Risk factors and a prediction model for pain recurrence after pancreatic stent removal in painful chronic pancreatitis.

Abstract

OBJECTIVES Endoscopic pancreatic stenting (EPS) is an effective treatment modality for painful chronic pancreatitis. However, little is known about the factors that cause pain recurrence after stent removal, and there are no clear criteria for stent removal. We aimed to develop a prediction model for pain recurrence by identifying its risk factors. METHODS We retrospectively reviewed 95 patients who underwent EPS due to pain for the first time using a single plastic stent between January 2007 and July 2022 at our institute. Univariate and multivariate stepwise Cox proportional hazards models were used to identify the risk factors for pain recurrence, and a prediction model was developed based on the identified factors. RESULTS Of the 95 enrolled patients, 89 (93.7%) achieved pain relief and 73 (76.8%) did stent removal. Of the 69 patients with a follow-up period ≥6 months after stent removal, 29 (42.0%) had pain recurrence during the median follow-up period of 59 months. Serum lipase level (p = 0.034) and pancreatic parenchymal thickness (p = 0.022) on computed tomography or magnetic resonance imaging were identified as independent risk factors for pain recurrence. The prediction model based on the identified factors had good discrimination ability, with a concordance index of 0.74, and could stratify pain recurrence rates. CONCLUSIONS We identified the risk factors and developed a new prediction model for pain recurrence following stent removal. This model might be useful for decision-making in pancreatic stent management, such as deciding whether to remove a pancreatic stent, continue EPS, or convert to surgery.