Examining the Prevalence of Homologous Recombination Repair Defects in ER+ Breast Cancers.

Purpose/Objective(s) Cancers with defects in the homologous recombination repair (HRD) pathway exhibit increased sensitivity to platinum agents and PARP inhibitors (PARPi). HRD can occur secondary to germline or somatic acquired genetic alterations in HR genes or BRCA1 promoter methylation, but a plurality of HRD cancers do not have known explanatory mechanisms. HRD is known to be most frequent in triple negative breast cancers. We sought to determine the frequency of HRD, defined by genomic signatures, in ER+/Her2- breast cancers, the largest segment of the overall breast cancer population. Materials/Methods The Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium of the International Cancer Genome Consortium (ICGC) and The Cancer Genome Atlas (TCGA) has made the whole-genomes of 2,658 cases available for analysis. We used the available breast cancer cases in this dataset as our study substrate and further categorized them as ER+/Her2-, TNBC, or unlabeled. Large-scale state transitions (LST) are well-described ultra-structural “genomic scars” that can be assessed and scored using computational techniques, and serve as an inference for HRD. LST has been associated with responsiveness to PARPi, consistent with its use as a surrogate for HRD. We dichotomized LST scores > = 15 as HRD. We further examined the prevalence of the single base substitution signature 3 (SBS3) and indel signature 6 (ID6) previously associated with BRCA1/2 cases by the PCAWG project as additional surrogates for HRD. The Mann-Whitney test was used to assess differences across groups. Results There were 73 breast cancer genomes contributed to PCAWG by the EU breast cancer cohort, all of which were labeled ER+/Her2-. There were 44 breast cancer genomes contributed by UK (“UK cohort”) labeled TNBC. As expected, we observed statistically significant differences in mean LST scores and SBS3 between the ER+ and TNBC cohorts. Within the ER+/Her2- cohort, 11 of 73 cases (15%) had HRD as defined by high LST. The HRD status of these 11 cases is supported by elevated ID6 and SBS3 compared to the 62 cases with low LST. Conclusion Our work suggests that HRD may be as prevalent as 15% in ER+Her2- breast cancer. Given the preponderance of ER+ breast cancer, women with ER+ breast cancer with HRD may outnumber the 40% of women with TNBC with HRD. In a future with rapid, clinical grade assays for HRD, many more breast cancer patients may be eligible for synthetically lethal combination therapies than previously anticipated.