Background and aim: To evaluate the diagnostic value of integrated 18F-fluorodeoxyglucose (FDG) positron emission tomography and computed tomography (PET/CT) to discriminate malignant from benign ovarian tumors.
Methods: One hundred and eight women suspected of having ovarian cancer underwent preoperative FDG-PET/CT scans. FDG uptake was quantified by calculating the maximum standardized uptake value (SUV max) of each tumor. The receiver operating characteristic curve was drawn to determine the optimal cut-off values of SUV max that would best discriminate between benign and malignant tumors. Histopathologic results served as the reference standard. We assessed the association between SUV max and with International Federation of Gynecology and Obstetrics stage in borderline and malignant tumors, using one-factor analysis of variance and an unpaired t test with Bonferoni correction.
Results: The SUV max of benign (n=26), borderline (n=12) and malignant (n=73) lesions was 2.00 ± 1.02, 2.72 ± 1.04, and 7.55 ± 4.29, respectively. Although there were significant differences between benign and malignant, and borderline and malignant lesions (P<0.0001), there was no significant difference between benign and borderline lesions. Using an SUV max cutoff of 2.55, the sensitivity, specificity and accuracy of FDG-PET/CT scanning to detect malignant or borderline tumors were 82.4, 76.9, and 81.1%, respectively. The SUV max of stage I (n=35), stage II (n=8), stage III (n=34) and stage IV (n=8) was 3.59 ± 2.32, 5.18 ± 1.34, 8.72 ± 2.69, and 15.05 ± 3.77, respectively, and significant differences were observed between SUV max values and the various International Federation of Gynecology and Obstetrics stage (P<0.0001).
Conclusion: FDG-PET/CT scanning has a high diagnostic value in differentiating between malignant and benign tumors, and a low diagnostic value in differentiating between borderline and benign tumors.