Durvalumab and cediranib with and without olaparib in recurrent ovarian cancer: a phase II proof-of-concept study.

Abstract

Here we report the efficacy and translational findings of durvalumab, olaparib, and cediranib (D + O + C) and of durvalumab plus cediranib (D + C) from the recurrent ovarian cancer cohort within a single-center, multi-arm, non-randomized, multi-cohort phase I/II trial (NCT02484404). Sixty-eight patients were enrolled (39 in D + O + C, 29 in D + C). The primary endpoint was objective response rate (ORR); secondary endpoints included progression-free survival (PFS) and safety. ORR was 19.4% (95% CI, 9.5-43.5) for D + O + C and 29.6% (95% CI, 13.8-46.9) for D + C; D + C met the primary endpoint while D + O + C did not. Median PFS was 4.5 months in both arms, with four exceptional responders (PFS ≥ 12 months) per arm. Toxicity was manageable. Pre- and on-treatment biopsies and blood samples were collected for prespecified transcriptomic and immunophenotypic profiling; signature analyses and preclinical studies were conducted post hoc and were exploratory. Baseline tumors from exceptional responders and patients with clinical benefit (partial response or stable disease with PFS ≥ 4 months) demonstrated enrichment of immune activation and metabolic pathways, whereas tumors with no clinical benefit (NCB; progressive disease or stable disease with PFS < 4 months) exhibited upregulation of vascular adaptation and cytoskeletal remodeling pathways. These findings support the proof-of-concept clinical activity of D + O + C and D + C and identify molecular signatures with potential predictive value in subsets of recurrent ovarian cancer.