Optimization of semi-quantification in metabolic PET studies with 18 F-fluorodeoxyglucose and 11 C-methionine in the determination of malignancy of gliomas Katalin Borbe ´ly a , Istva ´n Nya ´ry a , Ma ´ria To ´th a , Kaj Ericson b , Bala ´zs Gulya ´s c,d, * a National Institute of Neurosurgery, Budapest, Hungary b Section for Neuroradiology, Department of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden c Department of Neuroscience, Karolinska Institute, Stockholm, Sweden d Psychiatry Section, Department of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden Received 14 April 2005; received in revised form 31 January 2006; accepted 10 February 2006 Available online 6 March 2006 Abstract The treatment of the glioma patient depends on the nature of the lesion and on the aggressiveness of the tumor. The management of gliomas continues to be a challenging task, because morphological neuroimaging techniques do not always differentiate them from nontumoral lesions or high grade tumors from low grade lesions. Positron Emission Tomography (PET) offers the possibility of the in vivo quantitative characterization of brain tumors. Despite decades of useful application of PET in the clinical monitoring of gliomas, no consensus has been reached on the most effective image analysis approach for providing the best diagnostic performance under heavy-duty clinical diagnostic circumstances. The main objective of the present study was to find and validate optimal semi-quantitative search strategies for metabolic PET studies on gliomas, with special regard to the optimization of those metabolic tracer uptake ratios most sensitive in predicting histologic grade and prognosis. 11 C-Methionine ( 11 C-Met, n = 50) and/or 18 F-Fluorodeoxyglucose ( 18 F-FDG, n = 33) PET measurements were performed in 59 patients with primary and recurrent brain gliomas (22 high grade and 37 low grade tumors) in order to correlate the biological behavior and 11 C-Met/ 18 F-FDG uptake of tumors. Data were analyzed by region-of-interests (ROI) methods using standard uptake value calculation. Different ROI defining strategies were then compared with each other for two of the most commonly used metabolic radiotracers, 18 F-FDG and 11 C-Met, in order to determine their usefulness in grading gliomas. The results were compared to histological data in all patients. Both ANOVA and receiver operating characteristic (ROC) analysis indicated that the performance of 18 F- FDG was superior to that of 11 C-Met for most of the ratios. 18 F-FDG is therefore suggested as the tracer of choice for noninvasive semi- quantitative indicator of histologic grade of gliomas. 11 C-Methionine has been suggested as a complimentary tracer, useful in delineating the extent of the tumor. The best diagnostic performance was obtained by calculating the ratio of the peak 18 F-FDG uptake of the tumor to that of white matter ( p <0.001; ANOVA). This metabolic tracer uptake ratio is therefore suggested as an easily obtained semi-quantitative PET indicator of malignancy and histological grade in gliomas. D 2006 Elsevier B.V. All rights reserved. Keywords: Glioma; Malignancy; 18 F-Fluorodeoxyglucose; 11 C-Methionine; PET; Semi-quantification; Tracer uptake ratio 1. Introduction Despite optimal surgical, radiation and medical therapy, survival of patients with gliomas remains disappointing and has not changed substantially during the last decades [1–6]. In the early imaging-based evaluation of gliomas, most investigators considered Computerized Tomography (CT) and Magnetic Resonance Imaging (MRI) enhance- 0022-510X/$ - see front matter D 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2006.02.015 * Corresponding author. Department of Neuroscience, Karolinska Institute, S-171 77 Stockholm, Sweden. Tel.: +46 8 5248 7858; fax: +46 8 308 572. E-mail address: balazs.gulyas@neuro.ki.se (B. Gulya ´s). Journal of the Neurological Sciences 246 (2006) 85 – 94 www.elsevier.com/locate/jns