MR imaging of perinatal brain damage: comparison of clinical outcome with initial and follow-up MR findings.

BACKGROUND AND PURPOSE The purpose of our study was to determine whether MR studies in the neonatal period are predictive of the neuroradiologic sequelae and clinical outcome in premature and term infants with perinatal brain injury. METHODS Thirty subjects (15 premature and 15 term infants) with abnormalities revealed by initial MR studies were reexamined approximately 1 year after birth with both MR imaging and a neurologic assessment. All initial MR studies were performed between 35 and 45 weeks corrected age in premature infants and within 28 days of life in term infants. The initial MR studies were evaluated for deep gray matter involvement, hemispheric parenchymal change, intracranial hemorrhage, and periventricular signal and/or morphologic changes. These MR findings were compared with the follow-up MR findings and with the neurologic outcome. RESULTS The development of cerebral palsy in premature infants was related to the following initial MR findings: subependymal hemorrhage associated with parenchymal destruction, periventricular signal alteration with irregularity of the ventricular wall, and widespread cerebral infarction. These MR findings were predictive of the subtypes of cerebral palsy. In term asphyxiated infants, T2 signal alterations of the deep gray matter rather than T1 shortening and diffuse involvement of the hemispheres were predictive of an unfavorable outcome. Both in term and premature infants, focal hemispheric parenchymal lesions alone (including infarction and intracerebral, subdural, intraventricular, and subarachnoid hemorrhage) did not produce poor outcomes. CONCLUSION MR studies performed at or near term in either premature or term infants with perinatal brain damage are effective in predicting both late neuroradiologic and clinical outcome.