Effect of drug interactions with apixaban on clinical outcomes in cancer patients with venous thromboembolism.

INTRODUCTION

It is unclear how drug-interaction with apixaban influences recurrent venous thromboembolism (VTE) and bleedings in cancer patients.

METHODS

A post-hoc analysis of a single-arm interventional clinical trial on apixaban treatment of cancer patients with VTE to investigate whether the occurrence of any of the endpoints could be associated with the concurrent use of an interacting drug. Drugs taken by the patients during the trial period were categorized as either increasing bleeding risk, increasing thrombosis risk, both or neither.

RESULTS

298 patients were divided into groups based on whether they used no interacting drugs (controls, n=74), drugs increasing bleeding risk (n=55), drugs increasing thrombosis risk (n=8), or both (n=161). Odds ratios (OR) were calculated for recurrent VTE, clinically relevant non-major bleeding (CRNMB), and major bleeding during the 36-month follow-up period. Each patient took a median of 13 different drugs over the study period. 67% of the patients used drugs expected to both increase bleeding and thrombosis. The use of fluconazole appeared associated with CRNMB (OR 3.6, 95% confidence interval (CI) 0.99-13), but not with major bleeding (OR 0.56, 95% CI 0.06 - 4.8). Non-steroid anti-inflammatory drugs were not associated with CRNMB (OR 1.0, 95% CI 0.25-4.1) or major bleedings (OR 0.72, 95% CI 0.14 - 3.6). Use of antiplatelet therapy was not associated with CRNMB (OR 0.75, 95% CI, 0.22 - 2.58) or major bleeding (OR 0.2, 95% CI, 0.02-1.6). There were no major bleedings in 23 patients using aprepitant nor in the 10 patients taking macrolides. We found no association between drugs and recurrent VTE, except that there were no recurrent VTE in 19 patients using bevacizumab.

CONCLUSIONS

Despite the high number of drugs taken that could potentially interact with apixaban, none were found to clearly influence clinical outcomes, except that fluconazole may increase the risk of CRNMB.