Faculty Bibliography

We have performed positron emission tomography (PET) with 2-deoxy-2[18F]fluoro-D-glucose (FDG) in eight infants and children (aged 18 days to 5 years) with medically refractory epilepsy of neonatal onset. It was hypothesized that in at least some of these infants a surgical approach (focal resection, cerebral hemispherectomy) might be of benefit in achieving seizure control, and that PET might assist in surgical selection. In three of the eight subjects, interictal PET revealed unilateral diffuse hypometabolism; following cerebral hemispherectomy in these three patients, all seizures ceased and there were no adverse effects. In one child, ictal PET showed hypermetabolism in the left frontal cortex, left striatum, and right cerebellum; a partial left cerebral hemispherectomy guided by intraoperative electrocorticography was performed, following which all seizures ceased. One infant had relative hypermetabolism in the right temporal and occipital lobes, right thalamus, and left frontal lobe on ictal PET, and EEG telemetry revealed a right occipitotemporal epileptic focus; this infant died from anesthetic complications following right occipitotemporal cortical resection. Of the three unoperated patients, one is a potential candidate for right frontal lobectomy, but the other two were not considered to be surgical candidates due to bilateral epileptogenicity. Neuropathologic correlation in our series revealed that PET is a sensitive test capable of detecting cytoarchitectural disturbances whereas CT and MRI failed in this regard. In addition, PET provides a very unique and important assessment of the functional integrity of brain regions outside the area of potential resection.

From over 100 children studied with 2-deoxy-2[18F]fluoro-D-glucose and positron emission tomography we selected 29 children (aged 5 days to 15.1 years) who had suffered transient neurological events not significantly affecting normal neurodevelopment. These 29 children were reasonably representative of normal children and provided an otherwise unobtainable population in which to study developmental changes in local cerebral metabolic rates for glucose (lCMRGlc). In infants less than 5 weeks old lCMRGlc was highest in sensorimotor cortex, thalamus, brainstem, and cerebellar vermis. By 3 months, lCMRGlc had increased in parietal, temporal, and occipital cortices; basal ganglia; and cerebellar cortex. Frontal and dorsolateral occipital cortical regions displayed a maturational rise in lCMRGlc by approximately 6 to 8 months. Absolute values of lCMRGlc for various grey matter regions were low at birth (13 to 25 mumol/min/100 gm), and rapidly rose to reach adult values (19 to 33 mumol/min/100 gm) by 2 years. lCMRGlc continued to rise until, by 3 to 4 years, it reached values of 49 to 65 mumol/min/100 gm in most regions. These high rates were maintained until approximately 9 years, when they began to decline, and reached adult rates again by the latter part of the second decade. The highest increases of lCMRGlc over adult values occurred in cerebral cortical structures; lesser increases were seen in subcortical structures and in the cerebellum. This time course of lCMRGlc changes matches that describing the process of initial overproduction and subsequent elimination of excessive neurons, synapses, and dendritic spines known to occur in the developing brain. The determination of changing metabolic patterns accompanying normal brain development is a necessary prelude to the study of abnormal brain development with positron emission tomography.

We employed positron emission tomography (PET) with 2-deoxy-2[18F]fluoro-D-glucose (FDG) to study local cerebral glucose utilization in 15 children who had Lennox-Gastaut syndrome. Our results show that LGS can be classified into four predominant subtypes, each with a distinct metabolic pattern: unilateral focal hypometabolism, unilateral diffuse hypometabolism, bilateral diffuse hypometabolism, and normal. Functional disturbances seen on FDG-PET did not always correlate with abnormalities revealed by x-ray computed tomographic scan. This classification of Lennox-Gastaut syndrome into four major metabolic subtypes not only provides a new perspective toward understanding cerebral function in this complex syndrome, but may also prove useful in the clinical management of these patients.

2-Deoxy-2[18F]fluoro-D-glucose positron emission tomography performed in human infants during development revealed progressive changes in local cerebral glucose utilization. In infants 5 weeks of age and younger, glucose utilization was highest in the sensorimotor cortex, thalamus, midbrain-brainstem, and cerebellar vermis. By 3 months, glucose metabolic activity had increased in the parietal, temporal, and occipital cortices and the basal ganglia, with subsequent increases in frontal and various association regions occurring by 8 months. These functional changes measured with positron emission tomography are in agreement with behavioral, neurophysiological, and anatomical alterations known to occur during infant development.