5742Increased bleeding events in patients co-administered rivaroxaban and either cyp3A4 or p-gp inhibitors

Abstract

Background: The ROCKET AF trial established the non-inferiority of rivaroxaban compared to warfarin for prevention of thromboembolic events in patients with non-valvular atrial fibrillation. Reduced dosing is recommended for patients with renal insufficiency. However little is known about clinical outcomes when rivaroxaban is used with cytochrome P450 CYP3A4 or P-glycoprotein (P-gp) inhibitors. Hepatic CYP3A4 and renal P-gp are thought to play important roles in rivaroxaban degradation and renal excretion respectively. Both are subject to competitive inhibition by a number of commonly prescribed cardiac medications. We hypothesized that bleeding rates might increase when patients take rivaroxaban in combination with either CYP3A4 or P-gp inhibitors. Methods: We reviewed the records of patients prescribed rivaroxaban over a 5- year period between January 1st, 2011 and December 31st, 2015. Patients were included in the study if they (1) had a prescription for rivaroxaban for at least 90 consecutive days, (2) were prescribed at least one concomitant medication of interest, and (3) had a medical bleeding event occurring at least 7 days after the prescription date of the interacting medication. In addition to CYP3A4 and P-gp inhibitors, the impact of platelet aggregation inhibitors (PAIs) and non-steroidal anti-inflammatory drugs (NSAIDs) on bleeding was evaluated. Bleeding-related diagnoses, classified by individual ICD codes, were used to tabulate the number of bleeding events. Chi-square tests were used to compare the proportion of patients with bleeding events in the group that had concomitant interacting medications versus the group that did not. Results: 1674 patients were identified who were prescribed rivaroxaban and either a CYP3A4 inhibitor, P-gp inhibitor, PAI, or NSAID. The most commonly co-prescribed class of drugs was P-gp inhibitors (49.8% of patients). 10.1% of patients on rivaroxaban were co-prescribed PAIs, 13.6% CYP3A4 inhibitors and 26.6% NSAIDs. The greatest bleeding risk was observed with CYP3A4 inhibitors (40.0% vs. 17.1% on rivaroxaban alone; p<0.01). Increased bleeding rates were also seen with co-administration of PAIs (28.4%, p<0.1), P-gp inhibitors (24.1%, p<0.01) and NSAIDs (21.1%, p=0.01). Discussion: Concomitant use of either CYP3A4 or P-gp inhibitors with rivaroxaban was associated with increased bleeding risk. CYP3A4 inhibitors were associated with the greatest risk, while P-gp inhibitors were associated with an increased risk similar to that seen for patients prescribed NSAIDs or PAIs. Our data is limited by the number of patients available for analysis and use of ICD codes to determine bleeding. However, our study provides evidence that co-administration of rivaroxaban with either CYP3A4 or P-gp inhibitors may be associated with an increased risk of bleeding.