Summary: With the use of positron emission tomog raphy (PET) and the 150 steady-state- [ISF]fluorode oxyglucose combined method, the local interrelationships between the cerebral metabolic rate for oxygen (CMR02) and the cerebral metabolic rate for glucose (CMRGlc) were investigated in control subjects and in stroke pa tients. In addition to the classic in vivo autoradiographic approach, a kinetic method was used to measure CMRGlc because it was expected to be more reliable in cerebral ischemia. In control subjects local coupling be tween CBF, CMR02, and CMRGlc was confirmed, and acceptable values for the CMR02/CMRGlc ratio were found; the latter, however, was lower in white matterThe recent development of independent methods for measuring in the human brain the rates of ox ygen consumption (CMR0 2 ) and glucose utilization (CMRGlc) using positron emission tomography (PET) has revived the study of the coupling be tween CMR0 2 and CMRGlc, previously restricted to the whole brain only (Finkle stein et aI., 1981; Baron et aI., 1982; Rhodes et aI., 1982). Such in vivo studies may help us understand better the con ditions required for the occurrence and the prog nostic significance of enhanced anaerobic glycol ysis in cerebral ischemia.Address correspondence and reprint requests to Dr. Baron at Service Hospitalier Frederic Joliot, CEA Departement de Biol ogie, 91406 Orsay, France.Abbreviations used: CBV, Cerebral blood volume; CMRGlc, cerebral metabolic rate for glucose; CMR02, cerebral metabolic rate for oxygen; CT, computerized tomography; GlcAV, arterio venous glucose difference; ICA, internal carotid artery; MR, metabolic ratio; OEF, oxygen extraction fraction; OM, orbito meatal; PET, positron emission tomography.140 than in gray. Uncoupling between CMR02 and CMRGlc was observed in all stroke patients, suggesting that (1) enhanced anaerobic glycolysis occurred both in reper fused recent infarcts and in chronically ischemic tissue, and (2) substrates other than blood-borne glucose were being oxidized at the borders of recent infarcts. However, methodological uncertainties presently make such obser vations only tentative. Finally, a coupled depression of CMRO, and CMRGlc was found in the contralateral cer ebellu m . Key Words: Cerebral glucose utilization-Ce rebral oxygen consumption-Oxygen-IS -Positron emission tomography. METHODS AND PATIENTS MethodsThe steady-state oxygen-IS method of measuring CBF and CMR02 (Jones et aI., 1976) was combined with the [ISF]fluorodeoxyglucose eSFDG) technique for measuring CMRGlc (Phelps et aI., 1979; Reivich et aI., 1979). A detailed account of the combined measurement has been given earlier (Baron et aI., 1982). Briefly, consecutive continuous inhalation of trace amounts of CI502 and 1502 was performed first. Once completed, 16 min (eight pe riods) were allowed to elapse before rapid (=20 s) intra venous injection of IsFDG (3-8 mCi). Generally, three contiguous head levels, parallel to the orbitomeatal (OM) line, were studied. The coincidence photons were col lected by an EC...

SUMMARY Using the '8f-fluoro-2-deoxy-d-glucose technique and positron emission tomography (PET), the local cerebral glucose utilisation (lCMRGlc) was measured in four non-demented patients with early-onset, bilateral Parkinson's disease characterised by the predominance of akinesia. The study was done twice, first in the untreated condition, and then after levodopa had been resumed. Despite a marked clinical improvement, we found no alteration in lCMRGlc between the first and second studies in any of the brain structure analysed. Compared to control values, lCMRGlc in the basal ganglia was moderately increased in both studies. These essentially negative findings agree with most previous human or animal studies, and indicate that the functional alterations in the central dopaminergic systems of patients with Parkinson's disease have metabolic correlates that are too small to be demonstrated by current PET devices.The typical symptomatic triad of Parkinson's disease, that is akinesia, rigidity, and tremor, implies that functional alterations take place in certain cerebral structures. Since energy metabolism in the brain has been shown to be coupled to function,' changes in regional energy mnetabolism specific to the functional abnormalitves of Parkinson's disease might be expected to occur. Such changes should be found in the structures deprived of their normal dopaminergic afferences, principally the striatum2 and the cortico-limbic areas3 and, as a secondary effect, in the structures receiving projections from these structures.In the attempt to use local energy metabolism as a marker of disordered dopaminergic transmission systems, several studies of the local cerebral glucose utilisation (lCMRGlc) using '4C-2-deoxy-D-glucose ('4CDG) autoradiography in rats with unilateral destruction of the dopaminergic systems have failed to show any consistent pattern of changes,4-9 although some reported striking alterations.4 5 Human studies of Parkinson' s disease patients have been, for methodological reasons, restricted to 2-dimensional '33xenon regional cerebral blood flow (rCBF)'0-'2 or Address for reprint requests: JC Baron, MD, Service Hospitalier Fr6d&ric Joliot,

Neuropsychological testing was carried out and the rate of oxygen metabolism in the brain was measured by PET in 15 highly selected patients with type 1 diabetes. The aim was to investigate the impact on the brain of hypoglycaemic comas resulting from insulin treatment. No significant difference was found between nine patients with a history of more than 10 hypoglycaemic comas and six others who denied any history of such events. These data suggest that intensified insulin treatment, although increasing the frequency of hypoglycaemic coma, may not always be harmful for the brain. This may be explained by the limited duration of hypoglycaemic coma induced by conventional insulin treatment.

Nine patients who had suffered strokes were examined between 10 and 34 days after onset using positron emission tomography. DMO labeled with carbon 11 was used to evaluate brain acid-base balance, and the oxygen-15 inhalation technique was used to measure regional cerebral blood flow, the oxygen extraction fraction, and cerebral metabolic rate for oxygen. [11C]DMO concentration and oxygen metabolism variables were measured in the infarcted area and in the symmetrical region in the contralateral cerebral hemisphere. [11C]DMO concentration was found to be unchanged or slightly increased in five cases and markedly increased in four cases. The apparent increase in tissue pH can be explained by the presence of a large extracellular fluid space with a pH nearly identical to that of brain plasma, or by an increase in intracellular pH, or by both phenomena. The change in [11C]DMO concentration in the infarcted area relative to that in the normal tissue was independent of the change in blood flow. Cerebral metabolic rate for oxygen was decreased in all cases. The increase in [11C]DMO concentration in the infarcted area was linearly correlated with the decrease in the oxygen extraction fraction in the same region; that is, it was correlated with the occurrence of perfusion in excess of metabolic demand. The overabundant local perfusion could play a role in the decreased H+ content.