Objective The optimal role of bevacizumab (Bev) in the treatment of ovarian cancer has not yet been established. Furthermore, it is unclear whether there is a benefit of Bev after progression on a Bev-containing regimen in ovarian cancer. The objective of this study was to compare response rates, progression-free survival (PFS), and overall survival between patients who were treated with chemotherapy and Bev after progression on Bev (BAB) versus patients who were treated with chemotherapy without Bev (CWOB).

Methods We conducted a retrospective chart review of all patients who received treatment with Bev (with or without cytotoxic chemotherapy) for recurrent ovarian cancer at a single institution. Patients who received additional therapy after progression while on Bev were included.

Results Forty-six patients were included (16 CWOB group and 30 BAB). The median number of previous chemotherapy regimens was 2.5 for CWOB compared with 4 for BAB (P = 0.11). Fifty-two percent of patients had an objective response to the first Bev regimen before progressing on Bev. Response rates for the regimen after progression on Bev were 19% (3/16) in the CWOB group and 23% (7/30) in the BAB group (P = 1). Twenty-five percent of the patients who responded to the first Bev regimen and 18% of those who did not respond to the first Bev regimen responded to the second Bev regimen (P = 0.72). The median PFS for patients in the CWOB group was 2.6 months (95% confidence interval [CI], 1.3–5 months), compared with 5.0 months (95% CI, 3.5–7.3 months) for patients in the BAB group (P = 0.01). Overall survival was similar, 9.4 months (95% CI, 5.0–12.0 months) for CWOB versus 8.6 months (95% CI, 5.8–15.5 months) for BAB (P = 0.19). One patient in the BAB group died of a bowel perforation.

Conclusions In patients previously treated with Bev for recurrent ovarian cancer, the subsequent addition of Bev to cytotoxic chemotherapy increased the PFS compared with patients not receiving a second course of Bev, but did so without an impact on overall survival. The response to the first Bev regimen did not predict whether a patient would respond again to the next Bev regimen. Randomized, larger studies will have to be performed to confirm this observation.