| Objective To evaluate the clinical value of 18F-FDG PET/CT imaging on the diagnosing and differentiating metastatic tumor of bone as well as detecting primary. To analyze PET/CT characteristics of different metastatic tumors and to evaluate the differential diagnosis value of SUVmax in different bone lesions.Materials and Methods 18F-FDG PET/CT was applied to 101 cases(55 males,46 females; mean age 63.6±14.0y) suspected with metastatic tumor of skeleton by conventional imaging examination(X-ray, CT, MRI and SPECT)..Data of PET scan were reconstructed and fused with CT slices. All of the PET images were reviewed visually and semi-quantitatively with Standardized Uptake Value. Diagnosis of skeletal metastases was mainly established on the detection of primary. Primary tumor was determined with morphological characteristics on CT slices, tracer concentration features and lesion distribution on PET images. Bone metastatic lesion was determined in destructive types, osteolytic, osteogenic or mixed, with morphological change of lesions shown and tracer distribution characteristics. Primary bone lesion and its nature was determined according to the location, amount, distribution, bone morphologically destructive characteristics and tracer distribution characteristics. Maximum standard uptake value (SUVmax) of the lesion was measured on the slices with the highest radiao activity. Final diagnosis of the patients were confirmed with clinical follow-ups or histopathology examination. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of PET/CT on skeletal metastatsese, primary malignant bone tumor and benign osseous lesion were computed at patient level and lesion level respectively.Results According to the final diagnosis,101 cases suffered from bone lesion were divided into three categories:metastatic, primary malignant and benign.1.50 cases were finally diagnosed with skeletal metastase, including 23 cases with lung cancer,7 prostate cancer,3 gastric cancer,3 lymphoma,2 renal cancer,2 hepatoma,2 double primary tumor and 1 sarcoma of uterus,1 carcinoma of sigmoid and 1 carcinoma of thyroid. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of PET/CT was 94.00%,94.12%,94.06%, 94.00% and 94.12%, respectively. The detection rate of PET/CT to identify primary tumor was 90%. There were 483 osteolytic lesions (SUVmax of 9.21±4.46) and 374 osteogenic (SUVmax of 3.80±1.56). The difference between the SUVmax of them was of statistical significance (t=19.49, P=0.000). Extra organic and tissue metastases were identified in 27 cases.2.25 cases (24.8%,25/101) were finally diagnosed with primary skeletal malignancy, including 16 cases with multiple myeloma (MM),6 primary lymphoma, 1 malignant primary fibrous histincytoma and 1 chondroma canceration. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value was 96%,93.42%,94.06%,82.76% and 98.61%, respectively. In the 16 cases of MM,996 lesions were found with SUVmax of 4.1±1.9. In the remaining 9 primary diseases,11 lesions were detected with SUVmaxof 13.1±8.0. There was statistical significance between the SUVmax of them (t=3.73, P=0.04).3.26 cases (25.7%,26/101) were finally diagnosed with skeletal benign lesion, including 15 cases of fracture,7 tuberculosis,2 vertebral hemangiomaas,1 fibrous dysplasia and 1 horacic eosinophilic granuloma. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of PET/CT was 84.62%, 100%,96.04%,100% and 94.94%, respectively. In the 26 benign bone lesion,96 lesion were detected, including 34 granuloma with SUVmax of 8.82±4.1 and 62 other benign lesions with SUVmax of 4.03±1.86. The difference between the SUVmax of them was of statistical significance (t=6.39, P=0.000).4. In the 101 cases,1,864 malignant bone lesions were detected with SUVmax of 5.5±3.8 and 96 benign lesions with SUVmax of 5.7±3.6. The difference of them was not statistically significant(t=0.39, P=0.70). In the 1,864 malignant lesions, there were 857 metastases with SUVmax of 7.2±4.5 and 1,007 primary malignant bone lesions with SUVmax of 4.2±2.4. Variance analysis showed that among osteolytic metastases, primary malignant bone tumor excluding MM and bone granuloma. The difference of SUVmax was no statistically significant. There was also no statistical significant among osteogenic metastases, MM lesions and benign lesions excluding granuloma. However, the difference between the former three groups (osteolytic metastases, primary malignant bone tumor excluding MM and bone granuloma) and the latter three(osteogenic metastases, MM lesions and benign lesions excluding granuloma) were statistically significant.5. In the 23 cases,341 osteolytic lesios were detected with SUVmax of 8.9±4.3. Sensitivity was 75.6% and specificity was 80.6% in analysis of ROC curve of SUVmax values of myeloma and multiple osteolytic metastatic tumor of bone. SUVmax of 5.65 was determined as the cut-off value to distinguish them. Difference between multiple osteolytic metastases and MM was statistically significant (t=17.84, P=0.000).Conclusion 18F-FDG PET/CT is releatively sensitive, specific and accurate in diagnosing skeletal metastase, primary malignancy and bone benign lesion. It has higher posibility to seek primary tumor and explore distribution scope of tumor lesions. PET/CT has potential value to differentiate multiple osteolytic metastase and multiple myeloma. Low to-moderate FDG uptake on PET and multiple destructive bone lesions on CT may indicate myeloma. Differentiation value of SUVmax to osteolytic metastase, primary malignant bone tumor excluding multiple myeloma and granuloma is not good as expected. Since 18F-FDG PET/CT is not tumor specific, histopathological examination may be necessary to reduce the possibility of image mis-explanation. |
PDF Full Text Request