Connected or disconnected: What's next after successful transmural drainage of pancreatic fluid collection?

Over the years, management of pancreatic fluid collection (PFC) has shifted towards minimally invasive strategies. Since the first case was described three decades ago,1 endoscopic ultrasonography (EUS)-guided transmural drainage of symptomatic PFC has spread worldwide and played a pivotal role as a minimally invasive alternative to conventional surgery and percutaneous drainage. PFC, a local complication of acute pancreatitis, is classified as either pseudocysts or walled-off necrosis (WON), based on the presence or absence of necrotic debris inside. Unlike pseudocysts, WON is more challenging to drain because it is more extensive and can contain a large amount of solid necrotic tissue. The recently developed lumen-apposing metal stents (LAMS), which are uniquely designed and placed under EUS guidance, have led to a paradigm shift in the management of PFC, including WON. The wide diameter of the LAMS allows repeated passage of the endoscope and its accessories and facilitates endoscopic necrosectomy in WON even if sufficient drainage is difficult to achieve with stent placement alone. According to a recent meta-analysis, technical and clinical success rates of EUS-guided drainage using LAMS are as high as 96% and 87%, respectively.2 Given the accumulating evidence of favorable clinical results, LAMS are being increasingly used in lieu of conventional plastic stents, especially for draining WON. Although the advent of LAMS has ensured short-term efficacy and safety of EUS-guided transmural drainage for PFC, issues regarding long-term management remain because PFC can recur in a proportion of cases after LAMS removal. Notably, disconnected pancreatic duct syndrome (DPDS) is a significant, but under-recognized, complication associated with acute necrotizing pancreatitis, and its adequate management is an unmet clinical need that endoscopists face despite successful endoscopic PFC drainage with LAMS. The reported incidence of DPDS among PFC cases varies from 16% to 74%,3-5 and its pathogenesis involves necrosis and subsequent complete disruption of the main pancreatic duct. Consequently, the upstream viable pancreas becomes an isolated functional segment that continues to secrete pancreatic juice containing digestive enzymes, ultimately resulting in the formation of a PFC. Hence, even though the presence of DPDS is one of the frequent causes of PFC recurrence, little attention has been paid to the management of DPDS after LAMS removal. In this issue of Digestive Endoscopy, Pawa et al. have addressed this matter by investigating the impact of plastic stent replacement after LAMS removal for preventing PFC recurrence. This approach helps maintain continuity of the transmural fistula created during LAMS placement. This study focused on cases with DPDS, which occurred in 52% (50/96) of the patients with symptomatic PFC, and they evaluated recurrence-free survival in PFC with DPDS that was managed with or without double pigtail stent (DPS) replacement after LAMS removal.6 Analysis of DPDS cases enrolled in the study revealed that the recurrence rate of PFC was 5% (1/21) in the DPS replacement group and 37% (10/27) in the non-DPS replacement group, with DPS patients showing significantly longer recurrence-free survival than those without DPS during a median follow-up duration of 20 months. Furthermore, the authors showed that extended duration of LAMS placement was negatively associated with successful DPS replacement. Based on these findings, the authors emphasize the importance of early exchange of LAMS with DPS to improve recurrence-free survival in patients with DPDS after successful transmural drainage. Although this was a single-center retrospective study in a small cohort, these results open a new avenue in the long-term management of PFC with DPDS after LAMS removal.6 Nevertheless, there remain unanswered questions about the long-term management of PFC with DPDS. The first question pertains to the definition and diagnosis of DPDS because its definitive diagnosis requires the following three criteria: (i) necrosis of at least 2 cm of the pancreas; (ii) total cut-off of the main pancreatic duct at the site of necrosis; and (iii) presence of viable pancreatic tissue upstream from the site of necrosis.7 As a diagnosis of DPDS, endoscopic retrograde cholangiopancreatography (ERCP) has contributed as the gold standard in the past for the diagnosis of the ductal disruption; however, since ERCP is an invasive diagnostic modality with a potential risk of procedure-related pancreatitis, the need to perform diagnostic ERCP in patients with a history of severe acute pancreatitis requires careful consideration. Further, even though magnetic resonance cholangiopancreatography (MRCP) has been described as an alternative imaging modality, there is currently no single accepted definition of DPDS on MRCP. Congruently, in Pawa et al.’s study,6 DPDS was confirmed by MRCP in only 69% (33/48) of the patients, with the remaining 31% (15/48) requiring diagnostic ERCP. Therefore, identifying a less-invasive diagnostic modality and criteria that can definitively diagnose DPDS is desirable. Recently, Bang et al. have proposed EUS as a useful method to diagnose DPDS,8 and they defined DPDS as the presence of a well-defined fluid collection along the course of the main pancreatic duct with the upstream pancreatic parenchyma and the duct terminating at the fluid collection. They demonstrated a perfect correlation between EUS findings and a diagnosis of DPDS in a series of 21 PFC patients with DPDS.8 Despite the small number of cases, this promising result can be expected to make EUS an attractive and minimally invasive diagnostic modality for DPDS. Furthermore, contrast-enhanced harmonic EUS may be helpful in the diagnosis of DPDS because it can delineate the pancreatic duct as an avascular structure. The second question is the optimal timing of LAMS removal. In Pawa et al.’s study,6 DPS replacement after LAMS removal was successful in only 44% (21/48) of the patients with DPDS, much lower than the 75% reported previously.4 To explain this low success rate, they explored factors associated with failed DPS replacement and identified a longer duration of LAMS placement as a factor.6 Notably, in their study, the median duration of LAMS placement was 28 days in the successful DPS replacement group and 53 days in the unsuccessful DPS replacement group.6 Therefore, the authors recommended the early replacement of LAMS with DPS. Additionally, from the perspective of potential adverse events, a LAMS indwelling time of greater than 4 weeks is considered to be a predictor of delayed bleeding and buried stent syndrome; however, LAMS removal within 4 weeks depends on the clinical course after endoscopic drainage. In contrast, a recent large-cohort study that included 952 PFC cases drained with LAMS revealed that the timing of LAMS removal (4–8 weeks vs. >8 weeks) did not affect delayed adverse events.9 Although factors affecting LAMS indwelling period were not evaluated in the Pawa et al.’s study,6 delayed LAMS removal is recommended when clinical success cannot be achieved without repeated necrosectomy after transmural drainage with LAMS. Therefore, it would be helpful to suggest an optimal timing of LAMS removal based on the degree of PFC reduction rather than the indwelling period. Additionally, deployment of DPS across LAMS at the time of initial drainage can minimize the risk of complete cavity collapse and thereby possibly improve the success of DPS replacement. The third question relates to the safety of a long-term transmural indwelling DPS left in situ. Basha et al. have evaluated the correlation between DPDS and new-onset diabetes mellitus (DM) in 274 PFC cases that were drained with LAMS.5 In that study, LAMS was removed without DPS replacement, regardless of the presence or absence of DPDS, and they show that a majority of the patients who developed PFC had DPDS, and that, surprisingly, new-onset DM was observed in one-third of cases with DPDS.5 Taken together, the findings of Pawa et al. and Basha et al. indicate that long-term DPS placement in situ can prevent PFC recurrence; however, these studies provide no answers about when to remove the DPS after long-term placement. Further, it is unclear whether long-term indwelling DPS helps reduce the incidence of new-onset DM, and whether there was a difference in new-onset DM between the two groups in Pawa et al.’s study (i.e., with and without DPS) is a matter of considerable interest.6 Regarding the safety of leaving a DPS in place for a long period, a recent report has stated that, during a median follow-up of 21 months, colonic perforations occurred in 8.3% (3/36) of the PFC patients with long-term DPS placement.10 We have also experienced a severe adverse event wherein a migrating DPS, left in place for more than 2 years, had penetrated the duodenal wall and required a laparotomy (unpubl. data). Currently, as there is no universal consensus on the optimal duration of an indwelling DPS, careful follow-up on a case-by-case basis is required wherein both benefits and risks are prudently weighed. Further studies are also needed to evaluate the impact of long-term indwelling DPS on pancreatic endocrine and exocrine functions. In summary, DPDS should not be overlooked after transmural drainage of PFC with LAMS. In this era of LAMS, the establishment of a long-term treatment strategy against DPDS, as well as drainage of PFC, is mandatory. Early replacement of LAMS with DPS appears to be a promising treatment option after successful endoscopic drainage with LAMS for PFC with DPDS. However, a few clinical questions require clarification, namely: (i) what is the accepted definition of DPDS and how to diagnose it? (ii) what is the optimal timing of LAMS and DPS removal? and (iii) is it safe to leave DPS in situ for a long period? Importantly, it should be noted that a significant fraction of recurrent PFC are asymptomatic and do not require reintervention.4-6 We hope that future large-scale trials will provide more evidence on these unanswered questions that can arise after successful transmural drainage of PFC with LAMS. Authors declare no conflict of interest for this article. None.