CLINICAL INVESTIGATIONS Correlation of Regional Metabolic Rates of Glucose with Glasgow Coma Scale After Traumatic Brain Injury

After traumatic brain injury (TBI), subcortical white matter damage may induce a functional disconnection leading to a dissociation of regional cerebral metabolic rate of glucose (CMRglc) between the cerebral cortex and deeper brain regions. Therefore, thalamic and brain stem CMRglc may have a closer correlation than does the cerebral cortex with depth of coma after TBI. Methods: Eleven adult healthy volunteers and 23 adult patients with TBI (median initial Glasgow Coma Scale score [GCSini], 8) underwent 18F-FDG PET within 5 d after injury. The CMRglc of cortical areas (remote from hemorrhagic lesions), striatum, thalamus, brain stem, cerebellar cortex, and whole brain was compared with severity of injury and the level of consciousness evaluated using GCSini and the Glasgow Coma Scale score at the time of PET (GCSpet). Results: The regional CMRglc of the brain stem is relatively unaffected by the TBI. Compared with healthy volunteers, TBI patients exhibited significantly depressed CMRglc in the striatum (3.9 1.3 vs. 5.1 0.9 mg/100 g/min, P 0.05) and thalamus (3.1 1.0 vs. 4.3 0.9 mg/100 g/min, P 0.05). CMRglc levels were not statistically lower in the cerebellum (2.9 0.8 vs. 3.5 0.8 mg/100 g/min, P NS) and brain stem (2.5 0.5 vs. 2.6 0.5 mg/100 g/min, P NS). However, compared between comatose and noncomatose patients, CMRglc values in the thalamus (2.7 0.7 vs. 3.6 1.2 mg/100 g/min, P 0.05), brain stem (2.2 0.4 vs. 2.8 0.5 mg/100 g/min, P 0.01), and cerebellar cortex (2.6 0.5 vs. 3.4 1.0 mg/100 g/min, P 0.05) were significantly lower in comatose patients. When individual values of regional CMRglc were plotted against posttraumatic level of consciousness, CMRglc values for the thalamus, brain stem, and cerebellum significantly correlated with the level of consciousness at the time of PET (r 0.58, P 0.01; r 0.66, P 0.01; r 0.64, P 0.01, respectively). CT or MRI findings were normal for the analyzed structures except for 3 patients with diffuse axonal injury of the brain stem. The presence of shear injury was associated with poor GCSini (P 0.05.) but was not related to GCSpet and brain stem CMRglc. Conclusion: A PET investigation using 18F FDG demonstrated a significant difference in glucose metabolism in the thalamus, brain stem, and cerebellum between comatose and noncomatose patients acutely after TBI. The metabolic rate of glucose in these regions significantly correlated with the level of consciousness at the time of PET.