Time-resolved phenotyping of seizure evolution in developmental and epileptic encephalopathy: insights and gaps from...

In developmental and epileptic encephalopathies (DEEs), the seizure phenotype is rarely static. Infantile epileptic spasms typically subside and are often replaced years later by other seizure types, and in several genetically defined DEEs seizure-free intervals precede seizure recurrence. Although these transitions are linked to long-term outcome, little is known about the molecular and cellular events that accompany them, because the interval is clinically silent and brain tissue is inaccessible. Animal studies have only partly filled this gap: most DEE models have been characterized at a single age, and models with continuous seizures provide no interval to investigate. We recently performed time-resolved deep phenotyping of Cyfip2^+/R87C mice, which recapitulate neonatal seizure-like events, a prolonged seizure-free interval, and adult-onset spontaneous recurrent seizures with premature death, using transcriptomic, ultrastructural, proteomic, and lipidomic analyses across four postnatal stages. Here we summarize the temporal landscape: excitatory synapse number falls while remaining synapses relocate to dendritic shafts, oligodendrocyte-lineage cells, microglia, and astrocytes are engaged sequentially, and lipid droplets accumulate progressively in astrocytes. We also highlight what remains unresolved. The highest-resolution analyses were performed only after seizure recurrence, illustrating a tradeoff between analytical depth and temporal coverage, so the juvenile interval preceding recurrence remains the central knowledge gap. Even so, the emerging picture is that the seizure-free interval is not a period of quiescence but one of active neuron-glia and metabolic remodeling. Defining this interval more precisely, and identifying biomarkers for patient staging, may transform a clinically silent phase into a well-defined therapeutic window for stage-specific intervention.