Background and objectives: Histopathologically confirmed lymph node metastasis is a prognostic factor in the surgical treatment of biliary tract cancer, however, preoperative diagnosis is still difficult even with computed tomography. FDG-PET has been used for the diagnosis of cancer and metastatic lesions. Herein, we retrospectively evaluated the utility of FDG-PET for detection of lymph node metastasis in biliary tract cancer.
Methods: We measured SUV(max) at each 190 surgically dissected lymph node area in 36 patients, and compared the values with histopathological diagnosis. The cutoff values for SUV(max) were defined from the ROC curve and the mean plus two standard deviations then used for detection of metastatic lymph node and prognostic value, compared with CT diagnosis.
Results: The sensitivity, specificity, and positive predictive value of FDG-PET were better than CT diagnosis (86%, 74%, 43% for SUV(max) ≥ 2.0, and 37%, 97%, 72% for SUV(max) ≥ 2.8, respectively). There was no relationship between SUV(max) and CT-determined lymph node dimensions. The presence of SUV(max) ≥ 2.8 lymph nodes was an independent determinant of prognosis after surgical treatment.
Conclusion: The detection of metastatic lymph nodes by FDG-PET is limited, but better than CT. SUV(max) for lymph nodes seems useful for clinical decision-making regarding treatment strategy including surgery.
Copyright © 2010 Wiley-Liss, Inc.