Whole-genome analysis of Nigerian patients with breast cancer reveals ethnic-driven somatic evolution and distinct genomic subtypes

Abstract

Black women across the African diaspora experience more aggressive breast cancer with higher mortality rates than white women of European ancestry. Although inter-ethnic germline variation is known, differential somatic evolution has not been investigated in detail. Analysis of deep whole genomes of 97 breast cancers, with RNA-seq in a subset, from women in Nigeria in comparison with The Cancer Genome Atlas (n = 76) reveal a higher rate of genomic instability and increased intra-tumoral heterogeneity as well as a unique genomic subtype defined by early clonal GATA3 mutations with a 10.5-year younger age at diagnosis. We also find non-coding mutations in bona fide drivers (ZNF217 and SYPL1) and a previously unreported INDEL signature strongly associated with African ancestry proportion, underscoring the need to expand inclusion of diverse populations in biomedical research. Finally, we demonstrate that characterizing tumors for homologous recombination deficiency has significant clinical relevance in stratifying patients for potentially life-saving therapies.

Fig. 1. Landscape of driver genes in breast cancer across different ethnic groups.

Genes were identified using two different detection methods (cDriver and MutSigCV; n = 13). Breast cancer drivers not detected due to insufficient statistical power (n = 173 independent tumors) but frequent in this dataset (≥2%) were also added to the overall list of drivers of breast cancer (n = 30) to visualize their distribution in the Nigerian, Black, and White groups. Multi-hit: more than one non-silent variant detected in a gene in one tumor.

Fig. 2. Manhattan plot for a genome-wide somatic non-coding variant enrichment analysis in the Nigerian group.

The dotted horizontal line represents the genome-wide significance threshold (pairwise Fisher’s exact test two-sided P-values adjusted by Benjamini–Hochberg FDR < 0.1) with two bins showing significant enrichment of somatic non-coding variants in two previously unreported breast cancer drivers.

Fig. 3. Mutational signatures in breast cancer tumors across different ethnic groups.

a, b Single-base substitution (SBS) signatures in all groups. a from top to bottom: number of tumors with SBS signatures across the entire dataset (the dotted line represents total sample size, n = 173) with signatures sorted left to right by descending frequency, number of mutations per sample (color representing groups) in respective signatures and proportion of samples carrying each signature in each group. b proportion of mutations assigned to each SBS signature across the three groups. c, d Identical plots as in a, b respectively for insertion/deletion (INDEL) signatures identified in the three groups. INDEL-B is a previously unreported signature characterized mainly by 5+bp insertions. Source data are provided as a Source Data file.

Fig. 4. Homologous recombination deficiency analysis in breast cancer tumors across different ethnic groups.

Analysis of homologous recombination deficiency (HRD) in all three groups with samples ordered based on CHORD HRD type (i.e., proficient, BRCA2-type deficient, and BRCA1-type deficient). a Probability score of BRCA1- and BRCA2-type HRD and b prediction of HRD type (P(BRCA1) + P(BRCA2) > 0.5 cut-off for HRD). Unlabeled samples in b are considered homologous repair proficient. c Somatic mutations identified in homologous recombination repair pathway genes BRCA1, BRCA2, and PALB2. d Structural variant (SV) signature extraction identified seven signatures. S3 and S2 & S6 correspond to previously identified SV signatures correlated to BRCA1- and BRCA2-deficiency, respectively. These results were compared to HRD-associated e single-base substitution signature 3 (SBS3) and f insertion/deletion (INDEL) signatures 6 and 8 (ID6/ID8). g Clinical subtype of each sample.

Fig. 5. Copy number landscape of Nigerian breast tumors.

a Genome-wide landscape of gain, loss of heterozygosity (LOH), and homozygous deletion (HD) events in the Nigerian group. The y-axis represents fraction of tumors with a particular event. LOH and HD are shown in opposite direction for better visualization. b Differential landscape (Nigerian versus White) of LOH events in HER2+ tumors. Events in the positive direction are more frequent in the Nigerian group. 14q LOH was virtually exclusive to the Nigerian HER2+ subtype. Source data are provided as a Source Data file.

Fig. 6. Chronological ordering of genomic events in Nigerian breast tumors.

Clonality-based ordering of significantly enriched copy number events (FDR < 0.05), whole-genome duplication (WGD), and key frequent mutational drivers (TP53 and GATA3). A Plackett-Luce model was used to order the events by sampling from all possible tumor phylogenies across the entire dataset (1,000 iterations). Violins represent the 95% confidence interval of the relative timing estimate for each event. The events are ordered early to late by the mean value of the relative timing estimates. The vertical dotted line represents the mean timing estimate of WGD across all samples. LOH loss of heterozygosity, HD homozygous deletion, Mut mutational driver.

Fig. 7. Somatic interaction analysis in breast cancer tumors.

Pairwise associations within and between mutational drivers and significantly enriched copy number aberrations were assessed using pairwise Fisher’s exact test. Significant associations (two-sided P-values adjusted for multiple testing, FDR < 0.05) are shown with positive associations (co-occurrence; OR > 1) having a positive sign (red) and negative associations (mutual exclusivity; 0 < OR < 1) having a negative sign (blue).

Fig. 8. Copy number burden across genomic subtypes identified in the Nigerian group.

Distribution of significantly enriched copy number aberrations in the form of gain, loss of heterozygosity (LOH), and homozygous deletion (HD) were assessed for all Nigerian patients. GATA3-positive tumors displayed a lower copy number burden than TP53-positive tumors but were similar to tumors negative for both TP53 and GATA3.