Quantitative evaluation of skeletal tumours with dynamic FDG PET: SUV in comparison to Patlak analysis.

This study was carried out to evaluate bone lesions using fluorine-18 fluorodeoxyglucose positron emission tomography (FDG PET) and to explore whether dynamic and quantitative PET data may help to differentiate benign lesions from malignant masses. Forty patients with primary bone lesions were studied. The final diagnosis was confirmed by histopathology. A 60-min dynamic FDG PET acquisition was undertaken in all subjects. From the dynamic PET images, indices such as the average and maximal standardised uptake values (SUVs), the tumour SUV-to-muscle SUV ratio (T/M) and the SUV at 60 min-to-SUV at 30 min ratio (averSUV60/30 min and maxSUV60/30 min) were produced. Patlak graphical analysis was used to obtain the influx constant (Ki), and the metabolic rate of FDG (MRFDG) was calculated. Based on the receiver operator characteristic curve, the sensitivity and specificity for each parameter in differentiating between malignant and benign lesions were evaluated. The histological results revealed 21 malignant tumours and 19 benign lesions in this group. The MRFDG and SUV indices in malignant lesions were significantly higher than those in benign lesions. However, each index showed a considerable overlap between benign and malignant lesions. Average SUV correlated positively with MRFDG (r=0.67). When a cut-off of 1.8 average SUV was used, the sensitivity and specificity for discrimination of malignancy from benign disease were 85% and 82.4%, respectively. MRFDG showed a similar sensitivity (82.4%) and a better specificity (92.9%). A combination consisting of a cut-off of average SUV (1.8) and averSUV60/30 min (1.1) resulted in an improvement of specificity to 93.3%, with a small reduction in sensitivity (81.3%) as compared with exclusive use of SUV. The results of this study indicate that a detectable difference in glucose metabolism exists between malignant and benign skeletal lesions. The static FDG uptake indices alone may not enable adequate differentiation between benign and malignant lesions. Quantitative dynamic imaging may provide more helpful information, but will not permit a definite diagnosis. The use of uptake indices may represent an alternative and interesting approach to the evaluation of bone lesions.