Objective: We describe the clinical and imaging studies of 11 full-term babies with neonatal stroke. We classify the neonatal non-hemorrhagic strokes as thrombotic, embolic, or global vascular insufficiency and determine if this classification is improved by adding magnetic resonance spectroscopy (MRS) and diffusion-weighted imaging (DWI) to conventional magnetic resonance imaging (MRI) and magnetic resonance angiography (MRA).
Methods: Clinically, eight of the 11 babies presented with seizures, one with apnea, and two with lethargy. Conventional MRI and DWI were used to classify each infarct as being either borderzone or vascular distribution. The location of infarction revealed the presumed vascular pathophysiology.
Results: Infants were classified as having either embolic (bilateral middle cerebral artery,n=1), global ischemic (bilateral borderzone, n=2), or thrombotic infarction (unilateral middle cerebral artery, n=7; bilateral posterior cerebral arteries, n=1). DWI and MRS detected a small infarct better than conventional MRI in one patient. MRA showed abnormal intracranial arteries in three, all of who were in the thrombotic group. Even though MRS was more sensitive than conventional MRI in detecting ischemia/infarction in one patient, in another there was no detectable lactate in the stroke region found on conventional MRI. Clinical presentation was similar in global ischemia and focal infarctions, but newborn stroke was more likely to present with lateralizing focal motor seizures. Seizures were the most common presenting sign, with a paucity of other focal neurological deficits.
Conclusion: MRI is the best approach to determine stroke pathophysiology. Brain infarction frequently presents with seizures. We speculate that the location and distribution of infarction might determine stroke timing, pathophysiology and outcome. Ongoing clinical studies will likely clarify this speculation.