Bevacizumab, olaparib, and durvalumab in patients with relapsed ovarian cancer: a phase II clinical trial from the GINECO group

Abstract

Most patients with advanced ovarian cancer (AOC) ultimately relapse after platinum-based chemotherapy. Combining bevacizumab, olaparib, and durvalumab likely drives synergistic activity. This open-label phase 2 study (NCT04015739) aimed to assess activity and safety of this triple combination in female patients with relapsed high-grade AOC following prior platinum-based therapy. Patients were treated with olaparib (300 mg orally, twice daily), the bevacizumab biosimilar FKB238 (15 mg/kg intravenously, once-every-3-weeks), and durvalumab (1.12 g intravenously, once-every-3-weeks) in nine French centers. The primary endpoint was the non-progression rate at 3 months for platinum-resistant relapse or 6 months for platinum-sensitive relapse per RECIST 1.1 and irRECIST. Secondary endpoints were CA-125 decline with CA-125 ELIMination rate constant K (KELIM-B) per CA-125 longitudinal kinetics over 100 days, progression free survival and overall survival, tumor response, and safety. Non-progression rates were 69.8% (90%CI 55.9%-80.0%) at 3 months for platinum-resistant relapse patients (N = 41), meeting the prespecified endpoint, and 43.8% (90%CI 29.0%-57.4%) at 6 months for platinum-sensitive relapse (N = 33), not meeting the prespecified endpoint. Median progression-free survival was 4.1 months (95%CI 3.5-5.9) and 4.9 months (95%CI 2.9-7.0) respectively. Favorable KELIM-B was associated with better survival. No toxic deaths or major safety signals were observed. Here we show that further investigation of this triple combination may be considered in AOC patients with platinum-resistant relapse.

Fig. 1. Flow-chart of the BOLD study.

Description of the patients screened for the study.

Fig. 2. Kaplan–Meier estimates of PFS in the platinum-resistant relapse (PRR) and platinum-sensitive relapse (PSR) cohorts.

Number of patients at risk is shown. Source data are provided as a Source Data file.

Fig. 3. Time on treatment, duration of response, and best overall response per RECIST 1.1, by patient.

Each horizontal bar represents a treated patient and arrows indicate treatment was ongoing at the data cutoff date. P indicates patients who received prior PARP inhibitor therapy. Three patients were not evaluable for response per RECIST 1.1 as progression was clinically symptomatic. Note: for two patients who experienced disease progression, treatment was maintained in the context of clinical benefit per protocol, given that further progression was not observed in subsequent evaluations and clinical status remained stable. Source data are provided as a Source Data file.

Fig. 4. Kaplan–Meier estimates for predictive biomarkers.

PFS (a) and OS (b) according to favorable versus unfavorable standardized (std) KELIM-B (≥1 or <1, respectively). PFS (c) and OS (d) according to favorable and unfavorable TIS (≥median and <median, respectively) in the platinum-resistant relapse (PRR) versus platinum-sensitive relapse (PSR) cohorts. PFS (e) and OS (f) according to combined favorable and unfavorable standardized KELIM-B and TIS in evaluable patients. Number of patients at risk is shown. Log-rank two-sided test: a Chisq = 4.979 on 1 ddl, b Chisq = 8.833 on 1 ddl, c Chisq = 5.243 on 1 ddl, d Chisq = 3.221 on 1 ddl, e Chisq = 4.215 on 1 ddl, f Chisq = 8.797 on 1 ddl. Source data are provided as a Source Data file.