Background: The purposes of this study are to evaluate the prognostic value of posttreatment18F-FDG PET/CT in predicting the survival of patients with ovarian carcinoma and to determine incremental value of combining posttreatment PET/CT with traditional prognostic factors in a multivariate model. Methods: This was an IRB-approved retrospective study. From July 2001 to July 2011, 48 patients who completed initial therapy for ovarian carcinoma with concurrent 3-to 9-month initial posttreatment18F-FDG PET/CT and serum CA-125 were identified from the radiology database. Prognostic value of posttreatment PET/CT, CA-125, age, race, and tumor stage were determined from Cox proportional hazard model using univariate and multivariate analyses. Time-dependent receiver operator curves were also calculated at various follow-up intervals. Results: In a univariate model, overall survival (OS) was associated with PET/CT (hazard ratio = 4.18; 95% CI 1.49–11.70) and CA-125 (hazard ratio = 11.09; 95% CI 4.27–28.79). When the effects of posttreatment PET/CT and CA-125 were combined in the multivariate analysis, hazard ratio for PET/CT increased to 4.84 (95% CI 1.59–14.73, p = 0.005) and hazard ratio for CA-125 increased to 14.43 (95% CI 4.65–44.84, p < 0.001). In the subset of patients with negative CA-125, posttreatment PET/CT had a hazard ratio of 2.98 (95% CI 0.86–10.37), supporting the role of posttreatment PET/CT in risk stratification of patients with negative CA-125. Time-dependent receiver operator curves showed that the combination of PET/CT and CA-125 improved prognostic accuracy compared to PET/CT or CA-125 alone at 12-, 24-, 30-, and 36-month follow-up. Conclusions: Posttreatment PET/CT can predict the survival of patients with ovarian carcinoma. The addition of posttreatment PET/CT to the CA-125 serum biomarker has an incremental value in improving prognostic accuracy, particularly in the subset of patients with negative CA-125.
- Ovarian carcinoma
- Patient survival
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging