Non-Operative Management of Pancreatic Pseudocysts

To the Editor:

I read with interest the article by Nealon et al on the surgical management of complications associated with percutaneous and/or endoscopic management of pancreatic pseudocysts.¹ The authors reported that clinical sepsis was encountered in 91% (72 of 79) of patients managed by nonoperative means and close to 85% required subsequent surgery. Moreover, complications pertaining to nonoperative management necessitated emergent or urgent surgery in greater than 50% of their patients. Also, 84% of patients who had pseudocysts in association with a noncommunicating main pancreatic duct (MPD) or stricture or chronic pancreatitis had poor outcomes when managed nonoperatively.

The rates of sepsis and failure reported by Nealon et al is in contrast to our recent report on endoscopic management of pancreatic ductal disruption by transpapillary stenting.² In this series, the largest reported to date, 55% of 97 patients had a successful outcome, defined as clinical resolution of ascites, peripancreatic fluid collection and/or fistula, and resolution of the PD disruption at the final ERCP. A further 8% had complete resolution of the fluid collection although follow-up pancreatogram revealed a MPD cut-off. Procedural complications were encountered in 6% and only 1 patient developed sepsis. On multivariable logistic regression, only a partially disrupted duct and a stent bridging the disruption correlated with a successful outcome.

Nealon et al rightly allude that the reason for high rate of sepsis and surgical intervention in their series was due to possible referral bias. Their cohort comprised a subset of patients who had failed nonoperative means of management. However, we did not encounter such high rates of sepsis following endotherapy in our referral group of patients. Also, 20% of the 97 patients in our study already have had a prior intervention (indwelling percutaneous drainage catheter). The authors do not provide any data on preprocedural antibiotic prophylaxis in their patients. We routinely prescribe prophylactic antibiotics in patients undergoing endotherapy for management of pancreatic fluid collections, a view endorsed by the American Society for Gastrointestinal Endoscopy.³ The authors note that the nutritional status of the patients who underwent nonoperative management was poor, with mean percentage weight loss of 12.3% and a mean serum albumin of 2.1 ± 1.1; it is unclear if any of the patients were on enteral nutrition. We generally place a direct percutaneous endoscopic jujenostomy feeding tube in almost all patients following drainage of pancreatic fluid collections to facilitate pancreatic rest and prevent bacterial translocation that could lead to pancreatic necrosis in patients with severe acute pancreatitis.^4,5 The authors have elegantly categorized ductal disruption and propose that patients with associated chronic pancreatitis have a poor outcome (Type 7 of their classification). This has not been our experience or that of 2 other large series that have evaluated this issue.^2,6,7 In fact, it appears that patients with pseudocysts in the setting of chronic pancreatitis have fewer complications, and outcomes of endotherapy in this cohort is comparable to those with acute pancreatitis. Endoscopists proficient with pancreatic endotherapy rarely venture and drain necrotic fluid collections as data in this area is scant and outcomes so far have been far from satisfactory.⁶ We do agree with Nealson and colleagues that patients with an abrupt duct cut-off (Type 5) have poor outcome when managed nonoperatively.

Another caveat in the study by Nealon et al is that less than 50% of their patients had undergone endoscopic means of drainage; also, more than two-thirds had undergone percutaneous drainage as initial therapy. Thus the true efficacy of endotherapy, even in this selected cohort, is hard to determine. It appears that only 5 of 20 patients managed by transpapillary stenting per se had resolution of ductal disruption. There is also lack of details of endotherapy such as type of stent and its relationship to the disruption in the pancreatic duct. Placing just a stent or any stent does not necessarily imply successful endotherapy. Two studies have clearly shown that placement of a stent that bridges the disruption is associated with a successful outcome.^2,7 In our series, 92% of patients with a bridging stent had a successful outcome versus 44% in whom the disruption could not be bridged.

Several factors determine success of pancreatic endotherapy, such as experience of the operator, placement of an appropriate stent (bridging versus nonbridging) depending on the nature of pancreatic duct disruption (complete versus partial), administration of preprocedural prophylactic antibiotics, initiation of enteral nutrition at early stages of the disease in patients with severe acute pancreatitis, and close collaboration with pancreaticobiliary surgeons and interventional radiologists. Only randomized trials comparing surgical, endoscopic, and radiologic interventions can provide a definitive answer to our quest for the most appropriate management strategy in these patients.

Shyam Varadarajulu, MD
Division of Gastroenterology-Hepatology
University of Alabama at Birmingham
School of Medicine, Birmingham, Alabama
svaradarajulu@yahoo.com