Objective This study was conducted to estimate the long-term clinical outcome of patients with recurrent ovarian carcinoma (ROC).

Methods Six hundred three patients with ROC were analyzed in this study. The pathological slides were evaluated under central pathological review. The prognostic significances of clinicopathologic factors were evaluated using both univariate and multivariate analysis.

Results The 5-year overall survival (OS) and postrecurrence survival (PRS) rates were 31.1 and 16.9%, respectively. On stratifying to treatment periods, the PRS has been prolonged over the last decade (year ≥2000) compared with before this period (year ≤1999) (P = 0.0002). In contrast, on stratifying to histological types and treatment periods, in both OS and PRS, the prognosis of patients with the nonmucinous/clear-cell histology, including serous, endometrioid, and other histological types, was significantly improved after 2000 compared with before (year ≤1999) (OS, P = 0.0009; PRS, P < 0.0001). In contrast, that of patients with the mucinous/clear-cell histology did not significantly differ regardless of the treatment period (≥2000 vs ≤1999: OS, P = 0.3887; PRS, P = 0.7617). In multivariate analysis, the stage, period of starting initial treatment, histological type, and the treatment-free interval were independent prognostic factors of a poor OS and PRS (OS/PRS: histological type: mucinous/clear-cell vs nonmucinous/clear-cell: hazard ratio, 1.300/1.498; 95% confidence interval [CI], 1.039–1.626/1.197–1.874).

Conclusions Despite the continuous administration of treatment for ROC, survival is poor, and the extent of therapeutic progress differs according to the histological type.