Cerebral glucose metabolism as a predictor of rehabilitation after ischemic stroke

Permanent neurological deficits after ischemic stroke are primarily determined by the location and size of an infarct, but social recovery and rehabilitation also depend on the functional status of brain tissue outside the infarct. Since neuronal loss and functional deactivation in peri-infarct tissue are reflected as changes in flow and metabolism, measurement of glucose consumption may yield an additional measure of rehabilitative capacity. Seventy-six nondiabetic patients (48 men, 28 women; mean age, 56.7 +/- 14.37 years) with a first unilateral supratentorial ischemic infarct were consecutively enrolled. At stable neurological and clinical condition 9 +/- 7.2 days after the attack, cerebral metabolic rate of glucose (CMRglu) in noninfarcted brain regions was measured by positron emission tomography of fluorodeoxyglucose. Outcome was assessed 21 to 77 (mean, 50.5 +/- 11.7) months after the stroke with a rehabilitation index for daily life activities. At time of assessment of outcome 16 patients had died (score, 0), 22 were completely recovered (score, 200), and 38 had partially improved (rehabilitation score, 5 to 195). Younger age, absence of arterial hypertension and cardiac disease, but also higher global, ipsilateral, and contralateral CMRglu were significantly related to a better final outcome (P = .001), whereas sex and neurological deficits in the subacute stage after stroke were not related to final outcome. To evaluate the significance of CMRglu further after adjustment for clinical prognostic variables, a multiple regression analysis of the effect of age and CMRglu on rehabilitation score in homogeneous subgroups of partially recovered patients was performed. It revealed a significant positive correlation of CMRglu (P = .016) with recovery in hypertensive subjects, while age was the dominant prognostic factor (P = .07) in patients with normal blood pressure. These results demonstrate that outcome after stroke is significantly influenced by several factors incapacitating brain function in addition to the ischemic attack. In addition to age and cardiac disease, hypertension is an important factor leading to widespread arteriopathy with neuronal loss and tissue damage. The significant correlation of CMRglu outside the infarct with functional recovery in hypertensive subjects probably reflects the extent of hypertensive tissue damage and subsequently reduced capacity to compensate for the focal ischemic insult.