Objective: Perinatal stroke commonly causes childhood neurological morbidity. Presumed perinatal ischemic stroke (PPIS) defines children presenting outside a normal perinatal period with chronic, focal infarction on neuroimaging. Infarcts are assumed to represent arterial strokes, but recent evidence suggests the periventricular venous infarction (PVI) of infants born preterm may also occur in utero and present as PPIS. Using the largest published cohort, we aimed to define arterial and PVI PPIS syndromes and their outcomes.
Methods: A PPIS consecutive cohort was identified (SickKids Children's Stroke Program, 1992-2006). Systematic neuroradiological scoring executed by blinded investigators included previously defined arterial stroke syndromes. PVI criteria included unilateral injury with at least four of the following conditions: (1) focal periventricular encephalomalacia, (2) internal capsule T2 prolongation, (3) cortical and (4) relative basal ganglia sparing, and (5) remote hemorrhage. Arterial and PVI classifications were validated and correlated with neurological outcomes (Pediatric Stroke Outcome Measure).
Results: In 59 PPISs (64% male), 94% of lesions fell within potential middle cerebral artery territories. Although arterial proximal M1 infarction was most common (n = 19; 35%), venous PVI was second (n = 12; 22%) and accounted for 75% of subcortical injuries. Motor outcomes (mean follow-up, 5.3 years) were predicted by basal ganglia involvement including leg hemiparesis, spasticity, and need for assistive devices (p < 0.01). Nonmotor outcomes were associated with cortical involvement, including cognitive/behavioral outcomes, visual deficits, and epilepsy (p < 0.01). Classification interrater reliability was excellent (correlation coefficients > 0.975).
Interpretation: Recognizable PPIS patterns predict long-term morbidity and may guide surveillance, therapy, and counseling. PVI is an underrecognized cause of PPIS and congenital hemiplegia.