Abstract No. 627 Proton pump inhibitors are associated with increased risk of hepatic encephalopathy after transjugular intrahepatic portosystemic shunt creation

Abstract

Hepatic encephalopathy is the most common complication after transjugular intrahepatic portosystemic shunt (TIPS) creation. Although proton pump inhibitors (PPIs) are regarded as a class of medications with an excellent safety profile, recent data suggests that in cirrhotic patients PPIs may be associated with an increased risk of hepatic encephalopathy (HE), perhaps by altering microbial flora in the gastrointestinal tract with resultant changes in metabolites in portal blood. The purpose of this study was to determine whether PPI usage was associated with an increased risk of HE exacerbations after TIPS. In this single-institution retrospective study, 199 patients (58 females, mean age = 56) underwent TIPS creation between 1/1/2005 and 4/1/2012 for ascites or hydrothorax (129), variceal hemorrhage (61), or both (9). The medical record was reviewed to determine dates of postprocedure PPI usage and HE exacerbations, as defined by hospital admission for HE or escalation in outpatient medical management. Negative binomial regression was used to analyze the rate of HE exacerbations per person-day following TIPS. Among 105 patients on chronic PPI therapy from the time of TIPS until death, transplant, or loss to follow-up, 73 patients experienced 135 HE exacerbations in 56207 person-days. Among 51 patients never on PPIs after TIPS, 15 experienced 16 HE exacerbations in 19517 person-days, a significantly lower rate of HE exacerbations per person-day compared to PPI users (p<0.001). In multivariate regression, older age (p = 0.04), higher pre-TIPS MELD (p<0.001), and PPI usage (p<0.001) were independent predictors of HE exacerbations. The remaining 43 patients had intermittent PPI usage after TIPS. This independent crossover sample experienced 48 HE exacerbations during 32387 days on PPIs and 25 HE exacerbations during 26836 days off PPIs, a non-significant trend toward increased risk while on PPIs (p = 0.08). In this study PPI use was associated with more frequent HE exacerbations after TIPS creation. These results raise the possibility that discontinuation of non-essential PPIs may reduce the risk of HE following TIPS.