Background: Bevacizumab (bev) may be part of first-line treatment for patients with “high-risk” epithelial ovarian cancer (OC) (GOG-0218 and ICON7 trials). The prevalence of high-risk patient subgroups, including bev use and its effect on survival in a real-world setting is not well studied. Methods: The observational multi-country cohort study RESPONSE collected retrospectively medical record data on women with newly diagnosed advanced high-grade serous or endometrioid OC (DOI: 10.1002/cncr.34350). We analysed patients defined as ‘high-risk’ according to ICON7 definition as either 1) stage IV disease, 2) inoperable stage III disease or 3) sub-optimally debulked (>1 cm residual disease) stage III disease. Unadjusted Cox proportional hazards regression analysis was performed to estimate the association of bev with the risk of death. Follow up time was calculated from time between response assessment at the end of primary chemotherapy and death or 20 months follow-up, whichever occurred first. Results: Of 954 patients in total, we identified 386 high-risk patients (40%) treated with platinum-based chemotherapy. Of these, 132 (34%) received bev maintenance treatment. Baseline characteristics did not differ between patients receiving bev or no bev. Bev use was associated with a statistically significant reduction in risk of death with a HR of 0.49 (95%Ci: 0.18-0.78, p=0.002) in the total group of high-risk patients. The effect on survival was most evident among inoperable stage III/IV patients (n=151) with a 63% reduction in risk of death (HR 0.37; 95% CI (0.18-0.78, p=0.009). Risk estimates for additional subgroups are given in the table. Conclusions: Forty percent of this real-world population are considered high-risk according to ICON-7 and a majority of these patients did not receive bev during first-line treatment. In line with ICON 7, this analysis supports the beneficial effect of bev on overall survival in this high-risk population. Patients who do not receive surgery are underrepresented in clinical trials and this real-world study confirms the strong benefit of bev particularly in this subgroup. Any real-world study may be limited by the potential presence of selection bias of patients to receive bevacizumab or not.

*High-risk patients included stage IV disease, inoperable stage III disease or sub-optimally debulked (>1 cm) stage III disease