Profound Neonatal and Infantile Asphyxia

Despite recent advances in obstetrics and neonatal intensive care, perinatal asphyxia remains a major cause of long-term neurologic sequelae in childhood (1, 2). Improvements in the resolution of computed tomography (CT) and magnetic resonance imaging (MR) that have occurred in recent years permit more accurate localization and quantification of hypoxic-ischemic cerebral injury, which, in turn, allow more reliable prediction of outcome. The thought-provoking article by Dr Barkovich in this issue of AJNR documents the topography of profound hypoxic-ischemic cerebral injury in the newborn and young infant on MR and CT and correlates neuroradiologic abnormalities with recent theories of pathogenesis. In this article, Dr Barkovich, an eminent neuroradiologist with extensive experience in imaging of the newborn brain (1), describes the topography of neuroradiologic abnormalities in 16 infants who had sustained severe hypoxic-ischemic insult (12 newborns and four infants under 2 years of age). The CT scans, performed without contrast enhancement in six infants (two newborns and four older infants), demonstrated decreased attenuation of cerebral tissue in central regions of the brain, ie, basal ganglia, thalamus. The detailed observations on MR scans in this study are of particular interest in view of the increasing evidence that this mode of imaging is becoming the technique of choice for precise anatomical localization of hypoxic-ischemic cerebral injury. Furthermore, because normative data on MR concerning stages of myelination in the immature brain have been established recently (3-6), it has become possible to follow the evolution of cerebral lesions more accurately using this technique. In the study by Barkovich, MR scans of the 12 newborns demonstrated abnormalities that were located predominantly in lentiform nuclei, thalamic, and hippocampal regions. In contrast, in the four older infants who sustained hypoxic-ischemic insult after the newborn period, there was extensive injury of corpus striatum and cerebral cortex, with relative sparing of perirolandic regions. MR scans obtained at a later age, eg, months to years following the original insult, demonstrated atrophy in the regions outlined above as well as in the lateral geniculate nuclei and hippocampus. Based on these observations, Dr Barkovich suggests that the anatomical pattern of hypoxic-ischemic cerebral injury in newborns may be different from that in older infants. However, in view of the small number of older infants (ie, four) in this study, these conclusions must be regarded with caution. It is relevant to consider the neuroradiologic observations in the context of existing neuropathologic data in experimental animals and in human newborns. It has been demonstrated that the neuropathologic pattern of hypoxic-ischemic cerebral injury correlates closely with the nature and duration of the insult and with the level of maturation of the immature brain at the time of injury. Thus, two distinct neuropathologic patterns have been documented which correlate closely with the nature of the hypoxic-ischemic insult ( 1, 2, 7-9). Acute, total asphyxia in rhesus monkey fetuses near term results in extensive destruction of brain stem and thalamus, with relative preservation of cortex and subcortical structures. In contrast, following prolonged, partial asphyxia, there is disproportionate involvement of the cortex and subcortical regions, either diffusely , or in parasagittal watershed zones between the vascular territories of anterior, middle, and posterior cerebral arteries (7). The MR abnormalities that were documented most commonly in the study by Dr Barkovich correspond more closely to the neuropathologic pattern observed following acute, total hypoxic-ischemic insult. On the basis