OBJECTIVES: To correlate the abnormalities on computed tomography, magnetic resonance imaging, and bone scan with fluorinated deoxyglucose positron emission tomography (FDG-PET) in patients with progressive metastatic prostate cancer, using a lesion-by-lesion analysis, and to preliminarily explore post-treatment changes in standard uptake value (SUV) with changes in prostate-specific antigen (PSA). METHODS: A lesional analysis compared abnormalities on FDG-PET with those on CT/MRI or bone scan. Patients had rising PSA levels and progressive disease according to the imaging findings. Changes in the SUV were compared with the PSA changes in patients who had serial scans after treatment. RESULTS: One hundred fifty-seven lesions in 17 patients were examined; 134 osseous lesions were evident on PET and/or bone scan, 95 lesions (71%) were evident on both, 31 (23%) were seen only on bone scan, and 8 (6%) were seen only on PET (adjusted McNemar's chi-square = 8.32, P = 0.004). All but one of the lesions seen only on bone scan were "stable" compared with the previous bone scans. All lesions seen only on PET proved to be active disease on subsequent bone scans. Twenty-three soft-tissue lesions were present on CT/MRI or PET, or both; 9 (39%) lesions were evident on both and 14 (61%) were evident only on one imaging modality. In 9 (75%) of 12 cases in which serial PET scans were available, the SUV changed in parallel with the PSA level. CONCLUSIONS: FDG-PET can discriminate active osseous disease from scintigraphically quiescent lesions in patients with progressive metastatic prostate cancer, but it is limited in detecting soft-tissue metastases. Post-treatment changes in the SUV tend to correlate with changes in PSA.
