| Objective: To investigate the clinical value of 18F-FDG PET/CT for breast cancer in differentiating breast malignant tumor from benign masses and its clinical staging.Methods: 27 women with 31 untreated primary breast masses examined with 18F-FDG PET/CT (Discovery LS4 PET/CT) in our center between Jun,2003 and Sep,2004 were studied retrospectively. To avoid breast compression and imaging artifacts, a special device allowing the breasts to hang freely was placed on the examination table for the patients to prostrate themselves onto it. After a dose of 5.55MBq/kg 18F-FDG with a radiochemical purity in excess of 95% was administered intravenously and a 40-minut-rest, Using a spiral plus 4scanner an plain scan was performed from the cranial end of the sternum to the infromammary fold, with a collimation of 5mm, table speed of 11.25mm/s,140kV, 80mA. Subsequently, PET metabolic imaging was performed initially. The field of view involved bi-breast, axilla and supclavicle zones. Emission scans were obtained with a total of 8-min acquision time for each position using 2D mode. After the calculation of the imaging count data in terms of iterative reconstruction algorithm, The three dimensions (axial, coronary, and sagittal )images and fused images of PET/CT were aquired.The questionable lesions were delayed for 3h plus a second1 ftscan. The F-FDG PET/CT images were analyzed semi-quantitatively and visually by three doctors who knew well both nuclear medicine and CT images and three radiologists analyzed the CT images together. All the lesions were confirmed histopathologilly by surgery or biopsy within 10 days after the examination of PET/CT.The findings of both modalities were compared with pathological results and were analyed statistically.Results: 31 breast tumors and 91 axillary lymph nodes were both founded by PET/CT and CT. Among these lesions, 21 breast cancer(2 cases of infiltrating lobular carcinoma ,1 case medullary carcinoma,2 cases of intraductal carcinoma, 16 cases of infiltrating ductal carcinoma)and 66 axillary lymph node metastasis were showed by pathology and the others were showed to be benign lesions. There were 10 cases of benign lesions including 1 case of sclerosing adenosis,lease of fibroadenoma with ductal epithelium hyperplasia,2 cases of lobule hyperplasia,and 6 cases of fibroadenoma. PET/CT allowed discrimination between 10 benign and 21 malignant breast masses with a sensibility of 80.95%, specificity of 90% and positive predictive value of 94.44% , but CT 90.48%, 60% and 82.61% respectively. And for the axillary lymph node metastasis, the sensibility, specificity and positive predictive value of PET/CT were 89.39%, 88% and 95.16% , but CT 86.36%, 52% and 82.61% respectively.In addition,different types of breast carcinoma had different SUV(standardized uptake value) value.The correlation between the findings of PET/CT and pathological results on axillary lymph status is positive (Pearson' correlation coeffient=0.64), so is CT (Pearson' correlation coeffient=0.37). The specificity of PET/CT in diagnosis of breast cancer and axillary lymph node metastasis was significantly higher than that of CT(P=0.04 < 0.05). The specificity of PET/CT in diagnosis of axillary lymph node metastasis was significantly higher than that of CT(x2=7.71,P < 0.05). The diagnostic accuracy of PET/CT is significantly higher than CT for the lesions < 2cm in size (P < 0.05).9 supclavicle lymph node metastasis(2 patients) one case of sternum and 1st lib metastasis werefound by both PET and CT and proven by pathology. 1 innermammary lymph node metastasis was found by PET/CT and CT .Conclusions: The clinical application value of PET/CT, which has both advantages of metabolic imaging and anatomic details delineation, for breast cancer in differentiating benign from malignancy and staging may be important. |