A population based analysis of proton pump inhibitor use following pancreaticoduodenectomy: What is the impact on marginal ulceration and gastrointestinal bleeding?

Abstract

Background: Upper gastrointestinal bleeding (UGIB) and marginal ulceration (MU) are significant postoperative issues following pancreaticoduodenectomy(PD). The role of proton pump inhibitor(PPI) therapy in preventing complications remains uncertain. A survey (Butler et al, 2015) of PPI prescription practice by HPB surgeons demonstrated a degree of consensus; with 38% of surgeons preferring PPI therapy for 6 months or more and 48% preferring lifelong PPI therapy. The aim of this study was to assess how such opinions translate into real world practice. A population based analysis of adherence to PPI prescription following PD and its impact on MU and UGIB rates was therefore undertaken. Methods: A retrospective review of 319 PDs performed in one institution between January 2011 and December 2016 was performed. Demographic, intraoperative and postoperative outcome data were recorded using electronic hospital administrative databases and linked to hospital and community based pharmacy prescription databases to assess duration of PPI therapy and associated medications (non steroidal anti inflammatory drugs(NSAIDs), oral anticoagulants). The number of postoperative PPI prescription days was quantified and categorised into 6 monthly intervals. Univariate statistical analyses were then carried out to assess association between PPI duration with MU and UGI bleeding complications. Further assessment was also made of PPI prescription and outcome with a logistic regression model adjusted for standard demographic and perioperative factors. Analysis was with Stata v14 with the significance level set at p<0.05. Results: The mean age was 63.7years (SD 14) (52% male). Median follow up was 666 days (95%CI 606–72) and median postoperative stay 13 days (IQR 10 – 20). The major complication rate was 10.7%. The marginal ulceration rate was 3.1% (95% CI 1.5 –5.6%) and postoperative GI bleeding was 5% (95%CI 2.9–8%). Median time to development of GI bleeding / ulceration was 7.5 months (IQR 1–28 months). The in-hospital postoperative PPI prescription rate was 89%. In the community (post-discharge) the rate reduced to 68.4%. Adjusting for postoperative death during the same period; 50.8%, 41.3%, 27.8% of PD patients had prescribed PPI therapy for a minimum of 6 months, 1 year and 2 years respectively. Only 37% of patients were prescribed PPI therapy for greater than 50% of their follow-up period and only 13% of patients were prescribed PPI for the full duration of their follow-up period. Inpatient PPI therapy had no impact on in-hospital major complications (11.1% v 10.6% p = 0.93) and was associated with an increased hospital stay (12.2 days v 18.5 days p = 0.02). The prescription of postoperative PPI did not impact upon GI bleeding or marginal ulceration rates (2.5 v 3.3% p = 0.723 for MU and 2.5% v 5.8% p = 0.24 for GI bleed). When stratified by 6 monthly durations, PPI therapy had no impact on MU (OR 1.2 95%CI 0.8–1.8 p = 0.371). Total PPI use of between 6 months and 12months duration was associated an increased odds of GI bleeding (OR 1.4 1.1–1.97 p = 0.04). In a model adjusted for age, sex and NSAID use no association between PPI therapy duration and GI bleeding was observed (OR 1.37 95% CI 0.97–1.94 p = 0.17). Conclusion: The pattern of postoperative PPI use amongst patients undergoing PD is variable, with only a third of patients using PPI therapy for the majority of their postoperative course. In-patient PPI therapy does not alter early postoperative complication rates. Longer term PPI use does not appear to reduce post PD gastrointestinal bleeding or marginal ulceration complications. This population based analysis of PPI use has demonstrated a lack of efficacy for routine PPI prescription in preventing marginal ulceration and upper gastro intestinal bleeding complications following pancreaticoduodenectomy. The routine long term prescription of postoperative PPI therapy following PD requires further prospective evaluation.