Opened Proton Pump Inhibitor Capsules Reduce Time to Healing Compared With Intact Capsules for Marginal Ulceration Following Roux-en-Y Gastric Bypass

Abstract

Marginal ulceration, or ulceration at the gastrojejunal anastomosis, is a common complication of Roux-en-Y gastric bypass (RYGB). Acidity likely contributes to the pathophysiology, and proton pump inhibitors (PPIs) frequently are prescribed for treatment. However, patients with gastric bypass only have a small gastric pouch and rapid small-bowel transit, which limits the opportunity for capsule breakdown and PPI absorption. Soluble PPIs (open capsules [OCs]) might be absorbed more easily than intact capsules (ICs). We compared time to ulcer healing, number of endoscopic procedures, and use of health care for patients with marginal ulceration who received PPIs in OC vs IC form. We performed a retrospective study of 164 patients diagnosed with marginal ulceration who underwent RYGB at the Brigham and Women's Hospital from 2000 through 2015. Patients received high-dose PPIs and underwent repeat endoscopy every 3 months until ulcer healing was confirmed. We used time-to-event analysis with a Cox proportional hazards model to evaluate the association between mode of PPI administration and time to ulcer healing, in addition to Cox multivariate regression analysis. Total charge (procedural and maintenance) was determined by comparison of categorized charges incurred from time of ulcer diagnosis to resolution. The primary outcome was time to healing of marginal ulceration in RYGB patients receiving high-dose PPIs in OC vs IC form. A total of 162 patients were included (115 received OC and 49 received IC). All patients were followed up until ulcer healing was confirmed. The median time to ulcer healing was 91.0 days for the OC group vs 342.0 days for the IC group (P < .001). OC was the only independent predictor of time to ulcer healing (P < .001) when we controlled for known risk factors. The number of endoscopic procedures (P= .02) and overall health care utilization (P= .05) were lower in the OC than the IC group. Patients with marginal ulceration after RYGB who receive OC PPIs have shorter ulcer healing times, fewer endoscopic procedures, and use less health care resources compared with patients who receive IC PPIs. Given these results and the high prevalence of marginal ulceration in this patient population, the use of OC PPIs is a low-risk, low-cost alternative that should be considered.