Pediatric stroke has become an increasingly recognized entity over the past decade. Although still relatively uncommon, the reported incidence and prevalence of pediatric stroke have increased with better data collection, improved imaging modalities, and better recognition and awareness among physicians. The incidence of ischemic stroke in childhood, for example, has increased from 0.63 per 100,000 per year (9) to 3.3 per 100,000 per year (10). This may be an underestimate, with reported incidence in some cases as high as 13 per 100,000 per year (11). If one includes the incidence of hemorrhagic stroke, the overall incidence of pediatric stroke is likely to exceed 6 per 100,000 per year (12). Stroke remains one of the top 10 causes of death in children, with a mortality rate of approximately 10% (11). Outcome, in general, is better than that seen in adults, mostly because of brain plasticity and the absence of ubiquitous underlying degenerative vascular disease such as atherosclerosis. Morbidity, however, is now recognized as a serious complication of pediatric stroke, and a majority of survivors will have residual and persistent neurologic and/or cognitive impairment. Neurologic impairment includes residual hemiparesis in about 66% of children, visual field deficits, and/or epilepsy. The recurrence risk for stroke is variable and depends on the underlying etiology and has been estimated to be 20% to 40% (13,14). The etiologies of stroke in childhood are multitudinous, and vary considerably from those seen in adults, with approximately 20% to 30% of cases remaining unresolved.

Seizures are more commonly seen as a heralding symptom in childhood stroke than in adult stroke. In childhood stroke, as compared with neonatal stroke, motor deficits are more commonly the presenting neurologic symptom than are seizures. The reported incidence of poststroke seizures and subsequent epilepsy in children has been
highly variable, based partly on population selection and small sample size. In the largest pediatric series to date (15), seizure incidence at presentation was 49% for arterial ischemic stroke and 64% for sinovenous thrombosis. Of these, 13% of the former and 23% of the latter had had seizure recurrence at follow-up. In another series (16), early seizures (within 2 weeks of presentation) were observed in 35% and 39% of ischemic and hemorrhagic stroke, respectively. A smaller study (17) found acute seizures to be the sole presenting feature of the stroke in 11% of their patients, with a further 11% presenting with hemiparesis accompanying the seizure at presentation. Nine (25%) of 35 children subsequently went on to develop epilepsy. In a series of 73 patients (18), acute seizures occurred in 49% of patients, with a further 7% presenting with early seizures (i.e., within 2 weeks of presentation). Eight percent of the cohort presented with late-onset seizures (more than 2 weeks after stroke presentation). Of those patients with seizures, two thirds had cortically based strokes. Epilepsy occurred in 28% of the patients: all with late-onset initial seizures and 50% with acute or early initial seizures. Subcortical (basal ganglia, thalamus) infarcts are also associated with seizures either as an isolated presenting feature or in combination with a hemiplegia (19). The semiology of the seizures is variable, and patients often have more than one seizure type, including focal motor, complex partial seizures, with or without secondary generalization, and, occasionally, primary generalized seizures. Status epilepticus has rarely been reported (20). The occurrence of seizure at the initial presentation of the stroke is associated with an unfavorable neurologic outcome (11,15,21).

Neonatal stroke differs from childhood stroke in a number of ways. It occurs in approximately 1 in 4000 term livebirths per year (22,23), and comprises 25% of all pediatric stroke. Two thirds of neonatal strokes have large-vessel infarcts (24), compared with childhood stroke in which more than 50% of strokes involve small-vessel territory. The anterior circulation is five times more commonly involved than the posterior circulation. The majority of neonatal strokes are ischemic (80%) rather than hemorrhagic (20%), and 60% to 65% involve the left middle cerebral artery territory (24, 25, 26, 27). Multiple infarcts are seen in 15% to 20% of cases. Approximately 12% to 14% of all newborn seizures are associated with cerebral infarction, with 80% to 90% occurring within 24 to 48 hours of the infarct (22,25,27, 28, 29, 30). Seizures are the most frequent presenting neurologic symptom. They may be obvious (generalized or focal tonic or clonic seizures) or subtle and nondescript (such as orofacial automatisms, episodes of decreased alertness).

In a large prospective series (31), 62 (69%) of 90 term infants who presented only with seizures (and without evidence of a more diffuse neonatal encephalopathy) showed magnetic resonance imaging (MRI) evidence of acute focal ischemia (35 of 62) or hemorrhagic brain injury (27 of 62). Only two of these children showed additional brain MRI evidence of an antenatal injury. Of the 245 term infants presenting with a neonatal encephalopathy, with or without seizures, only 8 (3.3%) had evidence on brain MRI scan of a focal infarction.

Neonatal seizures have been reported in more than 70% of cases of sinovenous thrombosis (32). In the Canadian Pediatric Ischemic Stroke Registry, the risk of epilepsy was 20% following an infarct caused by sinovenous thrombosis versus 15% when the stroke was caused by arterial ischemia. Interestingly, another series (24) showed no seizure recurrence in a cohort of 24 acute newborn strokes.

The electroencephalogram in neonatal stroke is highly variable and frequently normal. Abnormalities include focal or generalized slowing; focal, multifocal, or bilateral spike or spike-and-wave discharges; low-voltage rhythms; and burst suppression. Periodic lateralized epileptiform discharges (PLEDs) have also been reported in neonatal stroke in the term infant (33). The presence of an abnormal background on the electroencephalogram (24) and seizures are predictors of disability at outcome (30).

The etiology of neonatal stroke remains unknown in many instances; however, there is a growing body of evidence implicating thrombophilic abnormalities as a contributing factor to focal cerebral ischemic lesions. These abnormalities are noted in 30% to 68% of infants in some series (27,31). Some of these include genetic prothrombotic risk factors, while others are transient or acquired (acute-phase reactants) as a result of hypoxia or sepsis. Other factors predisposing to stroke include sinovenous thrombosis, possible thromboembolus from the placenta through a patent foramen ovale, sepsis, and neonatal asphyxia.

CVD is the most frequent underlying etiology of both single seizures and epilepsy in the elderly. In studies looking at both first seizure and new-onset epilepsy, in which sophisticated neuroimaging or high-resolution MRI was available (34, 35, 36, 37), the percentage of CVD patients was in the range of 25% to 39% (Fig. 35.1). It is very likely that discrete CVD lesions are often missed. Eighteen percent of patients older than 60 years of age with an unknown cause for new-onset seizures had previously undetected cerebral infarction (38). According to a recent evaluation of the large UK General Practice Research Database, 4709 patients older than age 60 years presenting with their focal seizures “of unknown etiology” were at significantly higher risk to develop stroke (p <0.0001) than was a random control group of 4709 individuals without seizures (39). This suggests that a first seizure in an elderly patient may actually be the tip of an iceberg of a subtle and otherwise undetected CVD and emphasizes the need for a rigorous standardized diagnostic protocol.

The incidence of stroke-associated seizures might be highly underestimated, as they often present as altered
mental status, sudden slowing, episodes of confusion, and unexplained episodes of loss of consciousness or memory loss. This variation of presentation is linked to their often extratemporal origin involving frontal areas, which are frequently affected by stroke (37). Within the same study, the Veterans Affair Cooperative Study, a prospective, currently ongoing, randomized, double-blind treatment study in patients older than age 60 years (37), 38.3% of patients presented with complex partial seizures, 14.3% with simple partial, 7.5% with mixed partial, and 39.9% with generalized tonic-clonic seizures, a third of which had focal onset (12.8%). In general, the leading seizure type is focal, and nearly half show secondary generalization (40). Status epilepticus was observed in 31 (19%) of 159 individuals with poststroke seizures in a cohort of 3205 patients, 4 of them presenting during the acute phase of stroke (41). A stroke-associated seizure with subsequent Todd paresis mimicking recurrent stroke is considered to be the most frequent misdiagnosis in referrals to stroke units. Experimental and clinical evidence suggests that poststroke seizures induce additional harm to the infarcted area, leading to an irreversible functional deterioration (42). The reported incidence of poststroke seizures varies from 4.4% to 42.8%, and the incidence of poststroke epilepsy ranges from 2.7% to 17%; this variation is a consequence of the confusion of terms and highly variable study designs (40).