Gestational breast cancer: distinctive molecular and clinico-epidemiological features

Abstract

Gestational breast cancer (GBC), defined as breast cancer (BC) diagnosed during pregnancy or the first-year post-partum, accounts for 6-15% of BC cases in women aged 20-44 years. GBC has worse prognosis than non-GBC, but reasons behind are not clear. The GEICAM/2012-03 Study (Molecular Characterization of Gestational Breast Cancer) is a multicenter prospective/retrospective observational registry of patients diagnosed with GBC. From November 2014 to June 2015 seventy patients diagnosed with GBC were included in the study, 30 diagnosed during pregnancy and 40 after delivery. Our current study was aimed to explore differences in epidemiological, clinico-pathological and gene expression features of GBC tumors, from the GEICAM/2012-03 Study, compared to non-GBC tumors from patients of similar age (< 43 years) from six different GEICAM studies, used as non- GBC control population. As per the main objective, the study found multiple differences showing GBC tumors as a different biological entity. GBC showed a more aggressive biology, with higher Ki67 levels, higher incidence of breast and/or ovarian cancer family history, and germline deleterious BRCA1/2 mutations, and are enriched in basal-like intrinsic subtype. GBC patients showed a lower number of tumor infiltrating lymphocytes, while specific genetic signatures highlight differences in GBC´s distinctive transcriptome. Our study shows that GBC is potentially a clinically and molecularly different entity, with specific epidemiological, clinical, and histological features, as well as a distinctive altered immune state and genetic signature. Nevertheless, further studies are needed to better understand the biology of GBC and to identify new targets against which develop new, more effective, targeted therapies.

Keywords: Breast cancer; Gene expression; Gestation; Gestational breast cancer; PAM50 intrinsic subtypes; Tumor infiltrating lymphocytes (TILs).

Fig. 1

Consort diagram. Flowchart describing the GBC and non-GBC populations used in the study, the assessments made and the number of patients analyzed in every step. FFPE: Formalin Fixed Paraffin Embedded; n: sample size; Pts: patients; TILs: tumor infiltrating lymphocytes

Fig. 2

Comparison of GBC against non-GBC control populations. A Presence of deleterious mutations in BRCA1 and/or BRCA2 in GBC vs. non-GBC control population from Van der Broek et al. (2016) [20]; *exact binomial test for proportions, p-value = 0.003; list of BRCA1 and BRCA2 deleterious mutations and variants of unknown significance found in GBC patients. B Distribution of PAM50 intrinsic subtypes in GBC vs. non-GBC control population from GEICAM/9906 [13] and Malaga cohort [14]. *p-value Fishers’ exact test < 0.05; **p-value Fishers’ exact test < 0.01. Fisher´s exact test for luminal A = 0.04, luminal B = 0.0694, HER2-enriched = 0.86, and basal-like = 0.000039. C Tumor-infiltrating lymphocyte (TIL) content in GBC vs. non-GBC control population from Denkert et al. (2018) [21] (in both populations patients ≤ 40 years old at diagnosis were considered). Chi. [2] test p-value = 0.034. D TIL content in GBC population according to the gestational status at the time of breast cancer diagnosis. Fishers’ test p-value = 0.048

Fig. 3

Principal Component Analysis of GBC vs. non-GBC. Principal Components analysis (PCA) of GBC (red) and non-GBC (blue) samples in (A) the global population (PC1, 2 and 3 in figure A1, and PC 2, 3 and 4 in figure A2) and (B) stratified by intrinsic subtype

Fig. 4

Metascape pathway enrichment analysis of differentially expressed genes in GBC vs. non-GBC across the different intrinsic subtypes. Multiple gene and gene sets related to cell cycle control, DNA repair and genomic stability pathways showed differential expression between GBC and non-GBC BC. Those gene and gene set are differentially distributed across all BC intrinsic subtypes