Response to first antiepileptic drug trial predicts health outcome in epilepsy

Purpose: Failure to respond to the initial antiepileptic drug (AED) is a predictor of increased risk of pharmacoresistant epilepsy. Whether response to the first AED also predicts adverse health outcomes is unknown. Methods: This longitudinal study compared rates of major adverse health outcomes (loss of driving privileges, unemployment, divorce/separation, injury, emergency room admission, hospitalization, and death) in 33 patients who failed the first AED (cases) and 30 patients who became seizure-free with the first AED (controls). Patient data were obtained by chart review and confirmed through a structured interview with each subject at 5-7 years after starting AED treatment. We also assessed between-group differences in quality of life, depression, and adverse AED effects by using standardized instruments completed by each subject at the end of follow-up. Key Findings: The number of major adverse health outcomes was similarly high during the first year of AED treatment [mean ± standard deviation (SD) 2.64 ± 0.99 for cases and 2.50 ± 1.14 for controls], but thereafter decreased to a greater extent in controls than in cases (p < 0.001). Controls had a higher cumulative probability of experiencing ≥1 year free from major adverse health outcomes compared to cases (p = 0.002). Two cases died during the follow-up, both of sudden unexpected death. Cases had worse quality of life ratings than controls, whereas no significant between-group differences were found for measures of depression and adverse AED effects. In a post hoc analysis limited to cases, patients who became seizure-free with subsequent AED treatments showed for the first 4 years major adverse health outcome rates similar to those recorded in patients with persisting seizures. After 4 years, however, cases who achieved late seizure freedom tended to show a more favorable outcome. Significance: Patients with epilepsy failing the initial AED trial are at increased risk of experiencing adverse health outcomes, at least for the first 4 years after diagnosis. Incorporating these findings into clinical decision making may aid in reducing delays in surgical referrals for pharmacoresistant epilepsy.