PARP inhibition with rucaparib alone followed by combination with atezolizumab: Phase Ib COUPLET clinical study in advanced gynaecological and triple-negative breast cancers

Abstract

Background: Combining PARP inhibitors (PARPis) with immune checkpoint inhibitors may improve clinical outcomes in selected cancers. We evaluated rucaparib and atezolizumab in advanced gynaecological or triple-negative breast cancer (TNBC).

Methods: After identifying the recommended dose, patients with PARPi-naive BRCA-mutated or homologous recombination-deficient/loss-of-heterozygosity-high platinum-sensitive ovarian cancer or TNBC received rucaparib plus atezolizumab. Tumour biopsies were collected pre-treatment, during single-agent rucaparib run-in, and after starting combination therapy.

Results: The most common adverse events with rucaparib 600 mg twice daily and atezolizumab 1200 mg on Day 1 every 3 weeks were gastrointestinal effects, fatigue, liver enzyme elevations, and anaemia. Responding patients typically had BRCA-mutated tumours and higher pre-treatment tumour levels of PD-L1 and CD8 + T cells. Markers of DNA damage repair decreased during rucaparib run-in and combination treatment in responders, but typically increased in non-responders. Apoptosis signature expression showed the reverse. CD8 + T-cell activity and STING pathway activation increased during rucaparib run-in, increasing further with atezolizumab.

Conclusions: In this small study, rucaparib plus atezolizumab demonstrated acceptable safety and activity in BRCA-mutated tumours. Increasing anti-tumour immunity and inflammation might be a key mechanism of action for clinical benefit from the combination, potentially guiding more targeted development of such regimens.

Clinical trial registration: ClinicalTrials.gov (NCT03101280).

Fig. 1. Baseline genetic and tumour immune microenvironment characteristics.

a Part 1 (all-comer ovarian and endometrial cancer dose-finding phase). b Part 2 (dose expansion). Patients in panels a and b are sorted from left to right by greatest reduction in SLD of target lesions. Tumour and biomarker characteristics are described according to the result from central review. However, patients could be enrolled or assigned to a cohort on the basis of local test results, to avoid delays in treatment initiation. In some cases, this resulted in a patient apparently being ineligible for the cohort in which they were treated. ^*BRCA1 mutation (p.D120H) with unknown functional consequence. ^†Patients who had NA/ND status for both BRCA mutation and LOH were enrolled based on local testing results. ^§Patient achieved CR despite best SLD change of –67% because the primary lesion (lymph node) decreased to <10 mm. CC clear-cell, CR complete response, EC endometrial cancer, HGS high-grade serous, HGS/C high-grade serous/carcinosarcoma, IC immune cell, LOH loss of heterozygosity, MSI, microsatellite instability, MSS microsatellite stable, NA not available, ND not done, OC ovarian cancer, PD progressive disease, PD-L1 programmed cell death-ligand 1, PFS progression-free survival, PP primary peritoneal cancer, PR partial response, RECIST Response Evaluation Criteria in Solid Tumours, SD stable disease, SLD sum of longest diameters, TMB tumour mutational burden, TNBC triple-negative breast cancer.

Fig. 2. Dynamic changes of tumour intrinsic and tumour immune microenvironment molecular features in longitudinal biopsies.

RNA sequencing transcriptome profiling was done in mandatory tumour biopsy samples collected before treatment, during rucaparib run-in (run-in), and at C2D1 of rucaparib plus atezolizumab (post-combination). a Schematic of biopsy collection timepoints in relation to treatment schedule. b Summary of enriched pathways (Ingenuity pathway analysis) from significantly differentially expressed genes (≥twofold change, P < 0.05) in paired post-combination vs. pre-treatment biopsies. Orange symbols: upregulated gene pathways; blue symbols: downregulated gene pathways. c Dynamic gene expression levels in the PARPi-regulated pathways for DDR, cell cycle, and apoptosis pathways. d Dynamic expression changes in immune signatures for CD274 (PD-L1 gene) and signatures representing CD8 T-cell activity and cGAS-STING. Each line plot in (c, d) represents a patient with confirmed CR or PR (left-hand panels) or SD or PD (right-hand panels). CnDn Cycle n Day n, cGAS cyclic GMP–AMP synthase, CR complete response, DDR DNA damage repair, OC ovarian cancer, PARPi poly(ADP-ribose) polymerase inhibitor, PD progressive disease, PD-L1 programmed cell death-ligand 1, PFS progression-free survival, PR partial response, SD stable disease, STING stimulator of interferon genes, TNBC triple-negative breast cancer.

Fig. 3. Dynamic changes in PD-L1 protein expression in ICs and CD8 T-cell infiltration in longitudinal biopsies.

a Left: representative PD-L1 IHC images in patient 10 (PR) at pre-treatment, run-in, and post-combination; Right: line plot graph of PD-L1 staining in ICs at different timepoints in responders (CR or PR) and non-responders (SD or PD). PanCK-positive tumour cells are purple and CD8-positive cells are brown. b Left: representative duplex IHC images for PanCK (tumour epithelial cells, magenta) and CD8 (T cells, brown) in patient 10 at different timepoints; Right: line plot graph summarising the percentage of CD8 infiltration over time by clinical outcomes and tumour indication. The black arrows point to PanCK (purple) stained tumour cells and the red arrows point to stained CD8 cells (brown). CR, complete response, IC immune cell, IHC immunohistochemistry, OC ovarian cancer, PD progressive disease, PD-L1 programmed cell death-ligand 1, PR partial response, SD stable disease, TNBC triple-negative breast cancer.