Abstract 3301: Breast cancer subtypes have distinct microbial and immune cytolytic transcriptomic signatures

Abstract

Abstract Background: The role of genetic mutations in breast cancer is well documented and has formed the basis for novel therapies. However, the impact of viral and bacterial infection in the pathogenesis of breast cancer is not well understood. Methods: We analyzed RNA sequencing (RNASeq) data generated by the Cancer Genome Atlas (TCGA) from over 900 pairs of breast tumor and adjacent normal tissues, using our pathogen discovery tool, PathSeq. We developed a novel metric for quantifying and normalizing microbial sequence abundance from RNASeq data by correcting for both the sequencing coverage and relative genome sizes of bacteria in the dataset. We also performed linear discriminant analyses (LDA) to test the hypothesis that the differences between the canonical breast cancer subtypes can be modeled on microbial abundance data. In addition, we interrogated the human gene expression data derived from the RNASeq dataset to determine the relationships between the breast cancer subtypes and immune cytolytic activity (CYT index; defined as the geometric mean of the expression levels of perforin and granzyme). Results: Viral sequences were rare but we identified bacterial sequences to varying degrees of abundance. Intriguingly, we demonstrated that microbial abundance data can be used to recapitulate the well-established PAM50 gene expression-based subsets. Comparison of tumors and adjacent normal tissues also revealed subtype-associated microbial profiles. In addition, we observed that the basal and Her2 subtypes of breast cancer had a higher frequency of tumors with high CYT index compared with luminal A and B subtypes. Conclusion: This study has revealed that the various breast cancer subtypes have unique microbial transcriptomic signatures as well as differential immune cytolytic activity profiles. These data suggest the potential for differential responses of breast cancer subsets to immunotherapy. Citation Format: Chandra Sekhar Pedamallu, Matthew Meyerson, Akinyemi Ojesina. Breast cancer subtypes have distinct microbial and immune cytolytic transcriptomic signatures. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3301.