ABSTRACT
Guidelines for the treatment of tubo-ovarian cancer patients beyond third line are lacking. We aimed to evaluate the effect of response in each line on patient's outcome as well as identify variables that predict response for additional line of chemotherapy. A cohort study was performed including all patients with advanced high-grade ovarian cancer. Survival analysis was performed using Kaplan-Meier curves and log-rank tests. Odds ratios and hazard ratios were calculated using multilevel, mixed-effects logistic regression and Cox regression, adjusting for repeated measures within individual patients. Two-hundred thirty-eight patients were included and underwent up to 10 lines of chemotherapy. The median progression-free survival was 15.6 and overall survival (OS) was 55.6 months. Response rates dropped with each additional line and by line 5, most patients (61%) became refractory and only 16% had any type of response (complete 4% or partial 12%). By line 2, whether a patient had partial disease (PR), stable disease (SD) or progressive disease (PD) did not have an effect on the OS. From line 2, whether a patient had PR, SD or PD did not have an effect on chemotherapy-free interval. Number of previous lines and time from previous line were the only variables that significantly correlated with both outcome of patients and response to the next line. In conclusion, time interval from the previous line of chemotherapy is the major clinical factor that predicts beneficial effect of another line of treatment in patients with ovarian cancer.