Date of Award

January 2016

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

Department

Medicine

First Advisor

Lajos Pusztai

Second Advisor

Christos Hatzis

Abstract

The causes of high versus low, or absent, immune cell infiltration in breast cancer is unknown. The goals of this analysis were to examine if total mutation load, neoantigen load, copy number variations (CNV), gene-level or pathway level somatic mutations or germline polymorphisms (SNP) are associated with the level of immune infiltration measured by immune metagene expression levels. We used RNA-Seq, DNA copy number, mutation and germline SNP data from the TCGA representing n=627 ER+, n=207 HER2+ and n=191 TNBC cancers. 13 published immune metagenes were used in correlation and multivariate linear regression analyses performed separately for the 3 major clinical subtypes. P-values were adjusted for multiple comparisons and permutation testing was used to assess false discovery rates. Overall mutation, neoantigen and amplification, or deletion loads did not correlate strongly with any of the immune metagenes in any subtype (Spearman coefficient 0.2). In ER+ cancers, mutations in MAP2K4 and TP53 were associated with lower and higher levels of immune infiltration, respectively. In TNBC, mutations in MYH9 and HERC2 were associated with lower immune infiltration. None were found in HER2+ cancers. Three SNPs (rs425757, rs410232, rs470797) in the exonic regions of the FHPR1 and MLP genes were associated with low immune infiltration in ER+ cancers, none in the other subtypes. Two amplicons in TNBC and 3 amplicons in HER2+ cancers were associated with lower immune infiltration. We also identified alterations in several biological pathways that were associated with immune infiltration in different breast cancer subtypes. At the individual patient level, each pathway were affected at different genes through distinct genomic mechanisms. Our results suggest that immune infiltration in breast cancer is not driven by a single global metric of genomic aberrations such as mutation, neoantigen or CNV loads, but by multiple different gene and pathway level associations that each affect small subsets of patients within each subtype.

Comments

This is an Open Access Thesis.

Share

COinS