Abstract
Objective
To characterize a cohort of children with epilepsia partialis continua (EPC) and develop a diagnostic algorithm incorporating key differential diagnoses.
Methods
Children presenting with EPC to a tertiary pediatric neurology center between 2002 and 2019 were characterized.
Results
Fifty‐four children fulfilled EPC criteria. Median age at onset was 7 years (range 0.6‐15), with median follow‐up of 4.3 years (range 0.2‐16). The diagnosis was Rasmussen encephalitis (RE) in 30 of 54 (56%), a mitochondrial disorder in 12 of 54 (22.2%), and magnetic resonance imaging (MRI) lesion‐positive focal epilepsy in 6 of 54 (11.1%). No diagnosis was made in 5 of 54 (9%). Children with mitochondrial disorders developed EPC earlier; each additional year at presentation reduced the odds of a mitochondrial diagnosis by 26% (P = .02). Preceding developmental concerns (odds ratio [OR] 22, P < .001), no seizures prior to EPC (OR 22, P < .001), bilateral slowing on electroencephalogram (EEG) (OR 26, P < .001), and increased cerebrospinal fluid (CSF) protein level (OR 16) predicted a mitochondrial disorder. Asymmetry or hemiatrophy was evident on MRI at presentation with EPC in 18 of 30 (60%) children with RE, and in the remainder at a median of 6 months (range 3‐15) after EPC onset. The first diagnostic test is brain MRI. Hemiatrophy may permit a diagnosis of RE with unilateral clinical and EEG findings. For children in whom a diagnosis of RE cannot be made on first scan but the clinical and radiological presentation resembles RE, repeat imaging every 6 months is recommended to detect progressive unicortical hemiatrophy, and brain biopsy should be considered. Evidence of intrathecal inflammation (oligoclonal bands and raised neopterin) can be supportive. In children with bihemispheric EPC, rapid polymerase gamma testing is recommended and if negative, sequencing mtDNA and whole‐exome sequencing on blood‐derived DNA should be performed.
Significance
Children presenting with EPC due to a mitochondrial disorder show clinical features distinguishing them from RE and structural epilepsies. A diagnostic algorithm for children with EPC will allow targeted investigation and timely diagnosis.
SEP