NSAIDs against post-ERCP pancreatitis – any doubts ?

NSAIDs against post-ERCP pancreatitis – any doubts ?

Thomas Rösch, Hamburg

Gastroenterology 2016 Apr;150(4):911-7; Epub 2016 Jan 9. Lancet 2016 Jun 4; 387(10035):2293-301. . Epub 2016 Apr 28.

CLINICAL-PANCREAS Rectal Indomethacin Does Not Prevent Post-ERCP Pancreatitis in Consecutive Patients
John M. Levenick,1,2 Stuart R. Gordon,1 Linda L. Fadden,1 L. Campbell Levy, 1 Matthew J. Rockacy,1 Sarah M. Hyder,1 Brian E. Lacy,1 Steven P. Bensen,1 Douglas D. Parr,3 and Timothy B. Gardner1
1Section of Gastroenterology and Hepatology, 3Investigational Pharmacy, Dart-mouth-Hitchcock Medical Center, Lebanon, New Hampshire; 2Section of Gastroentero-logy and Hepatology, Penn State Hershey Medical Center, Hershey, Pennsylvania

This article has an accompanying continuing medical education activity on page e19. Learning Objective: Upon completion of this test, successful learners will be able to: (1) define post-ERCP pancraetitis; (2) discuss the efficacy of rectal indomethacin to prevent post-ERCP pancre-atitis; (3) identify risk of UGI bleeding among ERCP patients given rectal indomethacin.

Backround & Aims

Rectal indomethacin, a nonsteroidal anti-inflammatory drug, is given to prevent pancreatitis in high risk patients undergoing endoscopic retrograde cholangiopancreatography (ERCP), based on findings from clinical trials. The European Society for Gastrointestinal Endoscopy guidelines recently recommended prophylactic rectal indomethacin for all patients undergoing ERCP, including those at average risk for pancreatitis. We performed a randomized controlled trail to investigate the efficacy of this approach.


We performed a prospective, double-blind, placebo controlled trial of 449 consecutive patients undergoing ERCP at Dartmouth Hitchcock Medical Center, from March 2013 through December 2014. Approximately 70% of the cohort were at average risk for PEP. Subjects were assigned randomly to groups given either a single 100-mg dose of rectal indomethacin (n=223) or a placebo suppository (n=226) during the procedure. The primary outcome was the development of post-ERCP pancreatitis (PEP), defined by new upper-abdominal pain, a lipase level more than 3-fold the upper limit of normal, and hospitalization after ERCP for 2 consecutive nights.


There were no differences between the groups in baseline clinical or procedural characteristics. Sixteen patients in the indomethacin group (7.2%) and 11 in the placebo group (4.9%) developed PEP (P 0.33). Complications and the severity of PEP were similar between groups. Per a priori protocol guidelines, the study was stopped owing to futility.


In a randomized controlled study of consecutive patients undergoing ERCP, rectal indomethacin did not prevent post-ERCP pancreatitis.
ClincialTrials.gov no: NCT01774604.

ClincialTrials.gov no: NCT01774604.

Routine pre-procedural rectal indometacin versus selctive post-procedural rectal indometacin to prevent pancreatitis in patients undergoing endoscopic retrograde cholangiopancreatography: a multicentre, single-blinded, randomised controlled trial
Hui Luo, Lina Zhao, Joseph Leung, Rongchun Zhang, Zhiguo Liu, Xiangping Wang, Biaoluo Wang, Zhanguo Nie, Ting Lei, Xun Li, Wence Zhou, Lingen Zhang, Qi Wang, Ming Li, Yi Zhou, Qian Liu, Hao Sun, Zheng Wang, Shuhui Liang, Xiaoyang Guo, Qin Tao, Kaichun Wu, Yanglin Pan, Xuegang Guo, Daiming Fan



Rectal indometacin decreases the occurrence of pancreatitis after endoscopic retrograde cholangiopancreatography (ERCP). However, the population most at risk and the optimal timing of administration require further investigation. We aimed to assess whether pre-procedural administration of rectal indometacin in all patients is more eff ective than post-procedural use in only high-risk patients to prevent post-ERCP pancreatitis.


We did a multicentre, single-blinded, randomised controlled trial at six centres in China. Eligible patients with native papilla undergoing ERCP were randomly assigned in a 1:1 ratio (with a computer-generated list) to universal pre-procedural indometacin or post-procedural indometacin in only high-risk patients, with stratification by trial centres and block size of ten. In the universal indometacin group, all patients received a single dose (100 mg) of rectal indometacin within 30 min before ERCP. In the risk-stratified, post-procedural indometacin group, only patients at predicted high risk received rectal indometacin, immediately after ERCP. Investigators, but not patients, were masked to group allocation. The primary outcome was overall ocurrence of post-ERCP pancreatitis. The analysis followed the intention-to-treat principle. This study was registered with ClinicalTrials.gov, number NCT02002650.


Between Dec 15, 2013, and Sept 21, 2015, 2600 patients were randomly assigned to universal, preprocedural indometacin (n=1297) or risk-stratified, post-procedural indometacin (n=1303). Overall, post-ERCP pancreatitis occurred in 47 (4%) of 1297 patients assigned to universal indometacin and 100 (8%) of 1303 patients assigned to risk-stratifi ed indometacin (relative risk 0·47; 95% CI 0·34–0·66; p<0·0001). Post-ERCP pancreatitis occurred in 18 (6%) of 305 high-risk patients in the universal group and 35 (12%) of 281 high-risk patients in the risk-stratifi ed group (p=0·0057). Post-ERCP pancreatitis was also less frequent in average-risk patients in the universal group (3% [29/992]), in which they received indometacin, than in the risk-stratifi ed group (6% [65/1022]), in which they did not receive the drug (p=0·0003). Other than pancreatitis, adverse events occurred in 41 (3%; two severe) patients in the universal indometacin group and 48 (4%; one severe) patients in the risk-stratified group. The most common adverse events were biliary infection (22 [2%] patients vs 33 [3%] patients) and gastrointestinal bleeding (13 [1%] vs ten [1%]).


Compared with a risk-stratified, post-procedural strategy, pre-procedural administration of rectal indometacin in unselected patients reduced the overall occurrence of post-ERCP pancreatitis without increasing risk of bleeding. Our results favour the routine use of rectal indometacin in patients without contraindications before ERCP.

What you should know about these papers

The application of non-steroidal anti-inflammatory drugs (NSAIDs) such as indometacin or diclofenac to prevent post-ERCP pancreatitis (PEP) has been examined in several randomized trials throughout the world and shown to be beneficial in almost all of them. As often the case, there are more meta analyses as there are original papers/randomized trials, namely 20 (!) meta analyses in a recent Pub med search with the search terms ERCP AND pancreatitis AND NSAID AND meta analysis: The three most recent of them included 91, 132, and 113 randomized trials. They reported a superiority of NSAID – risk reductions with ORs of 0.24 (diclofenac) or 0.59 (indometacin)1, overall 0.52 for both2, or 0.59 again for both substances3. The latter meta analysis is especially remarkable, it shows no effect for indometacin, none for men, and – none for sphincter of Oddi dyskinesia3, the primary example of a high-risk ERCP, with which everything started and large benefits for NSAIDs – notably with indometacin – had been shown in a 2012 NEJM publication4.

So, despite this plethora of meta analyses, at a closer look, the studies are not so easy to compare and several have significant limitations. Whether NSAIDs are beneficial in high-risk cases or all ERCP cases, is still somewhat unclear. To start with, the first question is whether indometacin and diclofenac are fully comparable; indometacin reaches maximal plasma levels 30-120 min after rectal administration, with a plasma elimination half life of 2 h. Diclofenac reaches maximal plasma levels 30 min after rectal application, plasma elimination half life is the same. Thus, they seem to be similar. An i.m. application of diclofenac leads to an earlier plasma peak (10-20 min). The timing of application might therefore be crucial; all studies on pharmacological treatment of (not ERCP induced) pancreatitis by drugs yielded negative results5, so the decisive fact here – in the prevention of pancreatitis – might be early administration before the attack on the pancreas (here ERCP), and the principle the earlier the better should apply. However, we do not know whether a drug applied rectally 30 min before ERCP is as effective as the same suppository given after ERCP: With the variation of resorption of indometacin, a pre-ERCP application may lead to a maximum plasma level during ERCP; while – with slow resorption, the suppository administered after ERCP may arrive there 2 h later, when the damage has already been done. Let´s look at more recent studies.

The first study comes from the US and planned to enroll some 1400 cases in trying to reduce the post-ERCP pancreatitis rate from 5% to 2.5% in unselected cases; a planned interim analysis after 25% of cases enrolled (n=449) showed no difference in PEP rates and the study was therefore terminated. This may give rise to some criticism, but we have to assume that study and journal statistitians have checked this carefully enough. Another major issue with this study is however, that about 30% of cases with prior ERCP and sphincterotomy were included in both groups, a patient subgroup with known low risk of post-ERCP pancreatitis at the repeated attempt. A subgroup analysis of different high risk cases is tried in the paper, but case number (with 27 pancreatitis cases overall) is by far not large enough here6. However, this is not the only negative study on NSAIDs recently published. Another study from Japan randomizing 407 primary ERCPs to rectal administration of 100 mg diclofenac at unclear times before or during ERCP (no details given), found no difference in pancreatitis rates (9.8% vs 9.4%)7. A study from Korea applying diclofenac by the i.m. route – the fastest way to maximal plasma levels (see above) – again failed to show a difference in 343 patients8.

The other trial comes from China and is huge. Case number calculation used exactly the same assumptions as in the US trial but arrives at double the case number (n=2600), just the power was 90% instead of the 80% in the US trial. However, the study used a different design, namely comparing routine use in all (primary) ERCPs versus use only in high-risk cases; here pancreatic rates were significantly reduced in the group with indometacin use in all cases. The study is already based on the assumption that NSAIDs are beneficial and the only question left is whether it should be used in all or only in high risk cases. The overall pancreatitis rates were 4% in the universal versus 8% in the selective group9.

This Chinese trial seemed to impress the Lancet reviewers so much by the mere case number that they might have overlooked that both groups were fundamentally different: The group in which indometacin was applied in all cases had the suppository given 30 min prior to ERCPs, whereas in the so-called high-risk group, this decision was often made during ERCP and the suppository administered after ERCP. Even more remarkable, when analyzing the high risk patients in both groups, those in the universal group, where all the patients received indometacin, also did better (PEP rates 6%) than their selective counterparts in the control group (12%), a fact which is very difficult to explain. Despite limited case numbers in these subgroups (n=305 vs n=281), the difference was statistically significant (p=0.0057). This could mean that earlier administration of rectal indomethacin is superior than later in high risk groups and underlines the inequity of the study groups.

Thus, despite all the meta analyses comparing apples and oranges, or more likely, apples and shoes, we may still be uncertain whether, and if so, how and when we should use NSAIDs in our daily ERCP routine. And – meta analyses are often not the solution, but rather the problem itself. More common sense would surely be appropriate when looking at these studies.


  1. Vadala di Prampero SF, Faleschini G, Panic N, et al. Endoscopic and pharmacological treatment for prophylaxis against postendoscopic retrograde cholangiopancreatography pancreatitis: a meta-analysis and systematic review. Eur J Gastroenterol Hepatol 2016.
  2. Sajid MS, Khawaja AH, Sayegh M, et al. Systematic review and meta-analysis on the prophylactic role of non-steroidal anti-inflammatory drugs to prevent post-endoscopic retrograde cholangiopancreatoraphy pancreatitis. World J Gastrointest Endosc 2015;7:1341-9.
  3. Rustagi T, Njei B. Factors Affecting the Efficacy of Nonsteroidal Anti-inflammatory Drugs in Preventing Post-Endoscopic Retrograde Cholangiopancreatography Pancreatitis: A Systematic Review and Meta- analysis. Pancreas 2015;44:859-67.
  4. Elmunzer BJ, Scheiman JM, Lehman GA, et al. A randomized trial of rectal indomethacin to prevent post-ERCP pancreatitis. N Engl J Med 2012;366:1414-22.
  5. Talukdar R, Vege SS. Acute pancreatitis. Curr Opin Gastroenterol 2015;31:374-9.
  6. Levenick JM, Gordon SR, Fadden LL, et al. Rectal Indomethacin Does Not Prevent Post-ERCP Pancreatitis in Consecutive Patients. Gastroenterology 2016;150:911-7; quiz e19.
  7. Ishiwatari H, Urata T, Yasuda I, et al. No Benefit of Oral Diclofenac on Post-Endoscopic Retrograde Cholangiopancreatography Pancreatitis. Dig Dis Sci 2016.
  8. Park SW, Chung MJ, Oh TG, et al. Intramuscular diclofenac for the prevention of post-ERCP pancreatitis: a randomized trial. Endoscopy 2015;47:33-9.
  9. Luo H, Zhao L, Leung J, et al. Routine pre-procedural rectal indometacin versus selective post-procedural rectal indometacin to prevent pancreatitis in patients undergoing endoscopic retrograde cholangiopancreatography: a multicentre, single-blinded, randomised controlled trial. Lancet 2016;387:2293-30

Related Posts

Scroll to Top