Hemiconvulsion-hemiplegia syndrome revisited: longitudinal MRI findings in 10 children

AimHemiconvulsion-hemiplegia syndrome (HHS) is a rare severe epilepsy of infancy consisting of unilateral convulsive status epilepticus immediately followed by transient or lasting ipsilateral hemiplegia. HHS may occur either in patients with previous brain pathology or without any identified cause, so-called idiopathic HHS'. MethodWe retrospectively analysed clinical and MRI longitudinal findings of a series of 10 patients (six females, four males) presenting with HHS. Age at the study inclusion ranged from 2years 6months to 15years (mean of 5y 10mo, median 4y 2mo). After defining magnetic resonance imaging (MRI) features as typical', i.e. strictly unilateral involvement, and atypical', i.e. bilateral, we compared clinical data from both groups. Cognitive level was assessed using Brunet-Lezine or Wechsler scales. ResultsHHS occurred at a mean age of 20.5months (range 8-48mo). In all cases, status epilepticus lasted for more than 1hour and was characterised by unilateral clonic seizures followed by ipsilateral hemiplegia (persistent in five patients). Two patients in this series died: the first from multi-organ failure 2weeks after the status epilepticus and the other from a second episode of ipsilateral intractable febrile status epilepticus 3years after the first episode. Early MRI (days 1-7 from status epilepticus) showed hemispheric cytotoxic oedema in all, extending to the contralateral side for one. T2 hyperintensity in the basal ganglia was disclosed in 70% of patients and in the hippocampus in 60%. After 1month (in intermediate and chronic phases), all surviving patients but one showed hemispheric cortical atrophy corresponding to the regions involved during the early stage. Comparing clinical features of patients presenting with typical' features, to those with atypical' findings, the second group presented psychomotor delay before status epilepticus. InterpretationThis series underlines the major value of early MRI for the prompt diagnosis of HHS, and shows that involvement of subcortical structures has been underestimated. Hippocampal involvement is not constant. This article is commented on by Abernethy on pages of this issue.