18F-FDG PET/CT usefulness in initial staging of ovarian cancer

440 Objectives: To evaluate the accuracy of 18F-FDG PET/CT in staging of ovarian cancer. Methods: Thirty patients with ovarian cancer underwent 18F-FDG PET/CT for staging (cTNM) before surgery. PET/CT results (cTNM) were compared with pathological staging (pTNM and FIGO criteria). Lymphadenectomy was performed in 8 pts (7 in stage I-II and 1 in stage IV); 145 pelvic and 159 retroperitoneal abdominal nodes were sampled. Results: pTNM revealed stage I in 6 pts, stage II in 4, stage IIIC in 18 pts and stage IV in 2. In 20/30 pts PET/CT and pTNM were concordant: 3 in stage II and 17 in stage IIIC. 7/10 discordant patients were upstaged at histology: 1 from cIA to pIC and 1 from cIIB to pIIC, 2 from cIIIC to pIV, 1 from cIA to pIB and 1 from cIB to pIIIC. In 3 patients (borderline cancer) PET/CT was negative (2 pIA, 1 pIB). For nodes, lesion based analysis provided an overall sensitivity, specificity, PPV, NPV and accuracy of PET/CT respectively of 87,5%, 100%, 100%, 99% and 99%; particularly this figures were 80%, 100%, 100%, 99%, 99% for abdominal nodes and all 100% for pelvic nodes. In a patients based analysis, the absence of metastatic lymph nodes was correctly diagnosed by PET/CT in 7/8 pts (88%). Conclusions: PET/CT staging was correct in 67% of patients. In early stage disease (I-II) PET accuracy was higher than 98% in nodal staging. PET/CT could be useful in selecting patients for abdominal lymphadenectomy, particularly in early stage (I-II).