Abstract
BACKGROUND: Polypharmacy, including use of inhibitors of CYP3A4 and P-glycoprotein (P-gp), is common in patients with venous thromboembolism (VTE) and is associated with increased bleeding. METHODS: In 8246 patients included in the EINSTEIN-VTE studies for acute VTE, we evaluated the effect of polypharmacy on bleeding and on the relative differences between rivaroxaban and enoxaparin/vitamin K antagonist (VKA). We assessed the incidence of clinically relevant bleeding (major and clinically relevant nonmajor bleeding) by number of comedications (none, 1-3, ≥4) at baseline, and by use of CYP3A4 and/or P-gp inhibitors. Interaction between rivaroxaban versus enoxaparin/VKA and comedication was assessed by Cox regression analysis with p(interaction) estimates. RESULTS: With increasing number of comedications, the incidence of clinically relevant bleeding rose from 5.7% to 13.3% in rivaroxaban recipients and from 9.1% to 11.1% in enoxaparin/VKA recipients. Whereas rivaroxaban was associated with a reduced bleeding risk compared with enoxaparin/VKA in patients without comedication (hazard ratio [HR] 0.6, 95% confidence interval [CI] 0.4-0.9), the risk was similar in patients with ≥4 comedications (HR 1.2, 95% CI 0.97-1.5, p(interaction) .002). Use of CYP3A4 and/or P-gp inhibitors was associated with a doubled bleeding risk compared with no use, without a difference between rivaroxaban and enoxaparin/VKA. CONCLUSION: We conclude that fixed-dose rivaroxaban as compared with enoxaparin followed by dose-adjusted VKA is not associated with an increased bleeding risk in patients with VTE administered polypharmacy in general and CYP3A4 and/or P-gp inhibitors specifically. This implies that the observed increased bleeding risks with polypharmacy and use of CYP3A4 and/or P-gp inhibitors are likely explained by comorbidities and frailty, and not by pharmacokinetic interactions.