To assess the performance of 18 F-fluorodeoxyglucose (FDG) PET/computed tomography (CT) versus 99m Tc MDP bone scan in assessment of metastatic osseous disease in breast cancer patients in relation to serum markers. We reviewed PET/CT studies and bone scans for 37 patients (mean age of 55.38 ± 13.08 years) with metastatic breast cancer to bone. To assess metastatic osseous burden, we used semiquantitative scores derived from PET/CT (PMS) and bone scans (BMS). We used McNemar test to compare lesion detection between both modalities and receiver operator characteristic analysis to define the cutoff value of serum CA 15-3 that best predicts additional value for PET/CT over bone scan. In 13 patients (35.1%), more lesions or higher-intensity lesions were detected on PET/CT, while only 4 patients (10.8%) had more prominent lesions on bone scans ( P = 0.049). Additional lesions seen on PET/CT are predominantly osteolytic or medullary (early phase). Most lesions with higher uptake on bone scans appear sclerotic (late phase). CA 15-3 was positively correlated to PMS ( r = 0.386; P = 0.018) but not to BMS ( r = -0.027; P = 0.874). However, serum alkaline phosphatase was positively correlated to both PMS ( r = 0.389; P = 0.017) and BMS ( r = 0.363; P = 0.027). CA 15-3 value of >47 U/ml best predicted additional findings on PET/CT compared to bone scans (area under the curve = 0.708; P = 0.0261). FDG PET/CT detects metastatic osseous lesions during an earlier phase. A higher CA 15-3 predicts a higher metastatic burden on PET/CT but not on bone scan. Bone scans are less specific, likely by missing early lesions and detecting persistent uptake in healing sclerotic lesions.
Read full abstract