Abstract Basal-Like Carcinomas (BLCs) are high grade ductal carcinomas identified by large scale transcriptomic analyses among carcinomas with triple-negative phenotype (ER- PR- HER2−). A particular relationship between this phenotype and the BRCA1 breast cancer susceptibility gene has been described based on several observations including high genomic similarities of sporadic and BRCA1 BLCs. BRCA1 mutated patients have been shown to strongly benefit from treatments using PARP inhibitors. However, significant benefit of this treatment was not demonstrated in sporadic triple-negative breast cancers in association to standard chemotherapy in phase III clinical trials, which emphasizes the necessity for a better selection of DNA repair deficient tumors. The goal of this study was to determine whether BRCA1 inactivation leads to particular genomic alterations that could be used to identify BRCA1 deficient sporadic BLCs. Genomic profiling was performed using SNP-arrays (Affymetrix and Illumina) and high quality profiles were obtained for 60 BLCs with more than 35% of tumor cells. The series contained 28 BRCA1 BLCs and 32 sporadic BLCs consisting in 10 sporadic BLCs with BRCA1 somatic inactivation by promoter methylation (BRCA1-like BLCs) and 22 sporadic tumors without BRCA1 methylation (nonBRCA1 BLCs). Genomic data was mined with GAP methodology (Popova et al, Genome Biol 2009), which allows absolute copy-number evaluation with DNA content calculation and clarification of Loss Of Heterozygosity (LOH) status. Genomic patterns of BLCs were characterized by frequent and highly recurrent allelic losses. Comparison between BRCA1, BRCA1-like and nonBRCA1 BLCs confirmed their overall similarity, and identified few significant differences enhanced when BRCA1 and BRCA1-like BLCs were considered as one group. Firstly, hypodiploidy characterized BLCs with BRCA1 germline and/or somatic inactivation. Secondly, the lpter region was found more often lost in BRCA1 and BRCA1-like BLCs than in nonBRCA1 BLCs. Thirdly, retention of inactive X chromosome was a characteristic of BLCs with BRCA1 germline and/or somatic inactivation. Finally, a BRCA1ness score based on the above described difference is proposed. In conclusion, despite a strong similarity of genomic patterns in BLCs, BRCA1 inactivation leads to few key genomic aberrations which constitute a new BRCA1 molecular signature namely hypodiploidy, 1pter loss and inactive X chromosome retention. This signature needs to be further evaluated in terms of response to PARP inhibitors in order to evaluate its capacity to identify patients that will benefit from this treatment. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-06-03.