Abstract
Breast cancer represents the number one cause of cancer-associated mortality globally. The most aggressive molecular subtype is triple negative breast cancer (TNBC), of which limited therapeutic options are available. It is well known that breast cancer prognosis and tumor sensitivity toward immunotherapy are dictated by the tumor microenvironment. Breast cancer gene expression profiles were extracted from the METABRIC dataset and two TNBC clusters displaying unique immune features were identified. Activated immune cells formed a large proportion of cells in the high infiltration cluster, which correlated to a good prognosis. Differentially expressed genes (DEGs) extracted between two heterogeneous subtypes were used to further explore the underlying immune mechanism and to identify prognostic biomarkers. Functional enrichment analysis revealed that the DEGs were predominately related to some processes involved in activation and regulation of innate immune signaling. Using network analysis, we identified two modules in which genes were selected for further prognostic investigation. Validation by independent datasets revealed that CXCL9 and CXCL13 were good prognostic biomarkers for TNBC. We also performed comparisons between the above two genes and immune markers (CYT, APM, TILs, and TIS), as well as cell checkpoint marker expressions, and found a statistically significant correlation between them in both METABRIC and TCGA datasets. The potential of CXCL9 and CXCL13 to predict chemotherapy sensitivity was also evaluated. We found that the CXCL9 and CXCL13 were good predictors for chemotherapy and their expressions were higher in chemotherapy-responsive patients in contrast to those who were not responsive. In brief, immune infiltrate characterization on TNBC revealed heterogeneous subtypes with unique immune features allowed for the identification of informative and reliable characteristics representative of the local immune tumor microenvironment and were potential candidates to guide the management of TNBC patients.
Highlights
More efficient prognostic and therapeutic strategies for triple negative breast cancer (TNBC) are still urgently needed
To assess the range and types of immune cell infiltration, the single sample gene set enrichment analysis (ssGSEA) method was used to estimate the enrichment of 27 types of immune cells for 299 TNBC patients derived from the METABRIC dataset who had existing transcriptome and clinical features data
By applying the unsupervised hierarchical agglomerative clustering, the TNBC tumors were subsequently re-classified into two heterogeneous clusters: low infiltration (176) and high infiltration (123) (Figure 1A)
Summary
11.6% of all cancers in women were found to be breast cancer, making it the most commonly diagnosed malignancy in this population as well as the leading cause of cancer death (Bray et al, 2018). In addition to TIL count by immune pathological evaluation, other methods recently emerged to assess the tumor immune landscape such as deconvolution algorithm to define the proportion of immune cells using genomic profiling (Loi et al, 2019), to identify the gene expression signatures that distinguish the immune-state, and to be a potential prognostic factor (Romero-Cordoba et al, 2019). PD-1 and PDL1 were highly expressed in TNBC in contrast to other subtypes, indicating that the patients benefit more from immune therapies (Zhou et al, 2018). PD-1 inhibitor sensitivity is not universal amongst all TNBC samples given the irregular expression of PD-1 and PD-L1 (Barrett et al, 2018). This challenges the current genomic-based breast cancer classification. We evaluated the ability of these signatures to predict patient chemotherapeutic response
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
