Interval breast cancer is associated with interferon immune response

Emilio Ugalde-Morales,Felix Grassmann,Keith Humphreys,Jingmei Li,Mikael Eriksson,Nicholas P Tobin,Linda S Lindström,Johan Vallon-Christersson,Åke Borg,Per Hall,Kamila Czene

doi:10.1016/j.ejca.2021.12.003

Emilio Ugalde-Morales, Felix Grassmann + Show 9 more

Open Access

https://doi.org/10.1016/j.ejca.2021.12.003

Copy DOI

Abstract

BackgroundThe aggressive nature of breast cancers detected between planned mammographic screens, so-called interval cancers, remains elusive. Here, we aim to characterise underlying molecular features of interval cancer. MethodsFrom 672 patients with invasive breast cancer, we analysed gene expression differences between 90 ‘true’ interval cancer cases (i.e. women with low-dense breasts defined as per cent mammographic density <25%) and 310 screen-detected tumours while accounting for PAM50 subtypes and thus overall tumour aggressiveness. We computed an interval cancer gene expression profile (IC-Gx) in a total of 2270 breast tumours (regardless of interval cancer status) and tested for association with expression-based immune subtypes in breast cancer. In addition, we investigated the contribution of inherited and somatic genetic variants in distinct features of interval cancer. ResultsWe identified 8331 genes nominally associated with interval cancer (P-value < 0.05, fold-change > 1.5). Gene set enrichment analysis showed immune-related pathways as key processes altered in interval cancer. Our IC-Gx, based on 47 genes with the strongest associations (false discovery rate < 0.05), was found to be associated mainly with immune subtypes involving interferon response. We isolated an interaction network of interval cancer and interferon genes for which a significant load of somatic and germline variants in class I interferon genes was observed. ConclusionWe identified novel molecular features of interval breast cancer highlighting interferon pathways as a potential target for prevention or treatment.

Highlights

Breast cancers diagnosed after a negative mammographic screening examination and before the scheduled mammogram are known to exhibit more aggressive phenotypes than their screendetected counterparts [1e3]
Through Gene set enrichment analysis (GSEA) analyses, we identified biological processes likely to be altered in interval cancers after controlling for molecular differences explained by the PAM50 subtypes (Table 2)
From 672 RNA-sequenced invasive breast cancer tumours obtained from the Linne-Brost 1 (LIBRO-1)/KARMA cohort, we identified 90 ‘true’ interval cancers and 310 screen-detected tumours (Table 1)

Summary

Introduction

Breast cancers diagnosed after a negative mammographic screening examination and before the scheduled mammogram (interval cancers) are known to exhibit more aggressive phenotypes than their screendetected counterparts [1e3]. Interval cancers are more likely to have unfavourable tumour characteristics [4,5] and are associated with poorer prognosis independently of clinicopathologic features [6,7]. This subset of aggressive tumours is enriched among women with mammographically low-dense breasts [8]. Methods: From 672 patients with invasive breast cancer, we analysed gene expression differences between 90 ‘true’ interval cancer cases (i.e. women with low-dense breasts defined as per cent mammographic density

Objectives

Methods

Results

Discussion

Conclusion