ABSTRACT: This study aimed to evaluate the efficiency of 18F-FDG PET/CT in suspected recurrence of epithelial ovarian cancer, after treatment, comparing outcomes of PET/CT with histological tumor subtype, CA-125 serum levels, and findings of conventional diagnostic imaging modalities (CI).
Data from 121 women who underwent FDG PET/CT for suspected recurrence of epithelial ovarian cancer after treatment were reviewed retrospectively.
Of all patients, 80% had recurrent disease and 20% were disease-free on the final clinical diagnosis. PET/CT showed true-positive findings in 82% of patients, whereas CI demonstrated true-positives in 70% of cases. At the time of PET/CT scanning, only 55 patients had serum CA-125 level greater than 35 U/mL, whereas 52 patients presented with CA-125 levels in a reference range. PET/CT sensitivity (82%) was significantly higher than that of CA-125 (59%), whereas difference in sensitivity between PET/CT and CI (69%) was limited. PET/CT specificity (87%) was significantly better than that of CI (47%), although no difference in specificity between PET/CT and CA-125 (80%) was found. However, no difference in CA-125 serum levels between patients with local tumor relapse and those with distant metastases was found. PET/CT showed the highest positive predictive value (96%) and negative predictive value (55%) when compared with other modalities. In high-grade tumors (n = 66), PET/CT accuracy was 80%, better than that of serum CA-125 (64%) and that of CI (62%). Equally in low-grade ovarian carcinomas (n = 55), PET/CT accuracy (87%) was significantly higher than that of the tumor marker (60%) and also higher than that of CI (70%).
FDG PET/CT was proven to be more efficient than serum CA-125 assay and CI in detecting recurrences of ovarian cancer after treatment. The sensitivity of FDG PET/CT is not influenced by tumor histology. FDG PET/CT should be considered a useful diagnostic tool in the surveillance of patients that received treatment for epithelial ovarian carcinoma.
Clinical nuclear medicine 08/2012; 37(8):e184-8. · 3.92 Impact Factor