Breast cancer diagnosed during pregnancy is associated with enrichment of non-silent mutations, mismatch repair deficiency signature and mucin mutations

Abstract

Breast cancer diagnosed during pregnancy (BCP) is a rare and highly challenging disease. To investigate the impact of pregnancy on the biology of breast cancer, we conducted a comparative analysis of a cohort of BCP patients and non-pregnant control patients by integrating gene expression, copy number alterations and whole genome sequencing data. We showed that BCP exhibit unique molecular characteristics including an enrichment of non-silent mutations, a higher frequency of mutations in mucin gene family and an enrichment of mismatch repair deficiency mutational signature. This provides important insights into the biology of BCP and suggests that these features may be implicated in promoting tumor progression during pregnancy. In addition, it provides an unprecedented resource for further understanding the biology of breast cancer in young women and how pregnancy could modulate tumor biology.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1

Summary of the genome-wide copy number analysis of 87 controls and 38 BCP tumor samples. a–c Comparison of cancer cell fraction, ploidy and fraction of genome altered between controls and BCP. d Comparison of the CNA frequencies of controls (blue) and BCP (pink). e Heatmap of 35 CNA breast cancer driver genes according to their alterations; controls (blue) and BCP (pink). P, p-value derived for the Mann–Whitney U test

Fig. 2

Mutational landscape of individual BCP and controls. a Bar chart representing the absolute number of substitutions in BCP and controls, y-axis limited to 50,000 indicated by (*). b Bar chart representing the absolute number of non-silent mutations in BCP and controls (median: 20 vs. 12, P = 0.027, respectively). c Co-mutation plot showing genes harboring at least one non-silent mutation with a frequency of at least 5% across the whole cohort, and their corresponding frequencies in BCP and controls (right). d Proportion of breast cancer substitution signatures in each sample. Signatures are colored according to broad biological groups: 1 and 5 are associated with clock-like processes, 2 and 13 are APOBEC-related, 20 and 26 are associated with mismatch-repair deficiency, 3 and 8 are associated with homologous-recombination deficiency

Fig. 3

Enrichment of mucin mutations and upregulation in BCP. a Lollipop plots were generated using cBioPortal Mutation Mapper. Each lollipop denotes a unique missense mutation for MUC2, MUC17, and MUC20 in BCP. b MUCsig according to normal adult mouse mammary development (from pregnancy day 1 to involution day 2). c Comparison of MUCsig between controls and BCP. P, p-value derived for the Mann–Whitney U test

Fig. 4

Association of signature 20 with mutational load and clinical outcome. a Comparison of SNV mutational load between Sig20 negative and Sig20 positive tumors. b Comparison of MSH2 expression between Sig20 negative and Sig20 positive tumors. c Kaplan–Meier plot showing the difference in DFS between control patients (N = 18), BCP patients with Sig20 negative tumors (N = 22) and BCP patients with Sig20 positive tumors (N = 13). P, p-value derived for the Mann–Whitney U test