Electroclinical classification of idiopathic generalized epilepsy syndromes at initial evaluation: A prospective multicenter study
Abstract Objective To determine the extent to which electroclinical information available at initial evaluation allows classification of idiopathic generalized epilepsy (IGE) syndromes, and to assess the contributions of seizure semiology and age at seizure onset to early syndromic diagnosis. Methods We prospectively analyzed a cohort of patients with new‐onset seizures who underwent a structured clinical evaluation, including detailed history‐taking, 3‐h video‐EEG, and epilepsy‐protocol MRI. Seizure semiology was assessed at three levels: index seizure, initial seizure type, and seizure types identified at initial evaluation. Syndromic classification followed International League Against Epilepsy criteria. Final electroclinical diagnosis, established after longitudinal follow‐up, served as reference. Within patients ultimately diagnosed with one of the four IGE syndromes, baseline data were used to determine whether the final syndrome could be assigned at initial evaluation. Results Of 2699 patients evaluated, 498 (18.4%) met criteria for genetic generalized epilepsy, of whom 401 (80.5%) were classified as having one of the IGE syndromes. Based on initial evaluation, the correct syndrome could be assigned in 366 patients (91.3%), whereas 7 (1.7%) would have been misclassified and 28 (7.0%) were not assignable. Non‐assignability was confined to absence epilepsies, reflecting overlap between childhood (CAE) and juvenile absence epilepsy (JAE), whereas misclassification occurred only in juvenile myoclonic epilepsy (JME) presenting with isolated generalized tonic–clonic (GTC) seizures. Classification based on the index seizure alone showed limited accuracy (63.6%), whereas incorporation of seizure types identified through structured history improved classification to 95.0%. Myoclonic seizures were highly specific for JME, whereas absence and GTC seizures required integration with additional features. Age at seizure onset strongly differentiated CAE from JAE (AUC 0.935). Significance Electroclinical classification of IGE syndromes is frequently achievable at initial evaluation when based on seizure semiology and age at onset. When classification is not possible, limitations follow predictable electroclinical patterns, supporting a structured, history‐driven diagnostic approach. Plain Language Summary This study examined how accurately doctors can identify specific types of generalized epilepsy when patients are first evaluated. Among 401 patients with generalized epilepsy, the correct epilepsy syndrome could be identified in more than 90% at the first visit. Looking beyond the first seizure and asking about all seizure types together with age at seizure onset greatly improved diagnosis. While some overlap exists between specific types (like childhood versus juvenile absence epilepsy), a thorough clinical evaluation from the start provides reliable diagnostic direction for patients and supports early and accurate diagnosis.