ORTICOSTEROIDS are commonly administered, often for prolonged periods, to patients with brain tu- mors for the symptomatic relief of cerebral edema. Common consequences of long-term corticosteroid ther- apy include centripetal obesity with cushingoid facies, abdominal striae, disordered sleep patterns, and neuro- psychiatric disturbances ranging from mild behavioral changes to major psychoses. 10–12,25,33 Adrenal glucocorti- coids are known to affect cell glucose utilization through- out the body and studies have shown there is an effect on brain–glucose utilization, although the mechanism of neu- ropsychiatric abnormalities in patients with long-term cor- ticosteroid therapy is unclear. 26,33 In preliminary work 18 we noted by using positron emission tomography (PET) with the glucose analog [ 18 F]fluoro-2-deoxyglucose (FDG) that there was a marked reduction in cerebral glucose utiliza- tion in patients with Cushing’s disease who had high cir- culating levels of endogenous corticosteroids. In the present study our hypothesis was that exogenous corticosteroids reduce cerebral glucose metabolism (CMR glu ). Thus we measured glucose metabolism, using FDG-PET, in patients with brain tumors who displayed cushingoid symptoms from prolonged corticosteroid ad- ministration. However, because of potential effects of the brain tumor on ipsilateral cerebral structures and the premise that an effect from corticosteroids would be gen- eralized, we measured glucose metabolism in the unin- volved cerebral hemisphere. In addition, we compared values for glucose metabolism in the cushingoid patients to those obtained in treated and untreated brain tumor pa- tients who were not cushingoid and to values obtained for normal subjects because of the potential for confounding variables such as effects of radiotherapy, anticonvulsant medication, and transhemispheric functional disconnec- tion in patients. J. Neurosurg. / Volume 83 / October, 1995 J Neurosurg 83:657–664, 1995 Decreased cerebral glucose metabolism in patients with brain tumors: an effect of corticosteroids MICHAEL J. FULHAM, M.B.B.S., F.R.A.C.P., ARTURO BRUNETTI, M.D., LUIGI ALOJ, M.D., RAMESH RAMAN, M.D., ANDREW J. DWYER, M.D., AND GIOVANNI DI CHIRO, M.D. Neuroimaging Branch, National Institute of Neurological Disorders and Stroke, and Department of Diagnostic Radiology, Clinical Center, National Institutes of Health, Bethesda, Maryland  The authors measured cerebral glucose metabolism (CMR glu ) using [ 18 F]fluoro-2-deoxyglucose (FDG) positron emission tomography (PET) in patients with brain tumors to evaluate the effect of exogenous corticosteroids (in this instance, dexamethasone) on glucose metabolism. Fifty-six FDG-PET studies obtained in 45 patients with unilateral supratentorial brain tumors were analyzed. Patients with brain tumors were divided into three groups: 1) patients with cushingoid symptoms, who had been treated with combinations of radiotherapy and chemotherapy taking oral dexa- methasone; 2) patients not taking dexamethasone but treated with radiotherapy; and 3) patients not taking dexametha- sone who had not been treated with radiotherapy. Serial FDG-PET scans were obtained in eight of the cushingoid patients. Glucose metabolism was measured in the contralateral cerebral and ipsilateral cerebellar hemispheres in patients and compared to measurements taken from 19 normal volunteers. The authors found that in the cushingoid brain tumor patients there was a marked reduction in CMR glu compared to normal volunteers and other brain tumor patients (Kruskal–Wallis test; p 0.001). In the majority of patients who had serial FDG-PET scans, there was a decline in glucose metabolism over time and in one patient, in whom dexamethasone was reduced in dosage, there was a sub- sequent increase in CMR glu . The authors conclude that there is a generalized reduction in CMR glu in brain tumor patients taking dexamethasone compared to other brain tumor patients and normal volunteers, and that this effect is independent of radiotherapy, concurrent anticonvulsant medication, and transhemispheric functional disconnection (transhemispheric diaschisis). KEY WORDS • brain neoplasm • [ 18 F]fluoro-2-deoxyglucose positron emission tomography • glucose metabolism • corticosteroids C 657