The Effect of Levetiracetam Compared with Enzyme-Inducing Antiseizure Medications on Apixaban and Rivaroxaban Peak Plasma Concentrations

Background and Objective Post-stroke epilepsy represents an important clinical challenge as it often requires both treatment with direct oral anticoagulants (DOACs) and antiseizure medications (ASMs). Levetiracetam (LEV), an ASM not known to induce metabolizing enzymes, has been suggested as a safer alternative to enzyme-inducing (EI)-ASMs in patients treated with DOACs; however, current clinical guidelines suggest caution when LEV is used with DOACs because of possible P-glycoprotein induction and competition (based on preclinical studies). We investigated whether LEV affects apixaban and rivaroxaban concentrations compared with two control groups: (a) patients treated with EI-ASMs and (b) patients not treated with any ASM. Methods In this retrospective observational study, we monitored apixaban and rivaroxaban peak plasma concentrations (C-max) in 203 patients treated with LEV (n = 28) and with EI-ASM (n = 33), and in patients not treated with any ASM (n = 142). Enzyme-inducing ASMs included carbamazepine, phenytoin, phenobarbital, primidone, and oxcarbazepine. We collected clinical and laboratory data for analysis, and DOAC C(max )of patients taking LEV were compared with the other two groups. Results In 203 patients, 55% were female and the mean age was 78 +/- 0.8 years. One hundred and eighty-six patients received apixaban and 17 patients received rivaroxaban. The proportion of patients with DOAC C-max below their therapeutic range was 7.1% in the LEV group, 10.6% in the non-ASM group, and 36.4% in the EI-ASM group (p < 0.001). The odds of having DOAC C-max below the therapeutic range (compared with control groups) was not significantly different in patients taking LEV (adjusted odds ratio 0.70, 95% confidence interval 0.19-2.67, p = 0.61), but it was 12.7-fold higher in patients taking EI-ASM (p < 0.001). In an analysis in patients treated with apixaban, there was no difference in apixaban Cmax between patients treated with LEV and non-ASM controls, and LEV clinical use was not associated with variability in apixaban C-max in a multivariate linear regression. Conclusions In this study, we show that unlike EI-ASMs, LEV clinical use was not significantly associated with lower apixaban C-max and was similar to that in patients not treated with any ASM. Our findings suggest that the combination of LEV with apixaban and rivaroxaban may not be associated with decreased apixaban and rivaroxaban Cmax. Therefore, prospective controlled studies are required to examine the possible non-pharmacokinetic mechanism of the effect of the LEV-apixaban or LEV-rivaroxaban combination on patients' outcomes.