Objective: To describe the technique, applications and advantages of (18)FDG PET scanning in detection, analysis and management of musculoskeletal lesions.
Design and patients: Forty-five patients (19 males,26 females) aged 9 to 81 years had radiographs, routine radionuclide scans, CT and/or MRI of clinically suspected active benign or malignant musculoskeletal lesions. (18)FDG scans with a Siemens ECAT EXACT 921 dedicated PET unit (Knoxville, Tenn.) and FWH=6 mm images acquired as a 5-6 bed examination (6 min emission and 4 min transmission) used OSEM iterative reconstruction with segmented transmission attenuation correction and a Gaussian filter (cutoff 6.7 mm). Region of interest (ROI) 3x3 pixel image analysis based on transverse whole body images (slice thickness 3.37 mm) generated Maximum Standard Uptake Values (Max SUV) with a cutoff of 2.0 used to distinguish benign and malignant lesions.
Results: Thirty-nine studies were available for SUV ROI analysis. Overall sensitivity for differentiating malignant from benign osseous and non-osseous lesions was 91.7% (22/24), overall specificity was 100% (11/11) with an accuracy of 91.7%. All aggressive lesions had a Max SUV >2.0. Data separating benign from malignant lesions and aggressive from benign lesions were statistically significant ( P<0.001) in both categories. There was no statistically significant difference in distinguishing aggressive from malignant lesions ( P, ns).
Conclusion: (18)FDG PET contributes unique information regarding metabolism of musculoskeletal lesions. By supplying a physiologic basis for more informed treatment and management, it influences prognosis and survival. Moreover, since residual, recurrent or metastatic tumors can be simultaneously documented on a single whole body scan, PET may theoretically prove to be cost-effective.