Investigating the molecular diversity of luminal breast cancer arising in young women

Breast cancer (BC) is the most common malignancy in women and a highly heterogenous disease. In young women (<40y), BC is uncommon (approximately 7% of all BC) but clinically significant. Even after adjusting for tumor stage, grade and subtype, young patients exhibit worse response to standard therapies, higher risk of relapse and lower overall survival compared to older patients particularly evident in the Luminal subtype (LBC). However, young LBC biology remains poorly understood and models that accurately reflect the molecular characteristics of this pathology are still lacking. To capture LBC heterogeneity and to investigate the molecular mechanisms driving cancer progression and drug resistance in young patients we collected a large cohort of samples from young pre-menopausal (<45y, Y) and non-Y post-menopausal (>55y, nY) LBC patients and we performed a comprehensive mutational profiling analysis using Next Generation Sequencing (NGS) approaches. Samples were sequenced first using a small custom-made amplicon-based panel, targeting four of the most frequently mutated genes in LBC (PIK3CA, AKT1, TP53, GATA3), then by whole exome sequencing (WES). Targeted NGS analyses revealed that the Y group exhibited a higher number of mutated (81% vs 62%, Y vs nY) and co-mutated (30% vs 18%, Y vs nY) patients, in these four considered genes. In detail, mutations in PIK3CA/AKT1 and GATA3 were significantly enriched in Y patients (57% and 34%, respectively, vs 36% and 14%, Y vs nY). Interestingly, we found that Y women harbor also higher frequency of PIK3CA/AKT1 co-mutations with GATA3 (16% vs 7%, Y vs nY) or TP53 (17% vs 12%, Y vs nY), especially in the luminal B subtype. WES analyses confirmed the enrichment of PIK3CA (40% vs 32%, Y vs nY) and GATA3 (19% vs 5%, Y vs nY) mutations in Y patients, although it was evident that many subclonal ones were lost. Additionally, WES data revealed other genes mostly mutated in Y (ERBB2, FAT3, CBFB and MAP3K1) and in nY (LRP1B, EPHA3 and CSMD3) patients. Copy number variation (CNV) analysis revealed that Y patients displayed TP53 deletions (55% vs 44%, Y vs nY) and GATA3 amplification (45% vs 25%, Y vs nY) at higher frequency compared to nY ones. Despite the well-known relevance of PIK3CA and TP53 pathway in oncogenesis, the role and crosstalk of PIK3CA/TP53 co-mutations (PIK3CA/TP53co-mut) is still unclear, especially in the context of LBC. Therefore, we decided to investigate the biological impact of PIK3CA/TP53co-mut on mammary tumorigenesis, using primary luminal breast epithelial (BPE) cells carrying PIK3CAE545K, PIK3CAH1047H, TP53R175H mutants, either alone or in combination. Co-occurring mutations of PIK3CA and TP53 altered the growth of these normal cells, leading to a transformed phenotype, much more than each of the single mutant alone. Moreover, PIK3CA/TP53co-mut cells displayed lower response to endocrine therapy, chemotherapy and targeted therapy, such as CDK4/6 and PI3K inhibitors, that are commonly used to treat LBC patients. It has been recently demonstrated that occurrence of PIK3CA and TP53 mutations may be responsible of epigenetic dependency of cancer cells, increasing their sensitivity to epigenetic therapy approaches. However, whether this is true also in LBC is unknown. In search for new therapeutic vulnerabilities, we thus performed an epigenetic drug screening evaluating the response of BPE co-mutated clones to a broad spectrum of epigenetic regulators inhibitors. We discovered that inhibitors of EZH2 methyl-transferase consistently reduced the viability of PIK3CA/TP53co-mut cells, suggesting that they might act in a synthetic lethal manner. Overall, our results provide new insights into the molecular fingerprint of LBC patients supporting the hypothesis that BC arise in Y women may represent a unique biological entity.