PATIENT CHARACTERISTICS ARE ASSOCIATED WITH RESPONSE TO ERCP FOR SPHINCTER OF ODDI DISORDERS, BUT PHYSICIAN-DEFINED SOD SUBTYPES ARE NOT: THE RESPOND LONGITUDINAL COHORT STUDY

Background: Sphincter of Oddi Disorders (SOD) describes a contentious and heterogeneous diagnosis in patients whose abdominal pain, idiopathic acute pancreatitis (AP), or both may derive from a spastic or stenotic sphincter. Traditionally, SOD is categorized by ≥1 physician-defined factor: dilated bile duct ≥12mm, elevated liver (>2x ULN) or pancreas (>3x ULN) labs, and idiopathic AP. With broader use of noninvasive imaging and evidence suggesting no benefit of ERCP for abdominal pain alone, the Results of Ercp for Sphincter of Oddi Disorders (RESPOnD) study aimed to measure the benefit of ERCP in this controversial context.

Methods: RESPOnD was a prospective cohort conducted at 14 U.S. centers. Patients undergoing first-time ERCP with sphincterotomy for suspected SOD (biliary, pancreatic, or mixed type) were eligible. After informed consent, patients completed a baseline assessment including validated questionnaires to characterize pain, pain-related disability, and potential confounders such as somatization, mood disorder, and recent opioid use. At 12 months after ERCP, the primary outcome was defined as the composite of being “improved” or “much improved” by Patient Global Impression of Change, no new or increased opioids, and no repeat intervention. Missing data were addressed using a hierarchal, multiple imputation scheme.

Results: Of 316 patients screened, 216 were enrolled and 3 excluded due to inability to complete ERCP (2) or missing consent (1). Most (190/213, 89%) subjects presented with ≥1 physician-defined characteristics of SOD (duct dilation, abnormal pancreatobiliary lab, and/or AP history). Of 213 patients, an average of 122 (57.4% [95%CI 50.4-64.4]) met the multiply imputed primary outcome for success; 23/213 (10.8%) developed post-ERCP pancreatitis. Outcomes were similar across Rome III subtypes (table 1). Recent opioid use, irritable bowel syndrome history, somatization, depression, anxiety, and low physical or mental health were associated univariately with lower success (table 2). AP > 30 days after index ERCP (or follow-up ERCP, if performed) developed in 40/213 (18.8%) at a median of 6 months after index ERCP and was more likely if there was a prior history of AP (31.8 vs. 4.9%, p<0.0001).

Conclusion: Patient characteristics are associated with response to ERCP treatment whereas traditional physician-defined characteristics such as duct dilation are not. The overall response rate (57%) is numerically higher than in the EPISOD study (abdominal pain alone) but does not rule out a substantial placebo response. AP occurs frequently after sphincterotomy, especially if there was a history of AP prior to enrollment, suggesting no benefit. Additional examination needs to be done in a multivariable setting and future trials should focus on measuring the treatment effect in patients with the most favorable baseline characteristics.