Objective: Carbamazepine, widely used in the treatment of partial and generalized tonic-clonic seizures, has been associated with life-threatening Stevens-Johnson syndrome/toxic epidermal necrolysis among some Asians. The HLA-B*1502 genotype that occurs with varying frequency among Asians is recommended for screening prior to starting carbamazepine. Our goal is to explore the cost-effectiveness of screening for the presence of this genetic allele.
Methods: We constructed a Markov model in a hypothetical cohort of adult Asian patients with epilepsy in the United States being considered for carbamazepine to investigate the cost-effectiveness of two alternative strategies: (1) no HLA-B*1502 gene allele screening and using carbamazepine and (2) HLA-B*1502 gene allele screening and starting levetiracetam in the case of a positive screen.
Results: For the lifetime horizon, HLA-B*1502 gene screening was the cost-effective choice compared to no gene screening, with an incremental cost-effectiveness ratio of $27 058 per quality-adjusted life-year (QALY), below the $50 000/QALY threshold in 99.69% of probabilistic sensitivity analyses. Although gene screening strategy was more expensive than a no screening strategy, it was more effective, yielding more QALYs, across all Asian ethnic groups.
Significance: Our analysis confirms the 2007 US Food and Drug Administration recommendation to screen for HLA-B*1502 allele before starting treatment with carbamazepine in patients of Asian ancestry in the United States.
Keywords: QALY; HLA-B*1502; Markov model; Stevens-Johnson syndrome; carbamazepine; cost; cost-effectiveness; epilepsy; genetic screening; pharmacovigilance; public health; quality-adjusted life-years; toxic epidermal necrolysis.
Wiley Periodicals, Inc. © 2019 International League Against Epilepsy.