Abstracts

Rationale: Perinatal stroke is a significant cause of severe epilepsy, and is responsible for approximately 8% of all infantile spasms (IS). IS due to underlying structural lesions, such as stroke, often respond poorly to treatment and can frequently evolve into medically intractable epilepsy. Yet outcomes after infantile spasms due to perinatal stroke are not uniformly poor, and predictors of these outcomes are not well described. Previous case series have suggested that perinatal stroke secondary to IS is influenced by early seizure burden, as stroke presenting with neonatal seizures leads to higher rates of medically intractable epilepsy relative to infarcts presenting later in infancy (presumed perinatal ischemic stroke). Recent studies have created neuroradiologic guidelines to reliably differentiate perinatal ischemic stroke by arterial and venous etiologies and to create subtypes of both perinatal arterial strokes (PAS) and periventricular venous strokes (PVVS). It is possible that the heterogeneous outcomes seen with perinatal stroke, including the development of infantile spasms and progression to long-term epilepsy are influenced by the extent or location of perinatal injury. Methods: Therefore, we performed a retrospective review of clinical and neuroimaging data for children diagnosed with IS at a single tertiary pediatric institution between 2000 and 2014. We identified 400 patients with confirmed IS and imaging. All imaging was reviewed with our Neuroradiologist for delineation of perinatal arterial and venous infarcts. Comorbidities such as generalized hypoxic ischemic injury or meningitis were excluded. Results: From our database, we identified 9 cases of PAS and 13 cases of PPVS who developed IS, together accounting for 6% of the total IS cases. IS evolved into medically intractable epilepsy in (4/9) 44% of PAS and (7/13) 54% of PVVS. Three of these individuals, (2/9 of PAS and 1/13 of PPVS) required hemispherectomies for seizure control. Neonatal seizures were seen prior to IS in 55% (5/9) of PAS but only 7% (1/13) of PVVS, and there was no difference between the presence or absence of neonatal seizures in the progression to medically intractable epilepsy. While neonatal seizures were a more common comorbidity with arterial infarcts, there was a high rate of hydrocephalus with PVVS, as 62% (8/13) of venous infarcts developed hydrocephalus in the neonatal period and 88% of those required ventriculoperitoneal shunts. Conclusions: Both PAS and PPVS are associated with the development of IS, and the rates of transition into medically intractable epilepsy are significant in both groups. While isolated PAS appears to be a cause of IS, the high rate of comorbidities such as hydrocephalus in PPVS suggests that venous infarction may not be independently sufficient to lead to IS. In addition, both PAS and PPVS with IS are at high risk of progression to medically intractable epilepsy. More studies need to be performed to validate our findings. Funding: None