Background: The value of tertiary cytoreductive surgery (TCS) on overall survival (OS) of patients with relapsed epithelial ovarian cancer (ROC) is not well defined. Aim of the present study was to evaluate the operative and clinical outcome after TCS.
Methods: We systematically evaluated all consecutive patients undergoing TCS. Tumor dissemination pattern, operative morbidity, residual tumor, and survival are described based on a validated intraoperative documentation tool. Predictors of survival and complete tumor resection are analyzed with Cox regression or logistic regression models.
Results: Between October 2000 and December 2008, 135 patients (median age, 51 years; range, 22-80 years) of mainly initial FIGO stage ≥ III (106 patients, 78.5%) were evaluated. In 53 patients (39.3%) a complete tumor-resection was obtained. The 1-month operative mortality was 6%. During a median follow-up period of 9.6 months (range, 0.1-75 months), 78 patients (57.8%) died, while 52 patients (38.5%) experienced a further relapse. Median OS was 19.1 months for the total collective (95% confidence interval [95% CI], 14.84-23.35). Median OS was 37.8 months (95% CI, 12.7-62.7) for patients without residual tumor; versus 19.0 months (95% CI, 9.8-28.2) for residual tumor ≤ 1 cm and 6.9 months (95% CI, 3.05-10.7) for residual tumor > 1 cm (P < .001). The presence of peritoneal carcinomatosis did not seem to significantly affect OS. Complete tumor resection was identified as the strongest predictor of OS. Other independent predictors of OS were interval to primary diagnosis ≥ 3 years (hazard ratio [HR], 0.28; 95% CI, 0.14-0.59) and serous papillary histology (HR, 0.23; 95% CI, 0.09-0.56). A total of 42 patients (31.1%) presented at least 1 major complication. Multivariate analysis identified tumor involvement of the middle abdomen and peritoneal carcinomatosis as independent predictors of complete tumor resection.
Conclusions: Postoperative tumor residual disease remains the strongest predictor of survival even in TCS setting. To identify the optimal candidates for TCS, the predictive value of ascites and peritoneal carcinomatosis should be confirmed by future prospective trials.