Infantile spasms: Assessing the diagnostic yield of an institutional guideline and the impact of etiology on long‐term treatment response

Neuroimaging and genetic testing have been proposed for diagnostic evaluation of infantile spasms (IS), establishing etiology in ~60% of multicenter IS cohorts. A retrospective analysis of the yield of diagnostic etiology following an institutionally established guideline for investigation/treatment of IS was conducted, and the association between etiological subgroups and sustained response to standard treatment was evaluated.Etiology of IS, neuroimaging, and genetic results were extracted from clinical records. Etiology was categorized as acquired or nonacquired, the latter including syndromic patients, nonsyndromic patients with confirmed etiology, and unknown cases. Regression analyses, using clinical variables including subtypes of etiology, were conducted to determine which factors correlated with favorable (spasm freedom at last follow-up after two or fewer standard treatments) versus unfavorable treatment outcome (refractory spasms despite two standard treatments or relapse).We included 127 IS patients (60% males) with a follow-up of 2.4 years (range = .6-5 years). All patients had neuroimaging, and 95% of patients in the nonacquired category (103 of 108 patients) had genetic testing. Etiology was identified in 103 of 127 (81%, 95% confidence interval = .73-.86). At last follow-up, 42 (33%) patients had favorable treatment outcome. No difference in treatment response was observed between acquired and nonacquired etiologies. Among patients with nonacquired etiologies, developmental delay prior to spasms onset increased the odds of unfavorable treatment outcome (p = .014), whereas a clearly recognizable dysmorphic/syndromic etiology was associated with a lower risk for treatment failure (p = .034). In nonacquired etiology without a recognizable dysmorphic/syndrome but with a genetic etiology, unfavorable treatment outcome was more likely (p = .043).Rigorous evaluation with neuroimaging and genetic testing yields an etiological diagnosis in most patients with IS. Among patients with a nonacquired etiology, those with recognizable dysmorphic/syndromic diagnosis had a higher likelihood of a favorable treatment outcome, whereas the absence of such a finding, when associated with an identifiable genetic diagnosis, was associated with unfavorable treatment outcomes.