Objective: Numerous genetic testing options for individuals with epilepsy have emerged over the past decade without clear guidelines regarding optimal testing strategies. We performed a systematic evidence review (SER) and conducted meta-analyses of the diagnostic yield of genetic tests commonly utilized for patients with epilepsy. We also assessed nonyield outcomes (NYOs) such as changes in treatment and/or management, prognostic information, recurrence risk determination, and genetic counseling.
Methods:We performed an SER, in accordance with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), using PubMed, Embase, CINAHL, and Cochrane Central through December of 2020. We included studies that utilized genome sequencing (GS), exome sequencing (ES), multigene panel (MGP), and/or genome-wide comparative genomic hybridization/chromosomal microarray (CGH/CMA) in cohorts (n ≥ 10) ascertained for epilepsy. Quality assessment was undertaken using ROBINS-I (Risk of Bias in Non-Randomized Studies of Interventions). We estimated diagnostic yields and 95% confidence intervals with random effects meta-analyses and narratively synthesized NYOs.Results: From 5985 nonduplicated articles published through 2020, 154 met inclusion criteria and were included in meta-analyses of diagnostic yield; 43 of those were included in the NYO synthesis. The overall diagnostic yield across all test modalities was 17%, with the highest yield for GS (48%), followed by ES (24%), MGP (19%), and CGH/CMA (9%). The only phenotypic factors that were significantly associated with increased yield were (1) the presence of developmental and epileptic encephalopathy and/or (2) the presence of neurodevelopmental comorbidities. Studies reporting NYOs addressed clinical and personal utility of testing.Significance: This comprehensive SER, focused specifically on the literature regarding patients with epilepsy, provides a comparative assessment of the yield of clinically available tests, which will help shape clinician decision-making and