Paediatric-tailored modified IC-CoDE approach in non-lesional D/EE-SWAS

This study aimed to characterize the cognitive and behavioural profiles of children with non-lesional Developmental and/or Epileptic Encephalopathy with Spike-Wave Activation during Slow Sleep (D/EE-SWAS) using a multidomain neuropsychological model inspired by the International Classification of Cognitive Disorders in Epilepsy (IC-CoDE), and to explore associations with clinical features and outcomes. A retrospective analysis was conducted on 28 paediatric patients assessed at clinical nadir (mean age 6 years and 4 months; range: 2 years and 7 months - 13 years and 8 months) through a standardized neuropsychological battery encompassing five domains: Language, Visuospatial, Attention-Executive, Behaviour, and Motor. Most patients exhibited multidomain impairments, with Attention-Executive (75 %) and Language (61 %) deficits being most frequent. Greater involvement of these domains was associated with poorer long-term outcomes, and more severe language impairment were associated with EEG abnormalities in the left central region. Cognitive-behavioural phenotypes were defined based on the number and severity of impaired domains through a modified IC-CoDE approach. 17/28 patients (61 %) presented generalised impairment, 5/28 (18 %) showed bi-domain impairment, 5/28 (18 %) had single-domain impairment, and 1/28 (3 %) intact cognition. These phenotypes were not significatively different in terms of age at first clinical symptom or SWAS detection, IQ scores, or clinical long-term outcome. A four-graded severity score of neuropsychological impairment allowed an exploratory cluster analysis identifying three distinct profiles: Cluster 1 (32 %) with severe, generalized impairments; Cluster 2 (43 %) with predominant executive-behavioural deficits; and Cluster 3 (25 %) with selective cognitive impairments. Cluster 1 showed significantly worse functioning at long-term assessment compared to Clusters 2 and 3 (p = 0.041). Moreover, age at first SWAS detection was significantly lower in Cluster 1 than in Cluster 3 (mean 4.1 vs 6.7 years, p = 0.035), suggesting that earlier onset may be associated with more severe neurodevelopmental disruption. These findings highlight the clinical relevance of applying a structured, multidomain approach to D/EE-SWAS, enabling the identification of cognitive-behavioural phenotypes with prognostic value and supporting the integration of domain-specific profiling in this population.