Summary: This study examines the correlation between local CMRglc (LCMRglc) alterations and clinicopatholog ical changes in a chronic middle cerebral artery (MCA) occlusion model in the cat. The left MCA was occluded for a period of 2 h. The animals were grouped into mild, moderate, and severe ischemia based on the depression of the EEG 30 min after the MCA occlusion. Follow ing release of the clip, the animals were allowed to re cover for a week during which time daily neurological examinations were performed. On the seventh day P4C]2-deoxyglucose was injected for the determination of LCMRglc• Alternative blocks were processed for histo logical evaluation in which both neuronal and phagocytic changes were graded into four categories (0 = normal to 3 = severe). LCMRglc (J.LmollIOO glmin) in the ischemic hemisphere (all histological grades) was significantly lower than the metabolic rate in comparable regions of the sham MCA occlusion group. Regions with significant phagocytosis (grade 2 and 3) invariably exhibited acti vated glucose metabolism (57.4 ± 8.4 and 105.9 ± 6.8 J.Lmol/lOO glmin, respectively), which was significantlyThe recovery process of brain tissue after an isch emic insult may be closely related to its metabolic state. Although the recent development of positron emission tomography has offered the potential for investigating glucose metabolism of the human brain Phelps et al., 1982), few studies have been made early following a stroke (Kuhl et al., 1980;Lenzi et al., 1982;Wise et al., 1983;Baron et al., 1984;Kushner et al., 1987). These findings provided some information about Abbreviations used: 2-DG, 2-deoxyglucose; LC, lumped con stant; LCMRgl c, local CMRglc; MCA, middle cerebral artery.
535higher than in regions without phagocytosis (30.4 ± 0.8 J.LmolllOO g/min). There was a significant gradient of me tabolism in the central, peripheral, and boundary zone and the non-MCA territory in the animals with severe ischemic lesions. LCMRglc in the central MCA territory was well correlated with the EEG amplitude changes (r = 0.82, p < 0.05) and the morphological score (r = -0.89, p < 0.05). The metabolic rate was significantly depressed in both the ipsilateral and the contralateral central MCA territories in comparison with the sham occlusion ani mals. The depression in LCMRglc in the contralateral hemisphere correlated well with the concomitant depres sion in the contralateral EEG amplitude. These studies demonstrate that local heterogeneous metabolic alter ations and contralateral cortical diaschisis exist chroni cally following temporary MCA occlusion and that the increases in local cerebral glucose metabolism seen in chronic stroke may be due to phagocytotic activity.