Diffusion-weighted imaging and proton magnetic resonance spectroscopy in perinatal hypoxic-ischemic encephalopathy: association with neuromotor outcome at 18 months of age.

We evaluated early diffusion-weighted imaging findings, the quantitative apparent diffusion coefficient, and magnetic resonance spectroscopy (the presence of lactate and ratios of N-acetylaspartate to total creatine and choline to total creatine) in the prediction of the 18-month neuromotor outcome of term newborns with hypoxic-ischemic encephalopathy. Conventional T1- and T2-weighted and diffusion-weighted imaging was performed in 20 asphyxiated term newborns, with additional basal ganglia magnetic resonance spectroscopy in 15 newborns between 2 and 18 days of life (mean 7.3 days). Neuromotor outcome was dichotomized into normal and abnormal for statistical analysis. Statistically significant differences in the ratios of N-acetylaspartate to total creatine, but not apparent diffusion coefficient values and ratios of choline to total creatine, were found between infants with a normal and an abnormal outcome (Mann-Whitney U-test, P = .010). There was a significant association between the presence of a lactate peak and an abnormal outcome (chi-square test, P = .017). The presence of a lactate peak for predicting an abnormal outcome had a sensitivity of 100% and a specificity of 80%, and the odds ratio was 37.4. Ischemic lesions were more conspicuous and/or extensive on diffusion-weighted imaging in all except one neonate. The presence of normal findings on both diffusion-weighted imaging and conventional magnetic resonance imaging is predictive of a normal neuromotor outcome, whereas lactate and a reduced ratio of N-acetylaspartate to total creatine in the basal ganglia, but not an apparent diffusion coefficient, are associated with an abnormal outcome at 18 months of age.