Abstracts

Rationale: As a group, antiseizure medications (ASMs) tend to be involved in drug-drug interactions. We have recently demonstrated that several newer ASMs can induce, to a considerable extent, the activity of the major drug-metabolizing enzyme cytochrome P 450 (CYP)3A, approaching the induction magnitude of older ASMs. The drug efflux transporter P-glycoprotein (P-gp) is co-induced with CYP3A. However, current literature does not always clarify the induction magnitude by ASMs, which leads to sub-optimal therapeutic decisions. Accordingly, we aimed to compare the magnitude of P-gp induction across ASMs and relate it to the magnitude of induction by the prototypical inducer rifampin to revisit ASM classification as P-gp inducers.


Methods: In this rigorous systemic review and network meta-analysis, we conducted a comprehensive search of PubMed, Embase, and Cochrane Library for works describing interactions between ASMs and P-gp substrates. Studies had to be in humans and indicate an AUC change, and the duration of induction at the final dose had to be at least five days. We additionally searched the FDA’s Clinical Pharmacology and Biopharmaceutics Reviews for each ASM and FDA-recommended clinical P-gp substrates. The primary parameter was the AUC ratio (AUCR) between the P-gp substrate +ASM and the substrate only. For external validity, rifampicin studies were added to the analysis. IRB approval was not required. PROSPERO registry number: CRD42023473609.

Results: Data meeting the inclusion criteria were available for four ASMs with two carbamazepine doses, representing six studies. Four rifampin studies were included. The ASM with the highest probability of being a P-gp inducer was carbamazepine, which did not differ itself from rifampin (Figures 1, 2); P-scores for 300 md/day carbamazepine, 600 mg/day carbamazepine, and 600 mg/day rifampin were 0.74, 0.47, and 0.6, respectively). Carbamazepine had a greater effect on P-gp than levetiracetam (P-score 0.12) and eslicarbazepine (P-score 0.21). The magnitude of P-gp induction was lesser for carbamazepine (AUCR 0.714 vs. mean 0.211) and eslicarbazepine (AUCR 0.96 vs. 0.5, respectively) as compared to CYP3A4 induction.¹

Conclusions: Our findings have direct and practical implications for the selection of co-medications for patients undergoing treatment with ASMs that are moderate CYP3A4 inducers. Drugs primarily eliminated by P-gp may be a preferable choice over those eliminated by CYP3A4, but the co-administered drug requires careful monitoring. Neither CYP3A4 nor P-gp substrates are recommended for patients treated with strong enzyme inducers. These conclusions pertain to pharmacokinetic interactions; Pharmacodynamic interactions should be considered as well, as in the case of levetiracetam and new oral anticoagulants.
¹ Cohen H. et al. Epilepsia. 2024;65,(2):445-55.


Funding: This study received no funding.