Olaparib (O) + ceralasertib (C) in patients (pts) with metastatic triple-negative breast cancer (mTNBC): Translational analysis of the VIOLETTE trial.

1087 Background: O is a PARP inhibitor approved in a number of tumor types and for use in pts with germline (g) mutations in BRCA ( BRCAm) HER2-negative metastatic breast cancer. C is an inhibitor of DDR kinase ATR. Analysis of VIOLETTE (NCT03330847) in pts with mTNBC showed no improvement in PFS with a combination of C and O vs. O, whether in the context of pathogenic BRCAm, non-BRCA HRR pathway mutations (HRRm), or homologous recombination repair (HRR) wildtype (HRRwt) tumors. We report an exploratory translational analysis of predictive genomic biomarkers of response to O ± C beyond g BRCAm. Methods: Pts received O (300 mg twice daily) ± C (160 mg daily; days 1–7; 28-day cycles) as a second- or third-line treatment option. Pts were prospectively stratified into molecular strata based on pathogenic/likely pathogenic alterations in BRCA1/2 ( BRCAm; n=83) or in ≥1 of 13 selected non- BRCA HRR genes (HRRm; n=40) or HRRwt (n=103) using FoundationOneCDx (F1CDx) sequencing assay of archival tumor sample. Advanced genomic analysis included genome-wide loss of heterozygosity (a marker of HRR deficiency [HRD] with a 16% cut-off to define HRD+ [≥16%] vs HRD− [<16%]), zygosity, and predicted g/somatic (s) status of BRCA and HRR alterations based on a validated computational FMI algorithm. Efficacy was evaluated by blinded independent central review of PFS, RECIST response, and best percentage change from baseline in target lesion size. Results: Out of evaluable pts for origin of BRCAm (n=42), 81% (n=34) were of germline, and 19% (n=8) were of somatic origin. Of pts evaluable for zygosity (n=48), most BRCA alterations were biallelic (90%; n=43), and of pts evaluable for HRD status (n=47), most were HRD+ (94%; n=44). Of HRRwt pts evaluable for HRD status (n=79), 57% were HRD+ (n=45). Responses by biomarker are shown. Responses in BRCAm pts were seen in all subgroups (zygosity, HRD, or g/s origin of mutation), including s BRCAm (ORR: n=3/4 on O; n=2/4 on O+C). Responses were also seen in patients with heterozygous BRCA alteration (n=3/5) by genomic result and in BRCAm/HRD− pts (n=3/3) across both arms. Pts with select HRR genes ( BARD1, RAD51C/D, ATM, CDK12) also achieved responses on O or O+C. Conclusions: Responses to O were seen in pts with both g BRCAm and s BRCAm. Beyond BRCAm, the efficacy of O was observed in pts with alterations in select HRR genes, such as BARD1 or RAD51C/D; further clinical investigation is required. Small sample size and HRD analysis performed on archival tumor samples should be considered limitations. Clinical trial information: NCT03330847 . [Table: see text]