Diagnostic delay in infants with very early‐onset rare and ultra‐rare genetic epilepsies

AIM: To evaluate time to diagnosis in infants with very early-onset genetic epilepsies, identify contributing factors to diagnostic delay, and assess the impact of a definite diagnosis on clinical management. METHOD: Infants with genetic epilepsies and seizure onset before 3 months of age were retrospectively identified from a cohort of 500 children with epilepsy followed at Saint-Luc University Hospital, Belgium, between 2018 and 2025. A correct diagnosis was defined as recognition of electroclinical phenotype before or at genetic confirmation. Diagnostic delay was defined as time to diagnosis exceeding 4 weeks from seizure onset. RESULTS: Thirty infants with KCNQ2/3-, SCN2A-, SCN1A gain-of-function, CDD-, BRAT1-, KCNT1-, FGF12-, SMC1A-, STXBP1-, GNAO1-, PHF6-, CELF2-, and 2q24.3 microduplication-related epilepsies were included. Median time to diagnosis was 55 days (interquartile range 30-427). Twenty-four patients (80%) experienced more than 4 weeks' delay. The correct diagnosis prompted change in management in 25 infants (83%). Contributing factors included deferred diagnosis by paediatric neurologists (79%), unreported/undescribed phenotypes in ultra-rare epilepsies (33%), misinterpretation of ictal events by paediatricians (17%), and parental underestimation of events (13%). INTERPRETATION: Diagnostic delay in infants with rare epilepsies is a major issue, even in a universal health care context, mainly driven by mischaracterization of seizure semiology by paediatric neurologists.