Early endoscopic retrograde cholangiopancreatography (ERCP) has proven to be helpful in patients with acute biliary pancreatitis (ABP), especially when severe disease is predicted or combined with cholangitis.1, 2, 3, 4, 5, 6, 7 However, early ERCP remains to be controversial for other patients since improvement in prognosis has not always been observed when compared with conservative treatment. The rationale behind conservative treatment is that the stones that are small enough to obstruct the pancreatic duct tend to pass spontaneously and manipulations of ERCP itself might aggravate the papillary edema and do more harm to the vulnerable pancreas. However, the development of complications from conservative treatment is not negligible. Neoptolemos et al.1 and Fan et al.2 reported that 12% and 7% of their patients, respectively, experienced systemic and local complications, even when mild disease was predicted. Furthermore, because all of the patients with ABP presents with upper abdominal pain, and leukocytosis and jaundice are seen in most of them regardless of severity, concomitant cholangitis, which is diagnosed on the basis of Charcot's triad, cannot easily be excluded. In practice, rescue ERCP is frequently performed during conservative treatment for patients with predicted mild disease. In addition, more complications develop if the duration of ampullary obstruction by a sustained stone exceeds 48 hours.4 Therefore, if feasible, it is crucial to perform ERCP as early as possible, irrespective of severity, for patients with ABP.

Meanwhile, 1% to 7% of patients who receive ERCP in the absence of pancreatitis experience post-ERCP pancreatitis (PEP),8, 9, 10, 11, 12 and the risk may exceed 25% amongst the high risk populations.13 There are several potential procedure-related risk factors; among them, pancreatic injection, which frequently occurs during biliary cannulation and removal of stones, is considered to be a definitive one.9, 10, 14, 15, 16, 17 To overcome this risk factor, temporary pancreatic stenting after ERCP has been attempted, with success, to reduce the development of PEP; this method has been supported with concrete evidence.18, 19, 20, 21, 22, 23, 24 Moreover, its use for patients who are at high risk is recommended by the guidelines put forth by the European Society of Gastrointestinal Endoscopy.25 However, to the best of our knowledge, there has only a limited number studies evaluating the safety of leaving a pancreatic stent temporarily after ERCP in a patient who was already affected by acute pancreatitis. Therefore, this current study was designed to explore this safety and to determine its influences on clinical outcomes.

During the study period, ERCP was successfully performed in 587 patients with ABP and insertion of a pancreatic stent was attempted in 79 patients, in whom the pancreas was injected. Pancreatic stenting was adequate in 64 (81.0%) patients (the PS group); they were compared with 15 patients with unsuccessful―or inadequate―stenting (the no PS group). The two groups showed no meaningful differences in the baseline demographic and clinical characteristics (Table 1).

Table 1 Baseline Characteristics Click for larger image

1. Primary outcome measure

There were no significant differences in the overall complications and systemic inflammatory response syndrome (SIRS) between the two groups (Table 2). However, fewer local complications occurred in the PS group than in the no PS group (4.7% for PS vs. 20.0% for no PS; p=0.04) and the difference was the most remarkable in necrosis (1.6% for PS vs. 13.3% for no PS; p=0.03). There was no case of mortality in both groups (p=0.99).

Table 2 Primary Outcome of Interest Click for larger image

2. Secondary outcome measure

Serum amylase and lipase levels measured on the day after ERCP tended to decrease in both groups; no patient showed an increase in the levels after ERCP. There were no significant differences in time elapsed for patients to become eligible for cholecystectomy or discharge between the two groups. There was no case of perforation or significant bleeding (Table 3).

Table 3 Secondary Outcome of Interest Click for larger image

Although the benefit of early ERCP for patients with ABP other than predicted severe disease or concomitant cholangitis is debatable, the development of complications is not infrequent. Moreover, concurrent acute cholangitis is not excluded readily, and despite its absence at the time of diagnosis, rescue ERCP during conservative treatment is frequently performed due to the emergence of cholangitis not only in daily practices, but also in clinical trials. Neoptolemos et al.1 and Fölsch et al.3 reported 23% and 20% of patients from the conservative group, respectively, received rescue ERCP within 4 weeks from randomization, although the number of patients with predicted mild disease is not exactly specified. Remarkably, Fan et al. reported that 9 (15.6%) out of 58 patients with predicted mild disease from the conservative group received emergency ERCP with a median of 48 hours after admission.2 Among the 40 patients with predicted severe disease from the conservative group, 18 (45.0%) received emergency ERCP, with a median of 60 hours after admission. Because early ductal decompression is a mandatory part of the treatment of acute cholangitis, it is critical to perform ERCP as early as possible, if it can be performed relatively safely. Procedure-related complications, such as bleeding or perforation, were not reported to be increased by almost all of the relevant studies.1, 2, 3, 4, 28, 29, 30, 31 However, further injury by manipulations of ERCP itself is still a matter of concern. The exact mechanism by which ERCP insults the pancreas is not fully understood. However, a decrease in the flow of pancreatic juice due to papillary edema caused during manipulations is believed to be one of the most important causes.32 The papilla of a patient with ABP in whom ERCP is going to be performed is obstructed by a gallstone, in which point in time, the flow of pancreatic juice has already been disturbed. In such an unfavorable circumstance, even the very successful extraction of the stone with the finest manipulations always leaves a chance that might bring about further swelling of the papilla, which would surely impose an unbeneficial influence on ABP even after stone removal. We assumed that this can be overcome by temporary pancreatic stenting, which has been proven for preventing PEP.

There is still much to be elucidated about the use of a pancreatic stent in patients with acute pancreatitis, unlike patients with chronic pancreatitis in whom pancreatic stenting is most commonly used. Madácsy et al., in a case series, described the use of rescue pancreatic stenting to prevent the evolution of severe PEP.33 Fejes et al. reported that temporary pancreatic stenting can serve as a good bridging procedure in 27 patients with severe ABP in whom biliary EST was unsuccessful.34 However, the number patients in the study was small, and the lack of a control group limited the full understanding of the consequences of failed stenting.

In our study, it is admitted that patients in the no PS group might have a smaller and more tortuous pancreatic duct, making the stent insertion more difficult. Moreover, we cannot deny the possibility that a worse outcome in the no PS group resulted due to the failed attempts, as Freeman et al. suggested.35 To minimize possible insults, however, we limited the number of attempts for recannulation of the pancreatic duct to less than 5 times. More importantly, the development of overall complication of ‘unfavorable’ no PS group was not increased compared with the previous series. The complication rates of the early ERCP group of the prospective randomized trials were 10/59 (17%) by Neoptolemos et al.,1 17/97 (18%) by Fan et al.,2 and 58/126 (46%) by Fölsch et al.3 This implies that our new strategy, in which early ERCP is performed for all patients with ABP having radiologically-proven CBD stones regardless of severity, can be helpful if pancreatic stent can be inserted successfully after ERCP, as expected for most of them, with minimal harm even in failed cases. This is supported by our data, as no patient experienced paradoxical elevation of serum amylase and lipase after ERCP. By our strategy, unnecessary prolongation of hospital stays and increased medical costs, which are unavoidable in ‘the wait-and-see principle’, can be evaded.

Ideally, a study in which patients were randomized into either the PS group or the no PS group would have been a better comparison. However, such an approach can be achieved only after knowing the feasibility of pancreatic stenting in patients with ABP. We firmly believe that our study is a good cornerstone for future studies.

In conclusion, temporary pancreatic stenting after early ERCP for patients with ABP appears to be safe since there was no increase in the occurrence of complications even in the failed cases. However, if successful, local complications are significantly reduced.