(ASPOT26) Effect of drug-drug interactions involving direct-acting oral anticoagulants and statins on the risk of bleeding

Background: Certain statins (atorvastatin, lovastatin, and simvastatin) may potentiate the risk of bleeding associated with direct-acting oral anticoagulants (DOACs; rivaroxaban, apixaban, dabigatran, edoxaban) through drug-drug interactions via cytochrome P450 3A4 and P-glycoprotein.

Objectives: To examine the risk of major bleeding associated with concomitant use of DOACs and statins that interact with DOACs vs. concomitant use of DOACs and statins that do not interact with DOACs.

Methods: We used Merative Marketscan data (2011–2019) and emulated a target trial for the comparison of the risk of major bleeding between DOAC+interacting statin (atorvastatin, lovastatin, or simvastatin) use vs. DOAC+non-interacting statin (fluvastatin, pravastatin, or rosuvastatin) use. We included adults aged ≥18 years who concomitantly used DOACs and statins. The primary outcome was hospitalization with major bleeding. We followed all patients from the first date of concomitant use of DOACs and statins until the occurrence of outcome, discontinuation of DOAC or statins, switching to different statins (switching allowed within interacting and non-interacting stains) or the end of study period. A logistic model with demographic characteristics, comorbidities, coprescriptions, and components of CHA₂DS₂-VASc score and HAS-BLED score was used to derive inverse probability of treatment weights (IPTW). We used IPTW Cox proportional hazards model to determine the risk of major bleeding associated with concomitant DOAC+statin use.

Results: Among 143,952 concomitant DOAC-statin users, 104,327 (72.5%) used an interacting statin and 39,625 (27.5%) used a non-interacting statin. The cohort had a median age of 60 years and 32.8% were female, 50.8% had atrial fibrillation, and 22.6% had history of stroke, transient ischemic attack, and venous thromboembolism. The two predominant treatment combinations were rivaroxaban with atorvastatin (35,929, 25%) and apixaban with atorvastatin (28,734, 20%). In the IPTW weighted cohort, all variables were well-balanced across the two exposure groups. In the weighted Cox model, DOAC+interacting statin use was not associated with an increased risk of major bleeding compared to DOAC+non-interacting statin use (4.3 vs. 4.1 events per 100 person-years; adjusted hazard ratio [HR], 1.02; 95% confidence intervals [CI], 0.88-1.18).

Conclusions: Among concomitant DOAC and statin users, the risk of major bleeding was similar between those who used an interacting statin compared to those who used a non-interacting statin. Pharmacokinetic interactions between DOACs and statins through P450 3A4 and P-glycoprotein may not appreciably pose increased bleeding risk at the population level.