EUS-directed pancreatic duct drainage: Mainstream miracle or proceed with caution?

Abstract

Abdominal pain is the foremost adverse event (AE) of chronic pancreatitis (CP). The mechanism of pain in CP is incompletely understood and likely multifactorial, including mechanical (intraductal pressure/obstruction), inflammatory, malabsorptive, and neurogenic/neuropathic changes in the pancreas and surrounding organs.1Anderson M.A. Akshintala V. Albers K.M. et al.Mechanism, assessment and management of pain in chronic pancreatitis: recommendations of a multidisciplinary study group.Pancreatology. 2016; 16: 83-94Crossref PubMed Scopus (58) Google Scholar In addition, patients with CP can have nonvisceral pain associated with high levels of psychologic stress. The intensity of pain and frequency of pain attacks compromise the quality of life in CP patients. In addition to analgesic medications and pancreatic enzyme therapy, there is consensus that there is a role for endoscopy therapy in a subset of CP patients with evidence of ductal obstruction. Although endoscopic retrograde pancreatography (ERP) with stent placement remains the primary technique of endoscopic decompression of the PD, EUS-guided PD drainage (EUS-PDD) has emerged as an alternative approach in recent years, especially after failed ERCP. In this issue of Gastrointestinal Endoscopy, Krafft et al2Krafft M.R. Croglio M.P. James T.W. et al.Endoscopic endgame for obstructive pancreatopathy: outcomes of anterograde EUS-guided pancreatic duct drainage. A dual-center study.Gastrointest Endosc. 2020; 92: 1055-1066Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar report the results of their retrospective case series on outcomes of antegrade EUS-PDD. This was a dual-center retrospective review encompassing a total of 28 patients who underwent attempted antegrade EUS-PDD. The study included 26 patients with CP and 2 patients with stenosis of the pancreaticojejunostomy (PJS) after Whipple surgery. In the CP cohort, pancreaticogastrostomy (PG) was attempted during the index procedure. In patients who experienced clinical success with PG performed during the index procedure, repeated endoscopy session(s) were performed to attempt transpapillary drainage. In the PJS cohort, gastropancreatoentrostomy (also known as ring drainage) was attempted during the index procedure. The study endpoint was achievement of transpapillary/transanastomotic drainage, which was defined as definitive therapy. In the CP cohort (n = 26), the rates of technical and clinical success of PG performed in the index procedure were 81% and 75%, respectively, with an AE rate of 15%. With repeated endoscopic interventions (ranging from 1 to 6 additional procedures) in the 15 patients who experienced clinical success, transpapillary drainage (definitive therapy) was achieved in 100%. Clinical success after definitive therapy was maintained in 100% of patients during a median follow-up time of 4.5 months. In the PJS cohort (n = 2), the rates of technical and clinical success after ring drainage were 100% and 100%, respectively, which was maintained during a mean follow-up time of 18 months. The authors concluded that PG is a destination therapy and a step toward definitive therapy (transpapillary drainage) in CP patients. Transpapillary drainage was proposed as the endoscopic endpoint because once it is attained, future procedures (if needed) could be performed through conventional ERP, a widely available and reliable technique. Prior studies have shown that pancreatic interstitial pressure is elevated in patients with CP compared with control individuals3Bradley III, E.L. Pancreatic duct pressure in chronic pancreatitis.Am J Surg. 1982; 144: 313-316Abstract Full Text PDF PubMed Scopus (237) Google Scholar, 4Karanjia N.D. Widdison A.L. Leung F. et al.Compartment syndrome in experimental chronic obstructive pancreatitis: effect of decompressing the main pancreatic duct.Br J Surg. 1994; 81: 259-264Crossref PubMed Scopus (134) Google Scholar, 5Ebbehoj N. Borly L. Bulow J. et al.Pancreatic tissue fluid pressure in chronic pancreatitis: relation to pain, morphology, and function.Scand J Gastroenterol. 1990; 25: 1046-1051Crossref PubMed Scopus (92) Google Scholar and is higher in CP patients with pain than in those without.5Ebbehoj N. Borly L. Bulow J. et al.Pancreatic tissue fluid pressure in chronic pancreatitis: relation to pain, morphology, and function.Scand J Gastroenterol. 1990; 25: 1046-1051Crossref PubMed Scopus (92) Google Scholar Surgical series have reported a decrease in interstitial pressure after surgical ductal drainage.6Ebbehoj N. Borly L. Bulow J. et al.Evaluation of pancreatic tissue fluid pressure and pain in chronic pancreatitis: a longitudinal study.Scand J Gastroenterol. 1990; 25: 462-466Crossref PubMed Scopus (128) Google Scholar Although we assume that endoscopic PD drainage decreases both intraductal and interstitial pressures, there are limited data to support this hypothesis. In addition, a common argument against the widespread adoption of endoscopic therapy in the management of CP is the lack of high-quality studies that support sustained clinical improvement. The limitations of previous studies include lack of prospective design and blinding, lack of control/sham groups, variable measures for assessment of pain, heterogeneous study population, inadequate postprocedure follow-up, and failure to address quality-of-life measurements and cost effectiveness. Unfortunately, several of these limitations also apply to the current study. The primary outcomes of the study were to report the long-term outcomes of anterograde EUS-PDD in addition to establishing an endpoint for endoscopic therapy. The mean follow-up time after definitive therapy in the PJS cohort (n = 2) was 18 months, but it was a small fraction of the study population (7%). The median follow-up time after definitive therapy in the CP cohort (which was a majority of the study population, 93%) was only 4.5 months. Although previous studies of endoscopic therapy in CP might have had shorter follow-up periods, 4.5 months is a relatively short interval upon which to draw any definitive conclusion. The rationale of the authors to combine the 26 patients with CP (with normal anatomy) and 2 patients with PJS (with post-Whipple anatomy) is unclear. Although both groups had obstructive pancreatopathy in common, they were distinct groups with regard to endoscopic technical challenges; hence, combining them made the study population heterogenous. In addition, the use of different techniques (ERP, rendezvous-assisted ERP, electrohydraulic lithotripsy) and stent types (plastic, metal, lumen-apposing metal stent) during the index and subsequent endoscopies at the discretion of endoscopists adds to the heterogeneity and further clouds any conclusions drawn. Although failure of ERP and rendezvous-assisted ERP were mandatory before PG was attempted in the CP group, it is unclear whether extracorporeal shock wave lithotripsy (ESWL) was performed as part of the failed ERP. In most tertiary endoscopy centers, ESWL is routinely performed as part of the initial ERCP when a large obstructive PD stone is noted on cross-sectional imaging, with high rates of successful subsequent stone clearance either spontaneously or by the use of ERP. In such centers, the need to attempt EUS-PDD is rare. In patients with CP where it is used, rendezvous-ERP has a high failure rate because of the presence of PD calculi, which can be difficult–if not impossible–to pass by use of a guidewire. Finally, calcification and fibrosis of the pancreas creates significant difficulty in passing transgastric instruments across the gland into the PD. It is therefore not surprising that initial PG creation failed in almost 20% of patients in the CP cohort for this very reason. This study also noted an AE rate of 15%, all of which occurred in the CP cohort. This compares with the AE rate (19.3%) reported by a previous systematic review and meta-analysis on EUS-PDD.7Krishnamoorthi R. Jayaraj M. Mohan B.P. et al.Efficacy and safety of endoscopic ultrasound (EUS) guided pancreatic duct drainage: a systematic review and meta-analysis [abstract].Gastrointest Endosc. 2017; 85: AB143Abstract Full Text Full Text PDF Google Scholar Pancreatic pseudocysts and walled-off necrosis that are amenable to endoscopic drainage are almost always located adjacent to the stomach or duodenum and are associated with significant inflammatory changes that act as an adhesive, thus facilitating endoscopic fistula formation and drainage. Although the PD can be brought into close proximity to the stomach during EUS by insufflation of the gastric lumen, once the endoscope is removed, the stomach contracts in size. As such, migration of transgastric stents out of the PD as a consequence is to be expected and was seen in this study. This speaks to the lack of an appropriately designed endoprosthesis for the purpose of EUS-PDD. It is interesting to note that in those patients who underwent ring drainage, stent migration and PD leak was not seen, suggesting an “anchoring” effect of the pigtail stents in both lumens. A major limitation of the study is the use of a subjective scale (“no relief,” “partial relief,” and “complete relief”) for assessment of clinical success. Moreover, both partial and complete relief were considered clinical successes without any differentiation between them. The fact that only 1 patient (who was compliant with abstinence from alcohol and without AEs) experienced clinical failure (“no relief”) could be a reflection of the scale being too lenient. The possibility of bias while extracting such subjective data in a retrospective study is difficult to rule out. Although 100% of patients had clinical success (CS) after definitive therapy, only 55% of patients (who were taking narcotics before endoscopic therapy) were able to discontinue or reduce the dose of narcotics. Several tools for assessment of pain in CP have been previously described, including Izbicki pain score, Brief Pain Inventory. and McGill Pain Questionnaire.8Olesen S.S. Juel J. Nielsen A.K. et al.Pain severity reduces life quality in chronic pancreatitis: implications for design of future outcome trials.Pancreatology. 2014; 14: 497-502Crossref PubMed Scopus (59) Google Scholar, 9Bloechle C. Izbicki J.R. Knoefel W.T. et al.Quality of life in chronic pancreatitis: results after duodenum-preserving resection of the head of the pancreas.Pancreas. 1995; 11: 77-85Crossref PubMed Scopus (184) Google Scholar, 10Seicean A. Grigorescu M. Tanţău M. et al.Pain in chronic pancreatitis: assessment and relief through treatment.Rom J Gastroenterol. 2004; 13: 9-15PubMed Google Scholar It is important to use scales that assess different dimensions of pain experience (sensory, affective, and evaluative) rather than unidimensional scales. Although these rigorous tools can be inconvenient for use in clinical research, they are objective measures, and their use can help reduce bias. The study authors seem to have assumed a hypothesis that transpapillary drainage (definitive therapy) is superior to PG (destination therapy). On that basis, repeated endoscopic procedures (up to 6 sessions in 1 patient) were performed until the goal of transpapillary drainage was reached. There is no question that transpapillary drainage, once established, is more convenient for reintervention (if needed). However, there is no high-quality evidence to support the hypothesis, and the mode of PD decompression (transpapillary vs PG) may not make a difference in establishing or maintaining clinical success. In the absence of a comparative arm, the study conclusion that transpapillary drainage is the endoscopic endpoint is not entirely supported by the study results. If a patient is symptomatically doing well after the index PG, it may be reasonable to try 1 additional endoscopy to establish transpapillary drainage, but reattempting multiple times for the purpose of reaching a set, somewhat arbitrary, goal of “definitive” therapy may be unnecessary. Given the lack of data supporting a superiority of transpapillary drainage, the risk of AEs and the cost associated with each additional procedure requires consideration. Over the past 2 decades, there has been an increase in consensus that a subset of patients with CP benefits from endoscopic therapy to decompress the PD. EUS has emerged as an important tool in endoscopic therapy when ERP fails. What is clear from the study by Krafft et al2Krafft M.R. Croglio M.P. James T.W. et al.Endoscopic endgame for obstructive pancreatopathy: outcomes of anterograde EUS-guided pancreatic duct drainage. A dual-center study.Gastrointest Endosc. 2020; 92: 1055-1066Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar is that EUS-PDD is an advanced and invasive technique and that AEs occur at a high rate, even in expert hands. Moreover, appropriate patient selection is critical, and this procedure should not be used as a primary substitute for the lack of availability of conventional techniques, especially for patients with chronic calcific pancreatitis, for which ESWL has a long and successful track record of helping to achieve PD drainage. In the current time of increasing healthcare costs and decreasing reimbursements, evidence-based medicine and clearly defined treatment goals should guide endoscopists on the role for and goals of any endoscopic therapy in CP. Although EUS-PDD is clearly a technical tour de force and innovative, its widespread adoption requires additional evidence from controlled, prospective studies in addition to endoscopic tools that are specifically designed for the procedure being performed. Dr Ross is a consultant for Boston Scientific. The other author disclosed no financial relationships. Endoscopic endgame for obstructive pancreatopathy: outcomes of anterograde EUS-guided pancreatic duct drainage. A dual-center studyGastrointestinal EndoscopyVol. 92Issue 5PreviewAnterograde endoscopic ultrasound-guided pancreatic duct drainage (EUS-PDD) refers to transmural drainage of the main pancreatic duct via an endoprosthesis passed anterograde through the gastric (or intestinal) wall. Anterograde EUS-PDD is a rescue procedure for recalcitrant cases of benign obstructive pancreatopathy. Full-Text PDF