Clinical outcomes following endoscopic retrograde cholangiopancreatography-based therapy compared with surgery in chronic pancreatitis: evidence from a multicentric cohort.

Abstract

Background: Real-world comparative outcomes of endoscopic retrograde cholangiopancreatography-based endoscopic therapy vs. pancreatic surgery in chronic pancreatitis (CP) remain incompletely defined.

Methods: We performed a retrospective comparative effectiveness study using the TriNetX US Collaborative Network. Adults with CP undergoing endoscopic therapy or pancreatic surgery were identified, and propensity score matching generated 1451 patients in each cohort. Outcomes were assessed from 1-1095 days after the index event, using risk-based analyses as the primary comparative summaries and Kaplan-Meier analyses as secondary time-to-event summaries. Outcomes included chronic opioid prescriptions, opioid use disorder (OUD), celiac plexus block/neurolysis, pain codes, acute pancreatitis, exocrine and endocrine pancreatic insufficiency, emergency visits, and all-cause mortality.

Results: Mean follow up was 771.494 days for endoscopy and 827.285 days for surgery. New chronic opioid prescriptions occurred in 146/1092 (13.4%) endoscopic vs. 79/1032 (7.7%) surgical patients (odds ratio [OR] 1.862, 95% confidence interval [CI] 1.396-2.484; P<0.001; hazard ratio [HR] 1.861, 95%CI 1.415-2.448). Acute pancreatitis occurred in 167/530 (31.5%) vs. 65/565 (11.5%) patients (OR 3.539, 95%CI 2.578-4.858; P<0.001; HR 3.341, 95%CI 2.506-4.453). Exocrine pancreatic insufficiency occurred in 200/1213 (16.5%) vs. 132/1206 (10.9%) patients (OR 1.606, 95%CI 1.269-2.034; P<0.001; HR 1.585, 95%CI 1.272-1.975). Endocrine pancreatic insufficiency, mortality, emergency visits, and OUD did not differ significantly between cohorts.

Conclusions: Endoscopic therapy was associated with higher chronic opioid prescribing, acute pancreatitis, celiac plexus block/neurolysis, and exocrine pancreatic insufficiency. These findings are associative and should be interpreted in the context of residual confounding and coding-based outcome ascertainment.