Abstract

Importance

   Triple-negative breast cancer (TNBC) is the most aggressive breast
   cancer subtype and appears to have disproportionately higher incidence
   and worse outcomes among younger African American females.

Objective

   To investigate whether epigenetic differences exist in TNBCs of younger
   African American females that may explain clinical disparities seen in
   this patient group.

Design, Setting, and Participants

   This cross-sectional study used clinical, demographic, DNA methylation
   (HumanMethylation450; Illumina), and gene expression (RNA sequencing)
   data for US patient populations from publicly available data
   repositories (The Cancer Genome Atlas [TCGA], 2006-2012, and Gene
   Expression Omnibus [GEO], 2004-2013) accessed on April 13, 2021. White
   and African American females with TNBC identified in TCGA (69 patients)
   and a validation cohort of 210 African American patients from GEO
   ([38]GSE142102) were included. Patients without available race or age
   data were excluded. Data were analyzed from September 2022 through
   April 2023.

Main Outcomes and Measures

   DNA methylation and gene expression profiles of TNBC tumors by race
   (self-reported) and age were assessed. Age was considered a dichotomous
   variable using age 50 years as the cutoff (younger [<50 years] vs older
   [≥50 years]).

Results

   A total of 69 female patients (34 African American [49.3%] and 35 White
   [50.7%]; mean [SD; range] age, 55.7 [11.6; 29-82] years) with TNBC were
   included in the DNA methylation analysis; these patients and 210
   patients in the validation cohort were included in the gene expression
   analysis (279 patients). There were 1115 differentially methylated
   sites among younger African American females. The DNA methylation
   landscape on TNBC tumors in this population had increased odds of
   enrichment of hormone (odds ratio [OR], 1.82; 95% CI, 1.21 to 2.67;
   P = .003), muscle (OR, 1.85; 95% CI, 1.44 to 2.36; P < .001), and
   proliferation (OR, 3.14; 95% CI, 2.71 to 3.64; P < .001) pathways vs
   other groups (older African American females and all White females).
   Alterations in regulators of these molecular features in TNBCs of
   younger African American females were identified involving hormone
   modulation (downregulation of androgen receptor: fold change
   [FC] = −2.93; 95% CI, −4.76 to −2.11; P < .001) and upregulation of
   estrogen-related receptor α (FC = 0.86; 95% CI, 0.34 to 1.38;
   P = .002), muscle metabolism (upregulation of FOXC1: FC = 1.33; 95% CI,
   0.62 to 2.03; P < .001), and proliferation mediators (upregulation of
   NOTCH1: FC = 0.71; 95% CI, 0.23 to 1.19; P = .004 and MYC (FC = 0.81;
   95% CI, 0.18 to 1.45; P = .01).

Conclusions and Relevance

   These findings suggest that TNBC of younger African American females
   may represent a distinct epigenetic entity and offer novel insight into
   molecular alterations associated with TNBCs of this population.
   Understanding these epigenetic differences may lead to the development
   of more effective therapies for younger African American females, who
   have the highest incidence and worst outcomes from TNBC of any patient
   group.

Introduction

   Breast cancer is a heterogeneous disease divided into 4 main subtypes
   based on distinct pathological and molecular features. Triple-negative
   breast cancer (TNBC), which lacks estrogen receptor (ER) and
   progesterone receptor expression and human epidermal growth factor
   receptor 2 (ERBB2, formerly HER2) overexpression, is the most
   aggressive subtype, with more patients progressing to metastatic
   disease than any other breast cancer subtype. African American females
   develop TNBC at more than twice the rate of White females and have a
   one-third higher risk of dying from this disease.^[39]1,[40]2,[41]3
   Younger African American females have the highest incidence of TNBC
   (39%) compared with older African American females (14%) and White
   females of all ages (16%).^[42]4 The basis of this marked disparity in
   incidence and outcome is poorly understood.

   Studies have suggested that socioeconomic factors and treatment
   variables may account for the disparities in
   survival.^[43]5,[44]6,[45]7 However, studies from 2014^[46]2 and
   2009^[47]3 suggested that biological differences may also contribute.
   We and others found that after adjusting for socioeconomic, treatment,
   and tumor variables, younger African American females with TNBC had
   lower rates of response to chemotherapy than younger White
   females.^[48]8,[49]9 Furthermore, African American females had a higher
   proportion of basal-like and mesenchymal stem–like tumors, subtypes of
   TNBC associated with worse prognoses.^[50]10 Understanding molecular
   differences that contribute to the more aggressive behavior of TNBC in
   this patient population is essential for improving outcomes.

   Epigenetic modifications, such as DNA methylation (DNAm), represent
   changes to the structure of DNA that affect gene expression without
   altering the underlying gene sequence. The epigenetic landscape is
   determined by a person’s genetic makeup and a variety of environmental
   factors, such as diet, stress, and toxins. Importantly, epigenetic
   alterations have been implicated in breast cancer development and
   progression. A 2012 study^[51]11 suggested that TNBC was epigenetically
   distinct from other breast cancer subtypes, and we previously
   identified 4 prognostically relevant epigenetic subtypes within
   TNBC.^[52]12

   However, the contribution of epigenetics to the pathogenesis of TNBC by
   race and age remains poorly understood. In this exploratory study, we
   sought to identify whether TNBCs in younger African American females
   have epigenetic differences associated with altered molecular pathways
   that may contribute to the higher incidence and lower survival seen in
   this patient group.

Methods

   Data for this cross-sectional study were gathered from publicly
   available data sets and did not require institutional review board
   approval or patient informed consent per the Code of Federal
   Regulations (45 CFR §46.104). This study followed the Strengthening the
   Reporting of Observational Studies in Epidemiology ([53]STROBE)
   reporting guideline.

Data Collection and Normalization

   Clinical and demographic data from US patients with TNBC in the Cancer
   Genome Atlas (TCGA) 2006 to 2012 were obtained from the National Cancer
   Institute Genomic Data Commons portal. Race (African American, American
   Indian or Alaska Native, Asian, or White) and ethnicity (Hispanic or
   non-Hispanic) for each participant was self-reported and annotated by
   TCGA data abstractors. Ethnicity was considered to compare the
   incidence of TNBC by race and ethnicity group, but ethnicity was not
   used as a classification variable for the remainder of the study.
   Groups were generated based on race and age. Only treatment-naive
   tumors from African American and White patients were included in the
   study given the small number of patients of other races with available
   DNAm and gene expression data. DNAm (HumanMethylation450 array;
   Illumina) and gene expression (RNA sequencing) data were downloaded
   using the R statistical software package TCGAbiolinks version 2.16.4 on
   April 13, 2021. Patients of other races or without race or age data
   were excluded (88 patients). Groups represented in Figure 1 include
   non-Hispanic African American, non-Hispanic Asian, Hispanic or Latino
   White, and non-Hispanic White patients. DNA methylation and gene
   expression data were from patients self-reported as African American or
   White based on race alone, and ethnicity was not included (given the
   small number of Hispanic patients with DNAm data). African American
   patients with TNBC who had RNA sequencing data enrolled in the Gene
   Expression Omnibus (GEO; [54]GSE142102) 2004 to 2013 were included as a
   validation cohort. Gene expression data were normalized to Z score
   using the scale command in R. Probes not passing GenomeStudio quality
   controls (Illumina) and those with known cross-reactivity or missing
   data were excluded. All probes with single nucleotide variations
   (formerly single nucleotide polymorphisms) located in or near (within
   25 base pairs) repetitive elements were removed. DNAm levels were
   obtained as β values. Chromatin immunoprecipitation sequencing of
   estrogen-related receptor α (ESRRA) binding sites were downloaded from
   the University of California, Santa Cruz Genome Browser and androgen
   receptor (AR) binding sites from GEO ([55]GSE83860).

Statistical Analysis

   Patients were grouped based on race and age into 4 groups: younger
   African American (aged <50 years), older African American (aged ≥50
   years), younger White (aged <50 years), and older White (aged ≥50
   years) female patients. Age was dichotomized based on studies^[56]9
   finding an increased risk of mortality in African American patients
   younger than 50 years. The Uniform Manifold Approximation and
   Projection for Dimension Reduction (UMAP) technique was used to
   visualize the distribution of groups using the R M3C package version
   1.18.0. Heat maps were used to visualize hierarchical cluster analyses
   based on Euclidean distance using the R package gplots version 3.1.3.
   Correlation matrices were computed using R statistical software version
   4.2.0 (R Project for Statistical Computing) and plotted using the R
   gplots2 package. Phylogenetic trees with Spearman ρ distances between
   DNAm profiles of the 4 groups were generated using the FigTree software
   version 1.4.3 (Andrew Rambaut Lab, University of Edinburgh). We
   compared categorical variables using the Fisher exact test and age with
   the Kruskal-Wallis test. Differences between groups were computed using
   Dunn multiple comparison tests in R. Pathway enrichment scores were
   estimated using the hypergeometric test, and the χ^2 test was used to
   calculate the P value. A 2-sided P value < .05 was considered
   significant. Gene set enrichment analysis (GSEA) was performed using
   GSEA software version 4.3.2 (University of California, San Diego, and
   the Broad Institute) downloaded from the project
   website.^[57]13,[58]14,[59]15 Pathway visualization was performed using
   Cytoscape and the KEGGscape plugin.^[60]16,[61]17 Data were analyzed
   from September 2022 through April 2023.

   Differentially methylated sites (DMSs) were computed using the Wilcoxon
   test. All CpG sites with a differential mean β value of 15% or greater
   and a P value < .05 were considered DMSs. For the gene expression
   analysis, the number of counts was converted to a log2 value. The
   parametric t test was used to assess the significance of differentially
   expressed genes (DEGs), defined as an absolute fold difference between
   groups greater than 0.7 and a P value < .05. Expression-methylation
   quantitative trait loci (emQTL) involving 9398 CpGs in 870 genes were
   obtained from Fleischer et al.^[62]18 The Genomic Regions Enrichment of
   Annotations Tool (Stanford University) was used to identify genes with
   DMSs. Pathway enrichment analysis was performed using the R GOfuncR
   package version 1.16.0. Genes associated with relevant pathways were
   selected using the R biomaRT package version 2.52.0. Pathway enrichment
   was represented using the R ggplot2 package version 3.3.6.

Results

Patient and Clinicopathologic Data

   We identified 69 female patients (34 African American [49.3%] and 35
   White [50.7%]; mean [SD; range] age, 55.7 [11.6; 29-82] years),
   including 16 younger African American (mean [SD; range] age, 45.8 [4.1;
   35-50] years), 18 older African American (mean [SD; range] age, 64.9
   [9.4; 51-80] years), 12 younger White (mean [SD; range] age, 44.2 [6.4;
   29-50] years), and 23 older White (mean [SD; range] age, 61.4 [8.3;
   51-82] years) females. African American patients had proportionally
   higher rates of TNBC than patients with other races and ethnicities
   ([63]Figure 1A). For this analysis, self-reported Latino or Hispanic
   females were considered as a group to establish the proportion of
   patients with TNBC. This population included 4 Hispanic or Latino White
   (5.2%), 35 non-Hispanic African American (45.5%), 5 non-Hispanic Asian
   (6.5%), and 33 non-Hispanic White (42.8%) females. However, due to the
   small proportion in this group, ethnicity was removed as a
   classification factor for further analyses. When including only
   patients with DNAm data, our study population included 35 White
   individuals (2 Hispanic White and 33 non-Hispanic White individuals)
   and 34 non-Hispanic African American individuals. Tumor histology,
   location, stage, and nodal status were similar between groups (eTable
   in [64]Supplement 1).

Figure 1. DNA Methylation Differences in Breast Cancer by Race.

   Figure 1.
   [65]Open in a new tab

   A, The percentage of patients by race category among breast cancer
   subtypes is presented. B, The uniform manifold approximation and
   projection displays the distribution of triple-negative breast cancer
   (TNBC) specimens according to differentially methylated sites. C, The
   phylogenetic tree representing the DNA methylation–based Spearman ρ
   distance between TNBC tumors from different race and age categories.
   ERBB2 indicates human epidermal growth factor receptor 2; HR, hormone
   receptor.

Epigenetic Differences by Race

   We identified 164 DMSs between African American and White patients with
   TNBC (eFigure 1A in [66]Supplement 1), which demonstrated a modest
   ability to distinguish patients based on race ([67]Figure 1B; eFigure
   1B in [68]Supplement 1). Therefore, we added age to further stratify
   groups. By comparing DNAm levels of the top 5000 most variable genomic
   regions, we found that TNBCs of younger African American patients
   displayed a distinct epigenetic profile compared with TNBCs of other
   groups (younger White , older White, and older African American
   patients) (eFigure 1C in [69]Supplement 1). This observation persisted
   when we examined the top 2000 and top 1000 most variable regions
   (eFigure 1D-E in [70]Supplement 1). However, when we selected less than
   1000 regions, TNBCs of younger African American females clustered with
   those of older African American females and remained separate from
   White females of any age (eFigure 1F-G in [71]Supplement 1). This
   suggests that this small set of hypervariable genomic regions (<500
   regions) was associated more with race than age. TNBCs of younger
   African American patients displayed a longer statistical distance from
   other patient groups ([72]Figure 1C), indicating a distinct epigenetic
   landscape.

Association of Epigenetic Alterations With Pathways in Younger African
American Patients

   Among 1115 DMSs between younger African American patients and other
   groups, we identified 80 hypermethylated and 1035 hypomethylated
   regions (eFigure 2A in [73]Supplement 1). In hierarchical clustering
   and UMAP analyses, these regions effectively separated younger African
   American patients from other groups ([74]Figure 2A; eFigure 2B in
   [75]Supplement 1). We conducted associative pathways enrichment
   analyses using genes located near the younger African American
   patient–specific hypomethylated sites and found significantly increased
   odds of enrichment in hormone (odds ratio [OR], 1.82; 95% CI,
   1.21-2.67; P = .003), muscle (OR, 1.85; 95% CI, 1.44-2.36; P < .001),
   and proliferation (OR, 3.14; 95% CI, 2.71-3.64; P < .001) pathways
   compared with other groups (older African American females and White
   females of all ages) ([76]Figure 2B).

Figure 2. DNA Methylation, Gene Expression, and Pathway Enrichment in Younger
African American Compared With White and Older African American Patients.

   Figure 2.
   [77]Open in a new tab

   A, The uniform manifold approximation and projection displays the
   distribution of triple-negative breast cancer specimens according to
   differentially methylated sites between younger (aged <50 years)
   African American females and other groups (older [aged ≥50 years]
   African American, younger White, and older White females). B, Lollipop
   plots represent enrichment values for hormone, muscle, and
   proliferation pathways. C, The volcano plot represents the gene
   expression fold change and −log10 of the P value between
   triple-negative breast cancer from younger African American females and
   other groups. D, The bar plot shows the fold change (dots) and −log10 P
   value (bars) of genes associated with hormone, muscle, and
   proliferation pathways. AR indicates androgen receptor.

   We integrated gene expression profiles and identified 2864 DEGs, of
   which 1073 DEGs were upregulated and 1791 DEGs were downregulated in
   TNBCs of younger African American patients ([78]Figure 2C). A
   significant proportion of these DEGs had emQTL (OR, 1.29; 95% CI, 1.18
   to 1.40; P < .001), underscoring the importance of epigenetic
   regulation in maintaining the specific transcriptome of TNBCs of
   younger African American patients (eFigure 2C in [79]Supplement 1).
   Similar to epigenetic alterations, gene expression changes were
   identified in hormone, muscle, and proliferation pathways.
   Downregulation of AR and upregulation of ESRRA were among the most
   relevant DEGs involved in hormone metabolism. We additionally
   identified significant upregulation of FOXC1, a transcription factor
   involved in muscle cell metabolism, and upregulation of MYC and NOTCH1,
   2 key proliferative mediators ([80]Figure 2D). The gene set enrichment
   analysis using gene expression data revealed a decreased response to
   hormones, specifically estradiol, in younger African American patients
   and an activation of gene networks involved in cell cycle proliferation
   (eFigure 2D-E in [81]Supplement 1).

Epigenetic Differences by Age in African American Patients

   To investigate whether epigenetic differences detected in TNBCs of
   younger African American females were associated with race, we compared
   DNAm patterns between tumors of younger and older African American
   females. We identified 106 hypermethylated and 2800 hypomethylated
   genomic regions in younger African American females (eFigure 3A in
   [82]Supplement 1). DNAm levels of these genomic regions separated
   younger and older African American patients in multidimensional
   reduction analyses ([83]Figure 3A; eFigure 3B in [84]Supplement 1).
   Genes located near genomic regions that were hypomethylated in younger
   African American compared with older African American patients were
   also enriched in hormone (OR, 2.76; 95% CI, 1.96-3.81; P < .001),
   muscle (OR, 1.65; 95% CI, 1.30-2.09; P < .001), and proliferation (OR,
   2.76; 95% CI, 2.38-3.18; P < .001) pathways ([85]Figure 3B), the same
   alterations detected when comparing younger African American patients
   with other groups.

Figure 3. DNA Methylation, Gene Expression, and Pathway Enrichment in Younger
vs Older African American Females.

   Figure 3.
   [86]Open in a new tab

   A, The uniform manifold approximation and projection displays the
   distribution of triple-negative breast cancer specimens according to
   differentially methylated sites between younger (aged <50 years) and
   older (aged ≥50 years) African American females. B, Lollipop plots
   represent enrichment values of hormonal, muscle, and proliferation
   pathways. C, The volcano plot represents the gene expression fold
   change and −log10 of the P value between triple-negative breast cancer
   from younger and older African American females. D, The bar plot shows
   fold change (dots) and −log10 P value (bars) of genes associated with
   hormone, muscle, and proliferation pathways. AR indicates androgen
   receptor.

   In addition, comparative gene expression analysis revealed 405
   upregulated and 914 downregulated genes in TNBC tumors of younger
   compared with older African American patients ([87]Figure 3C). Several
   of these genes were associated with increased odds of emQTL (OR, 1.72;
   95% CI, 1.53-1.93; P < .001) (eFigure 3C in [88]Supplement 1). AR was
   also significantly downregulated ([89]Figure 3D), suggesting that
   epigenetic and transcriptomic alterations in hormone, muscle, and
   proliferation pathways identified in younger African American patients
   were different from those even in older African American patients.

Epigenetic Differences Between Younger African American and White Females

   Next, we examined whether alterations detected in TNBC tumors of
   younger African American patients were associated with age rather than
   race by comparing DNAm levels between younger African American and
   younger White patients. Analysis demonstrated 1515 differentially
   methylated sites (519 hypermethylated and 996 hypomethylated sites)
   (eFigure 4A in [90]Supplement 1), which efficiently separated the 2
   groups ([91]Figure 4A; eFigure 4B in [92]Supplement 1). These
   differentially methylated genomic regions were associated with muscle
   (OR, 1.50; 95% CI, 1.10-1.99; P = .008) and proliferation (OR, 2.14;
   95% CI, 1.78-2.56; P < .001) but not hormone (OR, 1.23; 95% CI,
   0.70-1.99; P = .46) pathways ([93]Figure 4B), although a higher
   enrichment was seen in younger African American patients. We also
   observed 2137 downregulated and 1625 upregulated genes in TNBCs of
   younger African American compared with younger White patients
   ([94]Figure 4C). Several DEGs were associated with higher odds of emQTL
   (OR, 1.10; 95% CI, 1.02-1.20; P = .02) (eFigure 4C in [95]Supplement 1)
   and involved in muscle and proliferation pathways ([96]Figure 4D).
   Genes involved in hormone response, such as estrogen receptor α (ESR1)
   and AR, were significantly downregulated in TNBCs of younger African
   American females. This observation suggests that although the
   association of hormones with outcomes in the TNBC epigenome was similar
   in younger females regardless of race, specific pathways, such as ER
   and AR response, were less active in TNBCs of younger African American
   females.

Figure 4. DNA Methylation, Gene Expression, and Pathway Enrichment in Younger
African American Compared With Younger White Females.

   Figure 4.
   [97]Open in a new tab

   A, The uniform manifold approximation and projection displays the
   distribution of triple-negative breast cancer specimens according to
   differentially methylated sites between younger (aged <50 years)
   African American and younger White females. B, Lollipop plots represent
   enrichment values of hormone, muscle, and proliferation pathways. C,
   The volcano plot represents the gene expression fold change and −log10
   of the P value between triple-negative breast cancer from younger
   African American and younger White females. D, The bar plot shows fold
   change (dots) and −log10 P value (bars) of genes associated with
   hormone, muscle, and proliferation pathways.

Upstream Factors Associated With Hormone, Myoepithelial, and Proliferation
Pathways in Younger African American Females

   Based on epigenetic and transcriptomic differences of TNBCs in younger
   African American females, we examined associations of upstream factors
   with significant changes in molecular functions. As reported
   elsewhere,^[98]19 we observed a downregulation of the AR gene in
   hormone receptor–negative and TNBC tumors (FC = −3.40; 95% CI, −3.93 to
   −2.87; P < .001) (eFigure 5A-B in [99]Supplement 1). Among patients
   with TNBC, younger African American patients had the lowest AR levels
   (FC = −2.93; 95% CI, −4.76 to −2.11; P < .001) ([100]Figure 5A), and
   the expression of AR was correlated with ESR1 expression, suggesting a
   common regulation of these factors (cor = 0.53; 95% CI, 0.33 to 0.68;
   P < .001) (eFigure 5C in [101]Supplement 1). Concurrent with a lower
   expression level of AR, a higher expression level of ESRRA in hormone
   receptor–negative and TNBC tumors (FC = 0.86; 95% CI, 0.34 to 1.38;
   P < .001) was observed (eFigure 5D-E in [102]Supplement 1),
   specifically in younger African American patients (FC = 0.86; 95% CI,
   0.34-1.38; P = .002) ([103]Figure 5A). ESRRA is an orphan nuclear
   receptor that has a negative correlation with ESR1, even in ER-positive
   tumors (cor = −0.29; 95% CI, −0.35 to −0.21; P < .001) (eFigure 5F in
   [104]Supplement 1). We explored the epigenetic status of genomic
   regions with binding affinity for ESRRA and AR and observed that ESRRA
   binding sites were significantly hypomethylated in TNBCs of younger
   African American patients, suggesting a functional activation. In
   concordance, genes associated with these hypomethylated elements were
   significantly upregulated (eFigure 5G in [105]Supplement 1), while
   genes with AR binding sites in their promoters were downregulated
   (eFigure 5H in [106]Supplement 1).

Figure 5. Basal-Like Phenotype–Associated Genes Differentially Expressed in
Younger African American Females.

   Figure 5.
   [107]Open in a new tab

   A, Hormone-associated genes (AR and ESRRA) are presented. B, Muscle and
   mammary gland differentiation–associated genes (FOXC1 and FOXA1,
   respectively) are presented. C, Proliferation-associated genes (MYC and
   NOTCH1) are presented. Younger was defined as aged less than 50 years.

   Owing to its significance in regulating hormone response, AR negativity
   is used to define a subset of TNBCs known as quadruple-negative breast
   cancers (QNBCs). We observed a correlation (r = 0.36; P < .001)
   (eFigure 5I in [108]Supplement 1) between gene expression profiles of
   TNBCs of younger African American patients and QNBC tumors of 161
   patients in the Molecular Taxonomy of Breast Cancer International
   Consortium (METABRIC) cohort (eFigure 5I in [109]Supplement 1).^[110]20
   For example, while MMP7 and FOXC1 genes were upregulated, MUCL1 and
   FOXA1, factors that maintain mammary cell differentiation, were
   downregulated in both groups. These similarities suggest that AR
   negativity may have been associated with the gene expression profile of
   TNBCs of younger African American patients, suggesting the importance
   of QNBCs in these patients.

Evaluation of Myoepithelial Gene Pathways in TNBCs of Younger African
American Females

   Our data suggest that in addition to having a highly
   hormone-insensitive phenotype, TNBCs of younger African American
   females had a greater basal-like phenotype. FOXC1, a transcription
   factor active in myoepithelial cells,^[111]21 was significantly
   upregulated (FC = 1.33; 95% CI, 0.62 to 2.03; P < .001) ([112]Figure
   5B), and FOXA1, involved in mammary gland differentiation and
   modulation of ER and AR functions in breast cancer,^[113]22 was
   significantly downregulated (FC = −2.74; 95% CI, −4.24 to −1.23;
   P < .001) ([114]Figure 5B). This pattern was also observed in patients
   with TNBCs compared with those with HR-positive cancers (FOXC1:
   FC = 3.17; 95% CI, 2.86 to 3.50; P < .001; FOXA1: FC = −4.73; 95% CI,
   −5.30 to −4.16; P < .001) (eFigure 6A-D in [115]Supplement 1). FOXC1
   was negatively correlated with FOXA1 in patients with TNBCs
   (cor = −0.71; 95% CI, −0.81 to −0.56; P < .001) (eFigure 6E in
   [116]Supplement 1), highlighting their opposing role in regulating gene
   expression programs involved in the basal-like phenotype.^[117]23
   Supporting this observation, we identified that all TNBCs of younger
   African American patients were classified as basal-like by the
   Prediction Analysis of Microarray 50 (PAM50) algorithm, while other
   groups had higher variability (eFigure 6F in [118]Supplement 1).

Proliferative Gene Networks Among TNBC Tumors by Group

   TNBCs of younger African American patients also displayed an
   upregulation of genes involved in proliferation pathways, such as MYC
   (FC = 0.81; 95% CI, 0.18-1.45; P = .01) and NOTCH1 (FC = 0.71; 95% CI,
   0.23-1.19; P = .004) ([119]Figure 5C). Additional genes involved in
   both pathways were significantly upregulated in these tumors,
   suggesting activation of these pathways in TNBCs of younger African
   American females (eFigure 7A-B in [120]Supplement 1). MYC (FC = 0.75;
   95% CI, 0.52-0.99; P < .001) (eFigure 8A-B in [121]Supplement 1) and
   NOTCH1 (FC = 0.97; 95% CI, 0.81-1.13; P < .001) (eFigure 8C-D in
   [122]Supplement 1) were also upregulated in TNBC tumors compared with
   other breast cancer subtypes.

   We validated results of this study using an external cohort of gene
   expression data from TNBCs of African American females generated by
   Purrington et al (210 patients, for a total of 279 patients in the gene
   expression analysis).^[123]24 In agreement with our core analysis, ESR1
   (FC = −0.42; 95% CI, −0.69 to −0.16; P = .002), FOXA1 (FC = −0.37; 95%
   CI, −0.62 to −0.11; P = .004), and AR (FC = −0.28; 95% CI, −0.53 to
   −0.02; P = .03) were downregulated in younger compared with older
   African American patients in the validation analysis (eFigure 9A-C in
   [124]Supplement 1). In addition, ESRRA (FC = 0.24; 95% CI, −0.04 to
   0.53; P = .09), FOXC1 (FC = 0.14; 95% CI, −0.15 to 0.42; P = .34),
   NOTCH1 (FC = 0.14; 95% CI, −0.14 to 0.43; P = .33), and MYC (FC = 0.11;
   95% CI, −0.18 to 0.40; P = .45) showed no significant upregulation in
   younger compared with older African American females (eFigure 9D-G in
   [125]Supplement 1).

Discussion

   TNBC is a particularly aggressive disease, with higher rates of death
   than any other breast cancer subtype. Despite significant advances in
   breast cancer research and treatment, disparities in incidence and
   outcome persist between female patients of different races and ages.
   Specifically, younger African American females develop TNBC at more
   than twice the rate of older African American females and White females
   of any age and exhibit lower overall survival.^[126]1 Beyond
   socioeconomic and treatment variables, understanding the biological
   factors that may be associated with this disparity remains a major
   challenge. In this exploratory cross-sectional study, we identified a
   distinct DNAm landscape of TNBCs of younger African American patients.
   Furthermore, we found that the significantly hypomethylated regions
   identified in this patient group were enriched in genes associated with
   hormone, muscle, and proliferation pathways. These epigenetic
   differences and their associated outcomes in gene expression programs
   may be associated with the increased aggressivity of TNBC reported in
   this patient group.

   Hormone pathways were consistently found to harbor epigenetic and
   transcriptomic alterations in younger African American females.
   However, we did not observe differential DNAm of hormone pathways
   between younger African American and younger White females, suggesting
   that these epigenetic mechanisms may be associated with the
   pathogenesis of TNBC in younger patients regardless of race. However,
   we identified significant differences in the expression of specific and
   relevant hormone pathway mediators, such as ESR1, AR, and ESRRA,
   suggesting a more hormone-unresponsive phenotype in TNBCs of younger
   African American patients, even compared with younger White patients.

   Prior studies^[127]25 have reported that African American females have
   higher rates in general of QNBC, a TNBC subtype characterized by the
   lack of AR expression. Our data suggest that the higher incidence of
   QNBC may be specific to younger African American patients, demonstrated
   by the significant downregulation of AR in TNBC tumors of this
   population compared with older African American patients. Furthermore,
   younger African American patients exhibited a significant upregulation
   in ESRRA expression, an orphan receptor with similar binding sites to
   ESR1. Notably, ESRRA has been found to have a negative correlation with
   ESR1 and is upregulated in TNBC in general, suggesting increased
   activity in patients with low estrogen signaling.^[128]26 Our data
   suggest that TNBCs of younger African American patients had even lower
   ESR1 and AR expression and higher ESRRA expression compared with TNBCs
   of other patient groups, indicating a greater hormone insensitivity in
   this patient group.

   The upregulation of FOXC1 combined with the downregulation of FOXA1
   suggests that TNBCs of younger African American patients may have a
   more basal phenotype in addition to a more hormone-unresponsive
   phenotype. FOXC1 is enriched in myoepithelial cells and mammary luminal
   progenitor cells, and therefore its upregulation and the concomitant
   downregulation of the mammary gland differentiation mediator FOXA1
   leads to a less differentiated phenotype.^[129]21 This phenomenon has
   been previously reported, with strong evidence indicating that
   FOXC1-high/FOXA1-low breast cancers tend to have basal-like phenotypes,
   regardless of the immunohistochemistry evaluation of ER, progesterone
   receptor , or HER2.^[130]23 Furthermore, this basal- and
   myoepithelial-like phenotype of TNBC of younger African American
   females supports a decreased hormone responsiveness given that FOXC1
   can silence ESR1 by competing with GATA3.^[131]27

   Finally, we identified epigenetic alterations associated with increased
   proliferation, an alteration that has been commonly described in
   basal-like breast tumors in general^[132]28 but not specifically in
   TNBCs of younger African American females, to our knowledge. We
   observed higher levels in younger African American patients of MYC and
   NOTCH1, 2 genes that are well known for their relevance in
   cancer.^[133]29,[134]30,[135]31,[136]32 Clinically, efforts are being
   made to develop anti-NOTCH targeted therapies, primarily through the
   use of γ-secretase inhibitors and monoclonal antibodies against Notch
   receptors.^[137]33 Younger African American patients with TNBC showing
   disproportionate enhancement of NOTCH1 could potentially benefit from
   these novel therapies.

Limitations

   This study has several limitations. The main limitation is that paired
   DNAm and gene expression data from patients with TNBC and
   well-annotated data, including race and age, were unavailable,
   resulting in our small sample size. This also precluded the ability to
   perform meaningful adjusted analyses. Additionally, our study involved
   the stratification of patients based on self-reported race. In the
   absence of ancestry tests to confirm race genetically, there is a risk
   that race groups in our study were misrepresented.

Conclusions

   In this exploratory cross-sectional study, we found that the epigenetic
   landscape of TNBC tumors of younger African American patients was
   epigenetically distinct, providing important insight into the potential
   mechanisms contributing to the aggressive behavior of this cancer
   within this patient group. Our findings have the potential to improve
   the clinical treatment of these individuals by unveiling possible
   targets for novel therapeutics and underscoring the significance of
   accounting for race- and age-associated epigenetic variations in TNBC.
   Future research should focus on discovering genomic regions
   contributing to these epigenetic discrepancies and their functional
   implications in associated changes in the behavior of TNBC in younger
   African American females.
   Supplement 1.

   eFigure 1. DNA Methylation Differences Between African American and
   White Females

   eFigure 2. Gene Expression Differences in Younger African American
   Females vs Other Groups

   eFigure 3. DNA Methylation and Gene Expression Differences Between
   Younger vs Older African American Females

   eFigure 4. DNA Methylation and Gene Expression Differences Between
   Younger African American vs Younger White Females

   eFigure 5. Differential Expression of AR and ESRRA by Breast Cancer
   Subtype, Correlation with ESR1, and Associated Gene Expression

   eFigure 6. FOXC1 and FOXA1 Expression Levels by Breast Cancer Subtype

   eFigure 7. Aberrantly Active Pathways on Triple-Negative Breast Cancer
   Tumors of Younger African American Females

   eFigure 8. Analysis of Proliferation Genes in Triple-Negative Breast
   Cancer

   eFigure 9. Key Genes Associated With Hormone, Muscle, and Proliferation
   Pathways in Validation Cohort
   [138]Click here for additional data file.^ (3.7MB, pdf)
   Supplement 2.

   Data Sharing Statement
   [139]Click here for additional data file.^ (97.6KB, pdf)

References