Real-world outcomes associated with bevacizumab combined with chemotherapy in platinum-resistant ovarian cancer.

5555 Background: In the pivotal Aurelia study, addition of bevacizumab (bev) to physician’s choice chemotherapy (paclitaxel, liposomal doxorubicin or topotecan) for platinum-resistant (PL-R) ovarian cancer (OC) was associated with improved progression-free survival (PFS; 6.7 vs 3.4 mos; hazard ratio (HR) 0.48, p = 0.001) but not overall survival (OS; 16.6 vs 13.3 mos HR 0.85, p = 0.174); the latter finding may relate to extensive crossover. In an exploratory subgroup analysis by treatment arm, benefits were particularly marked for bev + paclitaxel where median PFS (mPFS) increased from 3.9 to 10.4 mos (HR 0.46; 95%CI 0.30-0.71) and increases in OS approached statistical significance (HR 0.65; 95%CI, 0.42-1.02; 22.4 v 13.2 mos). Here we describe utilization of bev for PL-R OC and outcomes in routine clinical practice. Methods: The Ontario Cancer Registry and the New Drug Funding Program databases were utilized to identify all patients treated with bev plus chemotherapy (paclitaxel, liposomal doxorubicin or topotecan) for PL-R OC following its approval in 2017. Time on treatment (ToT) was defined as time from first to last bev treatment; this was used as a surrogate for PFS in routine practice. Median OS (mOS) was determined using the Kaplan-Meier method. Factors associated with ToT and OS were identified using a Cox proportional hazard model. A before and after comparison analysis was performed to determine mOS for patients treated pre- (2011-2017) and post-bev (2017-2019) approval. Results: From Oct 2017 to Dec 2019, 180 patients received bev + chemotherapy for first-line PL-R OC. Mean age was 63 years old, and 80% had serous OC. Bev was most often combined with liposomal doxorubicin (64 %) followed by paclitaxel (34%) and topotecan (2%). Median ToT was 3 months and OS was 11 months. ToT and OS were longer in patients who received paclitaxel as chemotherapy backbone (5 mos [ToT]; 14 mos [OS]) than those who received bev with liposomal doxorubicin (2 mos; 9 mos) or topotecan (2 mos; 6 mos). In multivariable models, ToT was superior in patients who received bev + paclitaxel than bev + liposomal doxorubicin (HR 0.40; 95%CI 0.28-0.57; p < 0.0001), and worse with longer time from diagnosis to bev start (1.03; 1.01-1.05; p = 0.0120). OS was also significantly longer in those who received paclitaxel vs liposomal doxorubicin (HR 0.54; 95%CI 0.30-0.98; p = 0.043). In a before and after analysis, patients treated in the pre- (n = 1290) and post-bev (n = 360) era had mOS of 8 and 9 months respectively. Post mOS increased for patients receiving paclitaxel (7 vs 12 months) but not with liposomal doxorubicin (9 vs 7 months). Conclusions: ToT and OS associated with bev for PL-R OC are shorter in a real-world population compared to results reported in AURELIA. ToT and OS were longer with bev + paclitaxel than with other chemotherapy agents.