VIOLETTE: Randomised phase II study of olaparib (ola) plus ceralasertib (cer) or adavosertib (ada) vs ola alone in patients (pts) with metastatic triple-negative breast cancer (mTNBC)

Abstract

Background Combining DNA damage response (DDR) inhibitors to inhibit multiple DDR pathways may enhance tumour cell death. Preclinical models show synergistic antitumour effects of ola (PARP1 inhibitor) + cer (ATR inhibitor) or ada (WEE1 inhibitor). VIOLETTE (NCT03330847), a phase 2 open-label study, evaluated ola ± cer or ada as 2nd–3rd line in pts with mTNBC. Methods Randomisation was stratified within each arm by mutation status (BRCAm, non-BRCAm homologous recombination repair pathway mutations [HRRm], or no HRRm [non-HRRm]) based on a prospective central tumour test and prior platinum therapy. Pts received ola 300 mg BID, 28-d cycle; cer 160 mg d 1–7 + ola 300 mg BID (cer+ola), 28-d cycle; or ada 150 mg BID d 1–3, 8–10 + ola 200 mg BID (ada+ola), 21-d cycle. Primary study endpoint was progression-free survival (PFS). Secondary endpoints included objective response rate (ORR), safety, and tolerability. Ada+ola was discontinued early due to higher than expected grade ≥3 haematologic toxicity, which limited its interpretation. We focus on cer+ola vs ola. The sponsor stopped VIOLETTE after reviewing BRCAm stratum efficacy data. Results Of 273 pts (450 planned; median age: 53 y), 114 received ola, 112 cer+ola, and 47 ada+ola. Primary analyses showed no statistically significant difference in PFS for cer+ola vs ola (BRCAm: n=83; HR 1.02 [90% CI 0.63–1.66, P=0.94], HRRm: n=40; 0.54 [0.28–1.03, 0.13], non-HRRm: n=103; 0.76 [0.50–1.14, 0.30]). No statistically significant difference in ORR was seen for cer+ola vs ola in BRCAm (50% vs 44%) or HRRm (20% vs 15%). ORR was higher in non-HRRm for cer+ola (15%, n=8) vs ola (4%, n=2) (odds ratio 4.45; 90% CI 1.30–21.20, P=0.04). In all pts, nausea and anaemia were the most common adverse events (AEs). Grade ≥3 AEs: 36% ola, 47% cer+ola, 78% ada+ola. Details of exploratory and subgroup analyses will be presented. Conclusions Cer+ola did not improve PFS vs ola as 2nd–3rd line for mTNBC. Clinical significance of higher ORR with cer+ola in non-HRRm pts is unclear. Cer+ola had a manageable safety profile consistent with known profiles of each. Further analyses may identify pts likely to benefit from each treatment.