Search

Aim: To prospectively assess the predictive value of asymmetrical myelination on MRI of the posterior limb of the internal capsule (PLIC) in newborn infants with an intraventricular haemorrhage (IVH) associated with unilateral haemorrhagic parenchymal involvement (PI), for subsequent development of a hemiplegia., Methods: 12 preterm infants (GA 25-36 wks) and 4 full-term infants were studied. Using cranial ultrasound (US), the pre-term infants were diagnosed to have an IVH with unilateral PI. The term infants presented with a porencephalic cyst (PC) on the first postnatal US, following an antenatal IVH with PI. MRI was performed at 40 wks postmenstrual age in the pre-term infants and during the first 2 weeks of life in the full-term infants, using a 1.5T magnet. Using an inversion recovery sequence, the myelination of the internal capsule was recorded as normal, abnormal or equivocal. Neurological assessment > or = 12 months disclosed the presence of a hemiplegia or asymmetry in tone pattern., Results: All 4 cases with a normal internal capsule had a normal outcome in spite of the development of a PC. All 9 cases with an abnormal PLIC developed a hemiplegia, while 1 of the 3 cases with an equivocal PLIC is normal on neurological assessment, one developed a mild asymmetry in tone and 1 a mild hemiplegia., Conclusion: While a symmetrical signal intensity within the internal capsule on MRI, performed at 40 weeks PMA, in infants with an IVH and unilateral PI appears to be strongly related to a normal outcome, an asymmetrical PLIC is an early predictor of future hemiplegia.

Unlabelled: During a 7-year-period, 1625 infants of 34 weeks gestation or less were enrolled in a prospective ultrasound (US) study. One hundred and eleven (6.8%) infants developed a large intraventricular haemorrhage (IVH) with or without unilateral parenchymal involvement (PI). Fifty-six of these 111 infants survived (50.4%) and in 23 (41%) of them a magnetic resonance imaging (MRI) study was performed beyond 12 months corrected age. There appeared to be a good agreement between neonatal ultrasound findings and MRI changes noted in infancy. Of the 10 cases with a large IVH without PI (group A), seven had a VP shunt with complete decompression of previously enlarged ventricles. Six of these seven infants had periventricular hyperintensity (PVHI) but none developed cerebral palsy (CP). Two of the ten cases without a VP shunt had irregular ventricular enlargement (VE) with PVHI in one. Both developed CP. Seven cases showed thinning of the corpus callosum. Of the 13 cases with a large IVH associated with PI (group B), the site of the PI could still be recognised on MRI and the degree of communication of the porencephalic cyst (PC) with the lateral ventricles correlated well with neonatal US findings. On MRI, VE was present in only 6 cases. Wallerian degeneration was present in 9/13 infants and all but one developed a hemiplegia. In 12/13 cases there was thinning of the corpus callosum, either focal or diffuse. PVHI was present in all infants. In 6/13 PVHI was only present around the PC. Neurodevelopmental outcome differed for both groups. CP was only present in 2/10 infants in group A, compared to 11/ 13 in group B. Global delay, in the absence of CP, was more common in infants with a large IVH than in those with associated PI., Conclusion: Combining neonatal US with MRI in infancy enhances our understanding of the long-term effects of severe haemorrhagic brain lesions, occurring in preterm infants.

Twenty-three infants with an infarct in the territory of the middle cerebral artery are reported. The diagnosis was made using cranial ultrasound in all, confirmed on postmortem in two cases and on MRI, performed during the neonatal period or in infancy, in 18 of the 20 survivors. Involvement of the main branch was present in 7 cases and three of these had a gestational age of less than 35 weeks. In the other 16 infants, involvement of a cortical branch or one or more of the lenticulostriate branches was present and all but three of these had a gestational age of 34 weeks or less. While involvement of the main branch was usually diagnosed on postnatal day 1 or 2 using ultrasound, involvement of the lenticulostriate branches was noted as a wedgeshaped echogenic lesion in the caudate nucleus, thalamus or putamen, between day 4 up till day 24, and at term age in one of the cases. Neurodevelopmental outcome of those with involvement of the main branch was disappointing as all survivors developed a hemiplegia, associated with epilepsy in two; while so far only three of the other 16 infants developed cerebral palsy, one a hemiplegia and one athetoid cerebral palsy. Global delay was present in a further three cases. Infarcts in the region of the middle cerebral artery can occur in both preterm as well as fullterm infants. Involvement of the main branch also occurred in infants with a gestational age below 35 weeks and resulted in the development of a hemiplegia in all survivors. Involvement of one of the other branches was especially common in preterm infants, who had a more favourable outcome. As the lesion in the latter group was usually not present before the end of the first week, serial ultrasound up till term age is needed in order to identify these lesions.

To test the hypothesis that cranial ultrasound correlated with post-mortem findings in neonates with hypoxic ischaemic encephalopathy (HIE), the brains of 20 infants who died after at least two real time ultrasound scans were examined. The ultrasound abnormalities detected in the periventricular/subcortical white matter, cortex or thalami were compared with the macroscopic and histological appearances. Comparing the last ultrasound scan which was performed no longer than 12 hours before the infant died, with histological data, the sensitivity and specificity for lesions in the thalamus was 100% and 83.3% respectively; for cortical lesions 76.9 and 100% respectively and for lesions in the periventricular white matter 80% and 75% respectively. The value of cranial ultrasound for detecting intracranial abnormalities in infants with HIE was considerably better than reported previously. This could mainly be attributed to the use of a 10 MHz transducer which was of critical importance to identify lesions in the superficial cortical layer.