Keywords

Introduction

The terms Hispanic and Latino are used in the United States to refer to Americans who trace their heritage to Latin America or Spain. The US Census takes a self-identification approach to categorizing race/ethnicity, including those who identify themselves as Hispanic or Latino [1]. Beyond any debate of who is or is not Hispanic/Latino, in this review we use the categories of Hispanic/Latino (Latina for women) to refer to the heterogeneous group of individuals who self-identify as such, and who are assumed to have been born or have ancestors who were born in a Latin American country. Genetic studies of the distribution of continental ancestry proportions in US Hispanics/Latinos and Latin Americans show that individuals in this group are genetically diverse, with varying proportions of European, Indigenous American, African, and to a lesser extent, Asian continental ancestry [2,3,4,5,6,7]. This heterogeneous group has often been treated as a monolithic unit in cancer epidemiology, partly out of necessity because of small sample sizes and the concomitant limitations in statistical power. The number of datasets with extensive information about Latino subgroups is limited, despite the fact that individuals who self-identify as Hispanics/Latinos make up the second-largest group within the US racial/ethnic categories [8].

Breast cancer is the most common cancer among US Hispanics/Latinas [9]; however, Hispanic/Latina women in the United States have historically shown a relatively low breast cancer incidence compared to non-Hispanic White (NHW) or African American/Black (AA/B) women [10]. In addition to the differences among racial/ethnic groups, variation within different Hispanic/Latina populations in the distribution of breast cancer subtypes by ancestry has been reported [11,12,13].

Breast cancer is a heterogeneous disease with intrinsic biological subtypes associated with different prognosis [14, 15]. Tumor profiling based on transcriptional profiles, such as PAM50, are considered the “gold standard” for intrinsic subtyping [14]. As routinely used in clinical practice, breast cancer intrinsic subtypes can be approximated by immunohistochemical markers, the most typically being estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) [16, 17]. Both population-based and hospital-based studies have shown that Hispanics/Latinas have about a 20–40% higher risk of developing ER−/PR− HER2+ and triple negative breast cancer (TNBC, a tumor subtype that is ER−/PR− HER2−) compared to NHW women [18,19,20]. ER/PR− tumors have fewer treatment options and poorer prognosis than other subtypes [21,22,23,24,25,26,27,28].

Ancestry-specific genetic variants have been shown to be associated with breast cancer risk in Hispanics/Latinas, and the strength of their association is tumor subtype-specific [29, 30]. Sample size in available data for breast cancer genetic studies in Hispanic/Latina populations has been limited, which represents a significant barrier for the further exploration of the possibility of additional germline variants that could predispose US Hispanics/Latinas or Latin American women to develop ER/PR− and HER2+ tumors [31,32,33,34]. Studies reporting associations between genetic ancestry and tumor subtype provide suggestive evidence that can be informative in the design of follow-up studies aimed at identifying factors that explain the observed association. These factors can be genetic, behavioral, environmental, and structural [29, 35, 36] . Additionally, observed associations could be the result of systematic sample bias [37].

This chapter summarizes results from studies that include breast cancer subtype and genetic ancestry information from women of Latin American origin, discusses possible explanations for observed associations, and proposes future studies that could contribute to a better understanding of tumor subtype-specific risk in US Hispanics/Latinas and Latin American women.

Indigenous American/European Genetic Ancestry and Breast Cancer Subtype in Hispanics/Latinas

The risk of developing breast cancer among Hispanics/Latinas varies by genetic ancestry [25, 38]. Women with higher proportion of Indigenous American ancestry (or lower European ancestry) have a lower risk of developing breast cancer [25]. An analysis based on data from the San Francisco Bay Area Breast Cancer Study (SFBCS) was the first to report on this association [25]. This study had a sample size of 440 cases and 597 controls and included 294 ER+ and 83 ER− patients, but had no information regarding HER2 status [25]. Genetic ancestry was estimated using a panel of ~100 ancestry informative markers. The investigators did not find a statistically significant difference in Indigenous American or European genetic ancestry proportion by ER status. However, higher European ancestry was found to be associated with increased breast cancer risk even after adjustment for known risk factors (OR = 1.39; 95% CI, 1.06–2.11; P = 0.013). A study of breast cancer survival and genetic ancestry conducted with additional cases from the SFBCS (476 ER+ and 169 ER− patients) with ancestry estimations based on genome-wide genotype data reported similar findings [39]. A subsequent study including cases from the Kaiser Permanente Pathways Cohort (408 ER+ and 94 ER−) also reported no association between ER/PR or HER2 status and Indigenous American or European ancestry proportions [36].

The Peruvian Genetics and Genomics of Breast Cancer (PEGEN-BC) is an ongoing case series study, including patients from the Instituto Nacional de Enfermedades Neoplasicas in Lima, Peru, diagnosed with invasive breast cancer since 2010 [40]. Results of an analysis including 1312 PEGEN-BC study patients showed a statistically significant association between proportion of Indigenous American ancestry and tumor subtype [40]. The average Indigenous American ancestry among participants with ER/PR− HER2+ tumors was 80%, compared with 75% among participants with the ER/PR+ HER2− subtype. Genetic ancestry was estimated from genome-wide genotype data and the program ADMIXTURE [41]. A multinomial logistic regression analysis was performed, which showed that the odds of ER/PR− HER2+ breast cancer increased by a factor of 1.22 per every 10% increase in Indigenous American ancestry when using ER/PR+ HER2− as the reference group (95% CI, 1.07–1.35; P = 0.001). This study included independent replication in a combined sample of Mexican and Colombian breast cancer patients from the COLUMBUS Consortium [40].

A study that analyzed breast tumor tissue samples from 232 Colombian women provided information on average Indigenous American ancestry proportions for luminal, HER2-enriched, and TNBC tumors [13]. Subtypes were inferred using immunohistochemical surrogates from the 2013 St. Gallen International Expert Consensus, and genetic ancestry was estimated using ~100 ancestry informative markers genotyped using DNA extracted from the tumor tissue [13]. Patients with luminal tumors had an average of 39% Indigenous American ancestry: those with HER2-enriched tumors had 35% and those with TNBC 37%. These differences in ancestry between subtypes, however, were not statistically significant. Another analysis in a subset of these Colombian study samples reported a suggestive correlation between expression of the ERBB2 gene, which is the gene that codes for the HER2 receptor, and Indigenous American ancestry. In this study, ERBB2 expression was higher among women above the median of Indigenous American ancestry compared to women below the median [42].

The Breast Cancer Health Disparities Consortium includes samples from the SFBCS, the Four Corners Study, and a case/control breast cancer study from Mexico [43]. A study using data from this consortium and including 1854 NHW women and 2326 Hispanics/Latinas analyzed the influence of risk factors and genetic ancestry on breast cancer risk in postmenopausal women. ER status information was available for a subset of the Hispanic/Latina patients, with 348 cases being ER+ and 106 ER−. They did not compare genetic ancestry proportions by subtype, and instead conducted analyses testing the association between quartiles of Indigenous American ancestry and overall breast cancer, ER+ breast cancer, and ER− breast cancer [43]. All analyses showed the same trend toward higher odds of developing breast cancer in the lower quartiles compared to the highest quartile of Indigenous American ancestry, which is consistent with results from previously published studies [25, 38].

African Genetic Ancestry and Tumor Subtype in Hispanic/Latina Breast Cancer Patients

Average African genetic ancestry in Hispanics/Latinas tends to vary greatly between countries and regions, from 5% or less in samples from Chile, Argentina, and Mexico to 10% or higher in Brazil, Cuba, or Puerto Rico [3, 44,45,46,47,48]. Assessing the association between African ancestry and tumor subtype requires large sample sizes in areas where this is a relatively minor ancestral component.

The previously mentioned study of tumor tissue samples from Colombia [13] reported an association between African ancestry and ER status, with average African ancestry being higher in ER− compared to ER+ cases (P = 0.02) [13]. Women with HER2-enriched tumors had a higher proportion of African ancestry (14%), followed by those with triple negative (12%) and luminal tumors (7%). The SFBCS study did not report a statistically significant association between African ancestry proportion and ER status, which was to be expected in a relatively small set of samples with limited African ancestry representation [25]. Average African ancestry proportion in the PEGEN-BC study samples was reported at 4% and differences between tumor subtypes for this component were not statistically significant, ranging between 3 and 5% [40].

Possible Reasons Behind Genetic Ancestry and Tumor Subtype Association Results

Studies that include US Hispanics/Latinas or Latin American women are heterogeneous. Not only because of the diversity within the Hispanic/Latino category discussed in the introduction but also because of various research designs that have been used in these studies. Most studies are based on a convenience sampling strategy within hospitals, while population-based studies with information on tumor characteristics are harder to find. In the United States, studies using cancer registry data to recruit cases can approximate a population-based design. As a result of both the intrinsic heterogeneity of the Hispanic/Latino population and the biases brought by convenience sampling, studies reporting associative findings or lack of association have to be taken with caution and should be followed up with additional research that allows for further evaluation of potential causal relationships between cancer phenotypes and etiological factors.

Differences in sample size could likely explain inconsistencies in reported associations between genetic ancestry and tumor subtype. For example, studies including a few hundred cases might not detect differences of 2 or 3% in average genetic ancestry between groups, as was the case in the Kaiser Permanente Pathways study [36], which only included a small sample of HER2+ patients. The PEGEN-BC study, with a sample of 1312 cases, had more power and identified a ~22% increase in the odds of having HER2+ tumors per 10% increase in Indigenous American ancestry [40], but was not informative about African ancestry, given that this component is minor, on average, among the study patients.

Genetic heterogeneity across Latin American countries can contribute to differences in ancestry and tumor subtype associations. Disaggregating Latin American populations based on the main genetic ancestry components will allow the association between some less represented genetic components to be tested, such as African ancestry, with specific breast cancer subtypes. For example, a study including women from Cuba or Puerto Rico will have more power than a study in Peru, with only 4% average African ancestry among patients [40]. For this purpose, additional efforts to run larger studies including participants from diverse ancestry backgrounds from Latin American countries are still needed.

Observed associations might also be explained by sampling bias. For example, within the Peruvian or Colombian studies [13, 40], there might be over-representation of patients with aggressive tumors from remote regions (such as the coastal region in Colombia or the Amazonian region in Peru). If ancestry proportion were correlated with region of origin within these countries, then by analyzing patients recruited in the cancer center might lead to an association between ancestry proportion and tumor subtype if region of origin is not accounted for in the analysis [40].

Correlation between behavioral, environmental, or socioeconomic factors and genetic ancestry could partly explain observed associations between genetic ancestry and tumor subtype. For example, socioeconomic status (SES) has been shown to be correlated with genetic ancestry [20, 49]. A study exploring the association between SES and tumor subtype among Hispanics/Latinas in data from the California Cancer Registry, including women diagnosed with invasive breast cancer from 2005 to 2010, reported that women in lower SES neighborhoods had greater risk of TNBC and ER/PR− HER2+ subtypes relative to ER/PR+ HER2− (P < 0.05) [20]. Additionally, Hispanic/Latina women who resided in low SES neighborhoods had significantly increased the risk of developing and dying from hormone receptor negative tumors than hormone receptor positive tumors [20]. This information leads to plausible explanations for the association between ancestry and subtype being related to SES and behavioral or environmental exposures associated with it.

Lifestyle may also play a part in the association between ancestry and breast cancer subtype. A study, including 2023 Hispanic/Latinas from the previously mentioned Breast Cancer Health Disparities Consortium, found that among women not using menopausal hormone therapy, weight gain was associated with increased risk of ER/PR+ breast cancer, but only among those with a low young-adult BMI [50]. The association with weight gain, for those who had information on genetic ancestry, was limited to women with lower Indigenous American genetic ancestry within this group [50] and interactions were not tested. It is possible that given the association between BMI and genetic ancestry among Hispanics/Latinas, associations between BMI and tumor subtype might lead to the observed link between ancestry and subtype.

Finally, underlying population-specific genetic variation could predispose specific populations to particular subtypes, thereby explaining observed ancestry and subtype associations. In a breast cancer genome-wide association study in Hispanics/Latinas, researchers identified two variants (rs140068132 and rs7157845) within the region 6q25, upstream of the estrogen receptor 1 gene (ESR1), and breast cancer risk [29]. The minor allele is protective (associated with decreased risk) against ER− breast cancer (OR = 0.60; 95% CI, 0.53–0.67) [29]. Other genetic variants such as these could explain the differences in breast cancer subtype by ancestry [30, 51].

Next Steps and Conclusions

In this chapter, we summarized the studies that have described associations (or lack of) between breast cancer subtypes and genetic ancestry among Hispanics/Latinas. Yet, more and larger studies are needed to increase the statistical power and better representation of the ancestral, cultural, and environmental diversity of Latin America. Ongoing collaborative efforts are aiming to do so by improving the design and collection of epidemiological and genetic data from breast cancer patients and the healthy population [30].

Future research is needed to understand the potential molecular mechanisms of population-specific variants as drivers of specific subtypes during breast cancer etiology and progression. This could explain, for example, the association between African ancestry and ER− tumors or Indigenous American ancestry and HER2+ disease. Additionally, building diverse cohorts with measures that reflect exposures associated with the individual (behavioral and biological), neighborhood, and structural level variables will provide the necessary information to disentangle the different components that contribute to observed associations between genetic ancestry and overall as well as subtype-specific breast cancer risk [30].

Currently, there are insufficient studies on tumor subtype etiology and outcome predictors among Hispanics/Latinas and African Americans that include estimations of genetic ancestry [30]. Additional support toward the enhancement of existing studies or the creation of news ones will be fundamental to improve our understanding of subtype-specific breast cancer risk, treatment efficacy, and outcome prognosis in admixed minority populations.