystectomy in patients with supported by the similar scores of surgical difficulty
mild gallstone pancreatitis.28 In this trial, patients were obtained between the same-admission and interval
randomly assigned to either cholecystectomy within 48 h groups. Studies on patients with other gallstone-related
and cholecystectomy 48 h or more after admission. The diseases such as cholecystitis or choledocholithiasis also
study was designed with length of hospital stay as the showed no differences in technical difficulty between
primary endpoint and was not powered to detect early and delayed cholecystectomy.34,35 Nevertheless,
differences in clinically relevant outcomes such as large, population-based studies might provide more
recurrent gallstone-related complications. Moreover, comprehensive data for a definitive appraisal of the
cholecystectomy within 48 h after admission in gallstone relative risk of surgical complications between same-
pancreatitis is controversial because patients can still admission and interval cholecystectomy.
develop pancreatic necrosis or organ failure during this Same-admission cholecystectomy has several benefits
phase of the disease, which both are viewed as contra- for both patients and health-care providers.
indications for early surgery.15,29,30 However, conversely First and foremost, the risk of readmission for recurrent
the randomisation criteria as applied in the present study pancreatitis and other gallstone-related complications is
(most notably a C-reactive protein concentration of minimised. Furthermore, same-admission cholecystectomy

1266 [Link] Vol 386 September 26, 2015

Downloaded for Ramesh Ardhanari (drrameshardhanari@[Link]) at Bangalore Baptist Hospital from [Link] by Elsevier on November 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
 Articles

prevents disabling colics that would otherwise have Declaration of interests

occurred in more than half of those patients awaiting MABo has received grants from Baxter, Ipsen, LifeCell, KCl,
Johnson & Johnson, and Abbot. MJB has received lectures and
elective surgery. An additional advantage is that both consultancy fees from Cook Medical and Boston Scientific. IHdH has
treatment and prevention of future gallstone-related received grants from Roche. NJS has received grants from Fonds
complications for acute pancreatitis are provided during a NutsOhra and ZonMw. HCvS received a career development grant from
single hospital stay. However, from the perspective of the Dutch Digestive Disease Foundation. All reported grants are outside
of the submitted work. All other authors declare no competing interests.
health-care providers, widespread implementation of this
strategy might be challenging, since it demands a shift Acknowledgments
This trial was supported by the Dutch Digestive Disease Foundation
from elective to acute care surgery, which will necessitate a (Maag Lever Darm Stichting), grant number WO 11-03. All authors
change in both the mindset towards the urgency of vouch for the accuracy and completeness of the data and analyses.
cholecystectomy in this particular patient group, and in References
logistics and infrastructure. In the setting of a randomised 1 Yadav D, Lowenfels AB. The epidemiology of pancreatitis and
trial, same-admission cholecystectomy did not prove an pancreatic cancer. Gastroenterology 2013; 144: 1252–61.
2 Venneman NG, van Brummelen SE, van Berge-Henegouwen GP,
obstacle for the 23 participating Dutch centres, but this van Erpecum KJ. Microlithiasis: an important cause of “idiopathic”
obviously does not guarantee worldwide implementation. acute pancreatitis? Ann Hepatol 2003; 2: 30–35.
However, several recent international studies have shown 3 Frossard JL, Steer ML, Pastor CM. Acute pancreatitis. Lancet 2008;
371: 143–52.
that quite straightforward organisational adjustments,
4 Working Group IAP/APA Acute Pancreatitis Guidelines. IAP/APA
such as direct admission to the surgical ward, can lead to evidence-based guidelines for the management of acute
improved efficiency in care for patients with gallstone pancreatitis. Pancreatology 2013; 13: e1–15.
pancreatitis.7,14,36 With respect to external validity, we should 5 Tenner S, Baillie J, Dewitt J, Vege SS. American College of
Gastroenterology Guidelines: management of acute pancreatitis.
note that our results cannot be extrapolated to patients Am J Gastroenterol 2013; 108: 1400–15.
older than 75 years of age with ASA class III or any patients 6 Bakker OJ, van Santvoort HC, Hagenaars JC, et al. Timing of
with a higher ASA classification. These patients are poor cholecystectomy after mild biliary pancreatitis. Br J Surg 2011;
98: 1446–54.
surgical candidates in whom the risk of perioperative
7 Nguyen GC, Boudreau H, Jagannath SB. Hospital volume as a
complications can outweigh the long-term protective effect predictor for undergoing cholecystectomy after admission for acute
of cholecystectomy, especially if endoscopic sphincterotomy biliary pancreatitis. Pancreas 2010; 39: e42–47.
has already been done. The optimum strategy in this 8 Lankisch PG, Weber-Dany B, Lerch MM. Clinical perspectives in
pancreatology: compliance with acute pancreatitis guidelines in
vulnerable patient group needs further investigation. Germany. Pancreatology 2005; 5: 591–93.
In conclusion, the results of this multicentre trial show 9 El-Dhuwaib Y, Deakin M, David GG, Durkin D, Corless DJ,
that same-admission cholecystectomy in patients with Slavin JP. Definitive management of gallstone pancreatitis in
England. Ann R Coll Surg Eng 2012; 94: 402–06.
mild gallstone pancreatitis was safe and reduced the risk 10 Sandzen B, Haapamaki MM, Nilsson E, Stenlund HC, Oman M.
of recurrent gallstone-related complications, as compared Cholecystectomy and sphincterotomy in patients with mild acute
with interval cholecystectomy. biliary pancreatitis in Sweden 1988–2003: a nationwide register
study. BMC Gastroenterol 2009; 9: 80.
Contributors 11 Pezzilli R, Uomo G, Gabbrielli A, et al. A prospective multicentre
DB supervised the study. SAB, MGB, SvB, OJB, HCvS, NJS, MABo, TLB, survey on the treatment of acute pancreatitis in Italy.
KB, MABr, MJB, ECC, CHD, PvD, CHvE, JJG, HvG, JH, IHdH, PMK, Dig Liver Dis 2007; 39: 838–46.
IQM, VBN, CR, AFS, JJS, BWMS, RT, BLW, BJW, BvR, HGG, and DB 12 Johnstone M, Marriott P, Royle TJ, et al. The impact of timing of
participated in the design of the study. DdC, SAB, NJS, SvB, TLB, MGB, cholecystectomy following gallstone pancreatitis. Surgeon 2014;
HCvS, OJB, MABo, KB, MABr, MJB, ECC, CHD, PvD, CHvE, JJG, HvG, 12: 134–40.
JH, IHdH, PMK, IQM, VBN, CR, AFS, JJS, BWMS, RT, BLW, BJW, 13 Working Party of the British Society of Gastroenterology,
AMV, EJSB, HSH, ERM, LAvdW, JvU, and DB collected the data. MGD Association of Surgeons of Great Britain and Ireland, Pancreatic
and DdC did the statistical analyses. DWdC, HCvS, SAB, MGB, NJS, Society of Great Britain and Ireland, Association of Upper GI
TLB, HGG, MGD, and DB interpreted the data. DWdC drafted the first Surgeons of Great Britain and Ireland. UK guidelines for the
and subsequent versions of the report. All authors read and appraised management of acute pancreatitis. Gut 2005; 54 (suppl 3): iii1–9.
the report and approved of the final version. 14 Monkhouse SJ, Court EL, Dash I, Coombs NJ. Two-week target for
laparoscopic cholecystectomy following gallstone pancreatitis is
Trial steering committee achievable and cost neutral. Br J Surg 2009; 96: 751–55.
St Antonius Hospital, Nieuwegein, Netherlands: D Boerma (chair), 15 Kelly TR, Wagner DS. Gallstone pancreatitis: a prospective
B van Ramshorst, R Timmer, and B L Weusten; Academic Medical Center randomized trial of the timing of surgery. Surgery 1988; 104: 600–05.
Amsterdam, Netherlands: M G Besselink, H C van Santvoort; Radboud 16 van Baal MC, Besselink MG, Bakker OJ, et al. Timing of
University Medical Center, Nijmegen, Netherlands: S A Bouwense, cholecystectomy after mild biliary pancreatitis: a systematic review.
H G Gooszen; University Medical Center, Utrecht, Netherlands: Ann Surg 2012; 255: 860–66.
O J Bakker; Leiden University Medical Center, Leiden, Netherlands: 17 Ito K, Ito H, Whang EE. Timing of cholecystectomy for biliary
A F Schaapherder; University Medical Center Groningen, Netherlands: pancreatitis: do the data support current guidelines?
V B Nieuwenhuijs; Gelderse Vallei Hospital, Ede, Netherlands: J Gastrointest Surg 2008; 12: 2164–70.
B J Witteman; Meander Medical Center, Amersfoort, Netherlands: 18 Hwang SS, Li BH, Haigh PI. Gallstone pancreatitis without
M A Brink; Erasmus Medical Center, Rotterdam, Netherlands: M J Bruno. cholecystectomy. JAMA Surg 2013; 148: 867–72.
19 Bouwense SA, Besselink MG, van Brunschot S, et al. Pancreatitis of
Adjudication committee
biliary origin, optimal timing of cholecystectomy (PONCHO trial):
St Antonius Hospital Nieuwegein: D Boerma; Academic Medical Center
study protocol for a randomized controlled trial. Trials 2012; 13: 225.
Amsterdam: M G Besselink, H C van Santvoort., Radboud University
20 Banks PA, Bollen TL, Dervenis C, et al. Classification of acute
Medical Center, Nijmegen: H van Goor; Maastricht University Medical
pancreatitis—2012: revision of the Atlanta classification and
Center: C H Dejong. definitions by international consensus. Gut 2013; 62: 102–11.

[Link] Vol 386 September 26, 2015 1267

Downloaded for Ramesh Ardhanari (drrameshardhanari@[Link]) at Bangalore Baptist Hospital from [Link] by Elsevier on November 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
 Articles

21 Giger UF, Michel JM, Opitz I, et al. Risk factors for perioperative 30 Nealon WH, Bawduniak J, Walser EM. Appropriate timing of
complications in patients undergoing laparoscopic cholecystectomy: cholecystectomy in patients who present with moderate to severe
analysis of 22,953 consecutive cases from the Swiss Association of gallstone-associated acute pancreatitis with peripancreatic fluid
Laparoscopic and Thoracoscopic Surgery database. J Am Coll Surg collections. Ann Surg 2004; 239: 741–49.
2006; 203: 723–28. 31 Heider TR, Brown A, Grimm IS, Behrns KE. Endoscopic
22 Forsmark CE, Baillie J. AGA Institute technical review on acute sphincterotomy permits interval laparoscopic cholecystectomy in
pancreatitis. Gastroenterology 2007; 132: 2022–44. patients with moderately severe gallstone pancreatitis.
23 Dutch Society of Internal Medicine. Dutch acute pancreatitis J Gastrointest Surg 2006; 10: 1–5.
guideline. 2005. [Link] 32 McAlister VC, Davenport E, Renouf E. Cholecystectomy deferral in
NyIcwbOy_VW0S2CSRAnVLQ/richtlijn_2005_Acute-Pancreatitis. patients with endoscopic sphincterotomy. Cochrane Database Syst Rev
pdf (accessed Aug 12, 2015). 2007; 4: CD006233.
24 Buddingh KT, Hofker HS, ten Cate Hoedemaker HO, van Dam GM, 33 Sanjay P, Yeeting S, Whigham C, Judson H, Polignano FM, Tait IS.
Ploeg RJ, Nieuwenhuijs VB. Safety measures during cholecystectomy: Endoscopic sphincterotomy and interval cholecystectomy are
results of a nationwide survey. World J Surg 2011; 35: 1235–41. reasonable alternatives to index cholecystectomy in severe acute
25 Mayumi T, Takada T, Kawarada Y, et al. Results of the Tokyo gallstone pancreatitis (GSP). Surg Endosc 2008; 22: 1832–37.
Consensus Meeting Tokyo Guidelines. J Hepatobiliary Pancreat Surg 34 Gurusamy KS, Davidson C, Gluud C, Davidson BR. Early versus
2007; 14: 114–21. delayed laparoscopic cholecystectomy for people with acute
26 Bergman JJ, van den Brink GR, Rauws EA, et al. Treatment of bile cholecystitis. Cochrane Database Syst Rev 2013; 6: CD005440.
duct lesions after laparoscopic cholecystectomy. Gut 1996; 38: 141–47. 35 Reinders JS, Goud A, Timmer R, et al. Early laparoscopic
27 Chan AW, Tetzlaff JM, Gotzsche PC, et al. SPIRIT 2013 explanation cholecystectomy improves outcomes after endoscopic
and elaboration: guidance for protocols of clinical trials. BMJ 2013; sphincterotomy for choledochocystolithiasis. Gastroenterology 2010;
346: e7586. 138: 2315–20.
28 Aboulian A, Chan T, Yaghoubian A, et al. Early cholecystectomy safely 36 Kulvatunyou N, Watt J, Friese RS, et al. Management of acute mild
decreases hospital stay in patients with mild gallstone pancreatitis: gallstone pancreatitis under acute care surgery: should patients be
a randomized prospective study. Ann Surg 2010; 251: 615–19. admitted to the surgery or medicine service? Am J Surg 2014;
29 Mofidi R, Duff MD, Wigmore SJ, Madhavan KK, Garden OJ, 208: 981–87.
Parks RW. Association between early systemic inflammatory
response, severity of multiorgan dysfunction and death in acute
pancreatitis. Br J Surg 2006; 93: 738–44.

1268 [Link] Vol 386 September 26, 2015

Downloaded for Ramesh Ardhanari (drrameshardhanari@[Link]) at Bangalore Baptist Hospital from [Link] by Elsevier on November 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.