In patients with atrial fibrillation using DOACs for thromboembolic prophylaxis, concurrent use of some medications may be associated with an increased risk of non-traumatic major bleeding

EBM Focus - Volume 12, Issue 42

Read the full EBM Focus and earn CME credit.

Reference: JAMA 2017 Oct 3;318(13):1250

  • In patients taking direct (non-vitamin K) oral anticoagulants (DOACs) for atrial fibrillation, evidence is limited regarding how the risk of bleeding is affected by the concurrent use of medications that share metabolic pathways with DOACs.
  • A recent retrospective cohort study investigated the possible influence of these medications on risk of non-traumatic major bleeding in over 90,000 adults in Taiwan who had nonvalvular atrial fibrillation and at least one DOAC prescription (dabigatran, rivaroxaban, and/or apixaban) from 2012 to 2016.
  • Fluconazole, phenytoin, rifampin, and amiodarone were each associated with an increased risk of non-traumatic major bleeding, with the 99% CIs for adjusted rate ratios across the 4 medications ranging from 1.02 to 3.07. Concurrent medications that were not associated with an increased risk include atorvastatin, azoles other than fluconazole, cyclosporine, digoxin, diltiazem, dronedarone, erythromycin/clarithromycin, and verapamil.

DOACs are often prescribed for patients with atrial fibrillation to reduce the risks of thromboembolism and ischemic stroke (Am J Cardiovasc Drugs 2015). DOACs are associated with an increased risk of bleeding, and how that risk is affected by concurrent medications that share metabolic pathways with DOACs is not known. This question was investigated in a recent retrospective cohort study of 91,330 adults (mean age 74 years) in Taiwan with nonvalvular atrial fibrillation who had at least one DOAC prescription (dabigatran, rivaroxaban, and/or apixaban) from 2012 to 2016 (edoxaban was not available during the study period). National health records were used to assess non-traumatic major bleeding events and concurrent use of any of 12 medications that share metabolic pathways with DOACs: P-glycoprotein competitors (digoxin, verapamil, diltiazem, amiodarone, and cyclosporine), CYP3A4 inhibitors (fluconazole and other azoles [ketoconazole, itraconazole, voriconazole, or posaconazole]), or both (atorvastatin, erythromycin/clarithromycin, dronedarone, rifampin, and phenytoin). The relative risks of non-traumatic major bleeding events were assessed for each person-quarter (3-month period per person) for patients taking each the 12 medications as compared with those not taking that medication. The risks were adjusted by the propensity score for the likelihood of the medication prescription based on demographics, comorbidities, other relevant medications, and health care utilization.

The most commonly prescribed concurrent medications were amiodarone, atorvastatin, digoxin, and diltiazem (each in 35%-55% of patients). The rate of non-traumatic major bleeding events among the entire cohort was about 4.3 per 100 person-years. Concurrent medications that were associated with an increased risk of major bleeding were fluconazole (adjusted rate ratio [RR] 2.35, 99% CI 1.8-3.07), phenytoin (adjusted RR 1.94, 99% CI 1.59-2.36), rifampin (adjusted RR 1.57, 99% CI 1.02-2.41), and amiodarone (adjusted RR 1.37, 99% CI 1.25-1.5). No significantly altered risks were found with the concurrent use of cyclosporine, diltiazem, dronedarone, verapamil, and azoles other than fluconazole. Conversely, atorvastatin, digoxin, and erythromycin/clarithromycin were each associated with a significantly reduced risk of major bleeding.

This study found that fluconazole, phenytoin, rifampin, and amiodarone were each associated with an increased risk of non-traumatic major bleeding events among patients taking DOACs for nonvalvular atrial fibrillation. The observational design limits the ability to infer a connection beyond association, and generalization to non-Taiwanese patients is limited. Further studies are needed to confirm these findings and describe why these specific medications but not others that also share metabolic pathways with DOACs were associated with an increased risk, and to describe why some medications were associated with a reduced risk. Data on possible drug interactions with DOACs are limited, and the results of this study will help guide concurrent medication choice among patients taking DOACs for nonvalvular atrial fibrillation.

For more information, see the DynaMed Plus topic Thromboembolic prophylaxis in atrial fibrillation. DynaMed users click here.


Other EBSCO Sites +