Epileptic Disord. 2026 Apr 8. doi: 10.1002/epd2.70232. Online ahead of print.
ABSTRACT
OBJECTIVES: Antiseizure medications (ASMs) can induce the activity of drug-metabolizing enzymes and drug transporters, including cytochrome P450 (CYP)2C9 and P-glycoprotein (P-gp). Our objective was to comparatively assess the effects of ASMs on exposure to clinical CYP2C9 and P-gp substrates.
METHODS: This systematic review and network meta-analysis (NMA) was registered in PROSPERO (CRD42023473609) and performed following PRISMA 2020 guidelines. MEDLINE, EMBASE, and Cochrane Library were searched until October 22, 2025, with additional searches conducted in the FDA and EMA databases and ClinicalTrial.gov. Studies were included if they were prospective and the ASM was used as monotherapy for ≥5 days. The primary endpoint was the substrate area under the curve ratio (AUCR) with/without the ASM. Treatments were ranked by P-scores (range 0-1, higher values reflect stronger induction). The point estimate for indirect pairwise comparisons was the standardized mean difference (SMD). Bias risk was assessed using the PKclin tool.
RESULTS: Twelve and six interventional pharmacokinetic studies with 227 and 97 participants were included in the CYP2C9 and P-gp NMAs, respectively. The ASM with the greatest CYP2C9 induction potential was carbamazepine (600 mg/day, P-score .78). The only statistically significant effect size estimate for CYP2C9 was obtained in the comparison between carbamazepine 600 mg/day and cenobamate 200 mg/day (SMD -.42; CI -.76, -.09). Carbamazepine (300 or 600 mg/day) was also the strongest P-gp inducer (P-scores, .79 and .55, respectively). The effects of its two doses did not differ, and 300 mg/day had a stronger effect on P-gp compared with the other ASMs.
SIGNIFICANCE: Despite variability in populations, substrate drugs, and doses, our findings demonstrate that carbamazepine is an inducer at 300 mg/day, and that ASMs can rank differently as CYP2C9 versus P-gp inducers. Therefore, the safety of ASM polytherapy cannot be extrapolated from one pathway to another for treatment selection, for example, for post-stroke epilepsy.
PMID:41948839 | DOI:10.1002/epd2.70232
