Diagnostic accuracy of 18F-FDG PET/CT in restaging of salivary gland cancer

1522 Objectives: The purpose of this study was to compare the diagnostic accuracy of CT or MRI and FDG PET/CT for detection of recurrent salivary gland cancer. Methods: PET/CT images obtained from November 2003 to October 2007 of salivary gland cancer patients who underwent surgery and/or radiation therapy were reviewed. Recurrent salivary gland cancers were classified based on site as follows: (1) local recurrence, (2) cervical lymph node (LN) metastasis, (3) distant metastasis. Also the peak standard uptake values (pSUV) of suspected lesions were recorded. PET/CT and CT or MRI were compared with the histological or clinical and follow up images (PET/CT, CT or MRI). Results: Total of 42 PET/CT images of 25 patients was included. Pathologic confirmation was available in 16 cases. 8 cases were concluded as local recurrence, and 7 cases as cervical LN metastasis. Sensitivity, specificity, and accuracy for local recurrence were 62.5%, 88.2%, and 83.3% in PET/CT, and 87.5%, 87.5%, 87.5% in the CT or MRI. For cervical LN metastasis, the sensitivity, specificity, and accuracy were 85.7%, 88.6%, and 88.1% in PET/CT, and 83.3%, 88.5%, 87.5% in the CT or MRI. In particular, 3 patients had distant metastasis demonstrated in PET/CT images, all in the lungs. The pSUVs between the malignancy positive group and negative group were not statistically different for the local recurrence (4.55 versus 4.19, p=0.835), nor for cervical LN metastasis (5.03 versus 2.77, p= 0.078). Conclusions: PET/CT is less sensitive than CT or MRI for detecting local recurrence of salivary gland cancer, and shows similar accuracy for cervical lymph node metastasis. However, PET/CT incidentally detected distant metastasis in 3 cases.