Intrinsic subtypes in Ethiopian breast cancer patient

Abstract

Purpose: The recent development of multi-gene assays for gene expression profiling has contributed significantly to the understanding of the clinically and biologically heterogeneous breast cancer (BC) disease. PAM50 is one of these assays used to stratify BC patients and individualize treatment. The present study was conducted to characterize PAM50-based intrinsic subtypes among Ethiopian BC patients.

Patients and methods: Formalin-fixed paraffin-embedded tissues were collected from 334 BC patients who attended five different Ethiopian health facilities. All samples were assessed using the PAM50 algorithm for intrinsic subtyping.

Results: The tumor samples were classified into PAM50 intrinsic subtypes as follows: 104 samples (31.1%) were luminal A, 91 samples (27.2%) were luminal B, 62 samples (18.6%) were HER2-enriched and 77 samples (23.1%) were basal-like. The intrinsic subtypes were found to be associated with clinical and histopathological parameters such as steroid hormone receptor status, HER2 status, Ki-67 proliferation index and tumor differentiation, but not with age, tumor size or histological type. An immunohistochemistry-based classification of tumors (IHC groups) was found to correlate with intrinsic subtypes.

Conclusion: The distribution of the intrinsic subtypes confirms previous immunohistochemistry-based studies from Ethiopia showing potentially endocrine-sensitive tumors in more than half of the patients. Health workers in primary or secondary level health care facilities can be trained to offer endocrine therapy to improve breast cancer care. Additionally, the findings indicate that PAM50-based classification offers a robust method for the molecular classification of tumors in the Ethiopian context.

Keywords: Breast cancer; Ethiopia; Intrinsic subtyping; PAM50.

Fig. 1

Gene Expression heatmap of the 50 loci used for the PAM50 classification of 334 BC samples. The 334 samples are grouped horizontally according to their intrinsic subtype, which are indicated above each block. Red tiles denote overexpression, while green tiles correspond to underexpression. The four horizontal bars above the heatmaps indicate the classification of samples according to IHC groups, tumor grade, tumor size and Ki-67 proliferation index (top-down, respectively, with color codes for each bar given at the right side of the figure). HR hormone receptor, HER2 human epidermal growth factor receptor 2

Fig. 2

Color-coded crosstable of 334 BC tissue samples grouped according to IHC groups (rows) and intrinsic subtypes (columns). The cell color gradient indicates the relationship in terms of a strong discordance (white) to a strong concordance (black) between IHC and intrinsic subtype classifications. HR hormone receptor, HER2 human epidermal growth factor receptor 2

Fig. 3

Scores of Estrogen Receptor, Progesterone Receptor, HER2 and Ki-67 Proliferation Index among intrinsic subtypes. The horizontal widths of the violin plots correspond to the frequency distribution of data points. Horizontal bars shown in the plots denote data available for those values. ER estrogen receptor, PgR progesterone receptor, IRS immune-reactive score, ER estrogen receptor, PgR progesterone receptor, HER2 human epidermal growth factor receptor 2