Insights into molecular classification of Triple Negative Breast Cancer: correlation with classical prognostic factors and impact on outcome.

Triple Negative Breast Cancer (TNBC) represents about 10-20% of invasive breast cancers: it is characterized by high level of molecular heterogeneity and lack of expression of HER2, Progesterone Receptor (PgR) and Estrogen Receptor (ER) that are both prognostic and predictive factors. Moreover, it is usually high grade and evolves in long distance metastasis. Due to lack of therapeutic strategies, TNBC is still an important matter in women's health. Nowadays, learning about its molecular and biological heterogeneity is of primary importance in order to study new markers to better classify TNBC and new molecular targets to develop targeted therapeutic strategies. The aims of our work were: (1) to classify at molecular level a wide number of triple negative breast cancers into different TNBC molecular subtypes; (2) to study the correlation between these molecular subtypes and patient’s outcome (3) to evaluate the type of relationship between TNBC molecular subtypes with presence of Tils and with the expression of specific immunohistochemical markers (such as AR, FOXA1, CK 5/6 and FGFR4). A total of 334 subjects were included in this study and for all patients we already had information about clinic history and tumor’s characteristics. Among all 334 patients we selected 44 subjects on the basis of the amount of material we had in our possession, and we did the Breast Cancer 360TM Biological Signatures molecular test. For what concerns the molecular classification, our results showed that: 45% of TNBC are composed by BLIA (basal like immuno-activated), while 26%, 19% and 10% are respectively classified as BLIS (basal like immuno-suppressed), LAR (androgen receptor group) and MES (mesenchymal type). In line with literature data, we observed that molecular subgroups are related with outcome, in particular BLIA showed a better prognosis with a smaller number of recurrences, otherwise BLIS presented the most aggressive outcome. Searching for a correlation with morphological and immunohistochemical data, we found that immunohistochemical expression of AR/FOXA1 are higher in LAR subgroup, while CK5/6 were more evident in MES. Moreover, the percentage of Tils (Tumor Infiltrative Lymphocytes) resulted major that 50% only in BLIA subgroup. Interestingly, we found out a curious positive correlation between AR expression and upregulation of Fibroblast Growth Factor Receptor 4 (FGFR4) gene. In line with this, we decided to investigate FGFR4 protein expression on the whole series. We found out that the immunohistochemical expression of FGFR4 correlated with lymph node metastasis, influencing both disease free and overall survival. In conclusion, our results confirmed the heterogeneity of TNBC at molecular level, also underlying their different impact on outcome. In addition, the present study paved the way to the investigation of FGFR4 receptor as a new prognostic factor in TNBC, also suggesting a new potential therapeutic target. Further studies using both in vitro and in vivo model should be useful to clarify the prognostic and predictive role of FGFR4.