Basal Ganglia and Frontal Lobe Glucose Metabolism A Reproducibility Positron Emission Tomography Study ABSTRACT Pos1tron emission tomography (PET) w1th '"F-fluoro-2-deoxy-D- glucose (FDG) 1s frequently used to study the metabolic correlates of movement and mental d1sorders. These stud1es generally focus on changes in the frontal cortex and the basal gangl1a. The re- producibility of glucose metabol1sm est1mates in these structures was tested 1n 13 normal subjects stud1ed at rest usmg a standard and s1mple protocol. A reproducible dorsoventral metabol1c gra- dient was demonstrated 1n the frontal cortex. Such a grad1ent was not present m the basal gangl1a when the upper region of interest 1n the caudate nucleus, where the lower metabol1c rate of glucose was probably attnbutable to part1al volume effects, was not considered. Absolute values of glucose metabolic rates varied by 6.4 to 12.5% m the frontal cortex and by 6.8 to 14.7% in the basal gangl1a. Vamt1ons 1n normal1zed values in the basal gangl1a ranged from 4.0 to 8.6%. The number of sub- jects requ1red to detect statistical differences in group com- panson or m test-retest stud1es was calculated for different anticipated levels of change. W1th the variability detected in this expenment, less than 10 subjects were expected to be suff1c1ent to detect a 15% change in most regions and in both types of stud1es. Goldman S, Dethy S, Lotstra F, B1ver F, Stanus E, Wikler D, Hildebrand J, Mendlew1cz J, Luxen A. Basal ganglia and frontal lobe glucose metabolism: a reproducibility positron emission tomography study. J Neuroimag 1995;5:219-226 Positron emissiOn tomography (PET) using 1 xF-ftuoro-2- deoxy-o-glucose (FOG) as a tracer allows evaluation of the regional distribution of glucose metabolism in the human brain. Today, PET with FOG (PET-FOG) is mainly used for Received Aug 15, 1994, and in revised form Oct 14. Accepted for publication Oct 17, 1994. Address correspondence to Dr Goldman. Serge Goldman, MD PET/Biomedical Cyclotron Unit ULB-H6pital Erasme 808, route de Lennik B-1 070 Brussels, Belgium Sophie Dethy, MD PET/Biomedical Cyclotron Unit Service de Neurologie ULB-H6pital Erasme Lotstra, MD Service de Psychiatrie ULB-H6pital Erasme Biver, MD PET/Biomedical Cyclotron Unit Service de Psychiatrie ULB-Hopital Erasme Etienne Stanus, PhD PET/Biomedical Cyclotron Unit ULB-Hopital Erasme David Wikler, MS PET/Biomedical Cyclotron Unit ULB-Hopital Erasme J. Hildebrand, MD Service de Neurolog1e ULB-Hopital Erasme Julien Mendlewicz, MD Service de Psych1atne ULB-Hopital Erasme Andre Luxen, PhD PET/Biomedical Cyclotron Un1t ULB-Hopital Erasme the detection of changes in patterns of local cerebral meta- bolic rates of glucose (ICMRglu) in relation to neuropsy- chiatric diseases such as Alzheimer's disease and focal epilepsy 111. In addition to Huntington's disease, for which PET-FOG proved useful early [21. neuropsychiatric diseases certainly or possibly involving the basal ganglia have been studied by PET-FOG. Most of these diseases were either movement disorders 13-51 or mental disorders with phar- macological clues for aminergic system implications such as schizophrenia, depression, eating disorders, and obsessive- compulsive disorders 16-12 [.These studies often focused on metabolic changes in striatal structures and in the prefrontal cortex, the brain structure implicated in the most complex human behaviors [ 13, 14]. Authors encountered two prob- lems when applying PET-FOG in this context. The first problem was related to the size. close to the resolution of most PET scanners, and the complex anatomy of some of the structures studied such as the caudate nucleus [15, 16]. The second problem was that reproducibility of glucose me- tabolism estimates in these structures is not specifically es- tablished even though this parameter is essential for the evaluation of natural evolution [17] or response to therapeu- tic interventions [17, 18]. For this study PET data from the basal ganglia and the frontal cortex of normal subjects were analyzed to elaborate an efficient strategy of image analysis applicable in any other center and, using this strategy, to determine the reproduci- bility of the glucose metabolism estimates. Materials and Methods Measurements of lCMRglu were performed on fasted sub- jects in a supine resting state with eyes closed and ears un- plugged, in a room with dimmed light, no conversation, and low ambient noise, mostly from the scanner gantry fans. Thirteen healthy (mean age: 30 yr; range: 18-38 yr) vol- unteers, 8 men and 5 women, participated in the study. They did not report familial or personal major psychiatric disor- ders and had no history of medical or neurological disease, Copyright© 1995 by the American Society of Neuroimaging 219