Breast Cancer Research in the Caribbean: Analysis of Reports From 1975 to 2017.

PURPOSE: Breast cancer is among the leading causes of death resulting from cancer in Caribbean women. Studies examining exogenous and genetically predetermined endogenous risk factors are critical to define breast cancer susceptibility in Caribbean women. The purpose of this systematic review is to assess the existing scientific literature in the last 42 years (1975 to 2017) to describe the body of research generated for the population of this region and determine future research directions.
METHODS: We selected published research articles using a combination of definite keyword searches in PubMed. Only articles presenting the Caribbean population as the focus of their research objectives were included in this analysis.
RESULTS: Studies on breast cancer in the Caribbean are limited. A majority of publications on Caribbean populations were descriptive, focusing on cancer trends and clinicopathologic factors. High incidence and mortality rates for breast cancer are reported for the region, and there seem to be some differences between countries in the frequency of cases according to age at presentation. A limited number of epidemiologic, behavioral, and genetic and molecular studies were conducted in more recent years.
CONCLUSION: A regional strategy for cancer registration is needed for the Caribbean to address possible underestimates of breast cancer incidence. Furthermore, behavioral, molecular, genetic, and epidemiologic investigations of breast cancer are critical to address the concerns related to currently described high incidence and mortality rates in the Caribbean.

INTRODUCTION

Recent reports reveal that breast cancer (BC) is a leading cause of death resulting from cancer in women in Caribbean nations.[1] On the basis of estimates by the WHO, geographic variations in BC incidence and mortality are reported among Caribbean nations.[2] Population-based studies suggest that BC susceptibility may be associated with aging, reproductive history, and lifestyle factors.[3-7] Since the 1990s, research has demonstrated that beyond the previously identified risk factors, familial clustering of BC points toward genetic risk factors, such as mutations in BRCA1 and BRCA2 genes.[8] The biologic heterogeneity of BC has also been highlighted through molecular profiling,[9-12] with four identified molecular profiles (basal-like, human epidermal growth factor receptor 2 [HER2]/neu [ie, HER2 positive], luminal A, and luminal B), characterized by distinct signatures and clinical outcomes.

As compared with research in other countries, BC research in Caribbean populations is still limited and requires stronger focus and development. Progress in such research would be significant because Caribbean populations are ethnically and culturally diverse (mixed genetic origins of predominantly African descent). Today, a gap exists in research investigating the epidemiology, clinicopathologic features, genetic determinants, and molecular classifications of BC in these populations. The African Caribbean Cancer Consortium seeks to build research capacity in the Caribbean to further investigation of factors associated with BC trends. We examined BC incidence and mortality in the Caribbean. We also analyzed 43 years of population-based studies (1975 to 2017) of BC and associated risk factors in Caribbean countries and discuss future regional research strategies.

METHODS

Literature Review and Inclusion and Exclusion Criteria

This analysis includes relevant publications in PubMed from inception to January 2017 using the search terms “breast cancer and Caribbean” and “breast cancer and [individual country name]” for all Caribbean countries (Appendix). The bibliographies of several review articles were examined to identify additional publications that might have been missed by our PubMed search.

Abstracts and full texts were reviewed by three independent reviewers and cross-referenced to address discrepancies and confirm eligibility for inclusion. For overlapping studies, the publication with the largest population and/or most complete information was included. All population-based analyses (case-cohort, cross-sectional, case-control, qualitative, and other observational studies) of BC in a Caribbean population were considered. Studies comparing Caribbean data with those on other populations in different geographic regions were included, but only data on indigenous Caribbean women are reported. Studies conducted in basic science or immigrant Caribbean populations and non-English articles were excluded. Screening prevalence rates in Caribbean women were reported in our previous review[13] and therefore were excluded from this analysis, with the exception of publications that reported factors associated with screening. On the basis of these inclusion and exclusion criteria, data from 92 publications were included in this analysis (Fig 1).

Fig 1

Flow diagram of study selection criteria.

Data Collection

From each publication, first author, publication year, topic of research focus, study sample size, and reported key findings were abstracted. The number of publications by country was examined, and countries were categorized as high (≥ 10 publications), intermediate (five to nine publications), or low (one to four publications).

For estimates of BC incidence and mortality, GLOBOCAN 2012[2] was used for the Caribbean and other world regions (for comparison). Incidence trends by age were also examined for each Caribbean nation and compared with world incidence trends. All incidence and mortality rates are reported as age standardized to the world population.

RESULTS

Estimated Incidence and Mortality of BC in Caribbean Populations

In all geographic regions, BC incidence was higher than mortality (Fig 2). The lowest BC incidence and mortality rates are estimated in Haiti (22.0 and 11.5 per 100,000, respectively) and the highest in the Bahamas (98.9 and 26.3 per 100,000, respectively). When age-specific incidence was examined, a majority (11 [79%] of 14) of Caribbean countries had peaks in BC incidence that were similar (age, 65 to 69 years) or at an older age (age, 70 to 74 years) than world estimates (peak at age 65 to 69 years). Among these Caribbean countries, Suriname, the Bahamas, and Guadeloupe had a second peak in BC incidence at < 60 years of age (age 45 to 49, 50 to 54, and 55 to 59 years, respectively). Only French Guiana, Dominican Republic, and Belize had a peak in BC incidence at a younger age (age 55 to 59 years).

Fig 2

Age-standardized incidence (bars) and mortality rates (circles) of breast cancer per 100,000 per year for individual Caribbean countries, the world, and US and other Latin American countries. Data adapted.[2] ASR(W), rate age standardized to the world population.

Status of BC Research in Caribbean Populations

From 1975 to 2017, 92 publications on BC research in the Caribbean region were identified, with an increase in frequency from the 1990s (Fig 3). Most articles focused on epidemiologic trends and clinicopathologic features. It was not until 2007 that an expansion of research focus was observed (Fig 3). Sixteen (53%) of 30 Caribbean countries were documented in publications focusing on BC in native Caribbean women, with a majority having intermediate to low numbers of publications (Fig 4).

Fig 3

Distribution of breast cancer publications by research focus over time: 1975 to 2017.

Fig 4

Distribution of studies according to country. NOS, not otherwise specified.

A majority (82 [89.1%] of 92) of publications involved one or more Caribbean coauthors, and many involved coauthors in other geographic regions. Most Caribbean researchers were the first author (66 [80.5%] of 82) on each publication, and a smaller proportion (61 [74.4%] of 82) were the corresponding author.

Key Findings From Caribbean BC Studies

Key findings of BC research in the Caribbean are summarized in Table 1. Sample size in these publications ranged from 16 to 9,389 women.

Table 1

Key Findings From BC Research Studies of Caribbean Populations

Descriptive cancer trends (incidence and mortality).

Eight publications from 1992 to 2016 described regional analyses of epidemiologic trends in incidence and/or mortality.[1,14-20] All of them suggested that BC was both the leading cancer site and cause of death resulting from cancer. Similar findings, using data from pathology databases or hospital or national cancer registries, were also published[21-50] and highlighted: earlier age at diagnosis in Barbados (age 50 to 54 years) in contrast to US black women (age 75 to 79 years)[24]; increasing BC incidence in Jamaica from 1958 to 1992,[34,36] followed by relatively stable incidence rates from 1993 to 2007[37,39,40]; lower BC incidence in Puerto Rico from 1969 to 2003 in comparison with mainland United States[46,47] and higher incidence and mortality rates in areas with high socioeconomic status[43]; similarly, from 1994 to 2003, lower BC incidence in Suriname than in the United States[48]; and increasing BC mortality in Trinidad and Tobago (henceforth, Trinidad) from 1970 to 2003,[49] with geographic residence and African ancestry emerging as strong predictors of BC incidence and mortality.[50]

Clinicopathologic features.

Eighteen publications described clinicopathologic features of women from seven Caribbean countries (the Bahamas, Cuba, Guadeloupe, Jamaica, Puerto Rico, Suriname, and Trinidad).[51-67,105] The most common histologic type was infiltrating ductal carcinoma. For countries that reported staging data, stage ≥ 2 at presentation was reported in the Bahamas,[51] Jamaica,[54,56] and Trinidad.[67] In contrast, one study in Puerto Rico reported that 88.9% of patients presented with stage I to III and T1/T2 BC.[60] Multiple studies of hormone receptor status (HRS) were conducted in Puerto Rico (four of five studies), whereas only a single study each was conducted in Cuba, Guadeloupe, Jamaica, and Trinidad. No studies of HRS were reported in the Bahamas or Suriname. The frequencies of estrogen receptor (ER) –positive and/or (progesterone receptor (PR) –positive BC were similar for patients in Puerto Rico (ER positive, 65.9%; PR positive, 51.8%), Guadeloupe (ER positive/PR positive, 65.1%), and Jamaica (ER positive, 63%) and slightly higher than the frequencies for patients in Cuba (ER positive, 53%; PR positive, 49%) and Trinidad (ER positive, 54%; PR positive, 46%).[52,53,57,59,66] The prevalence of triple-negative BC was reported only for patients with BC in Guadeloupe (14%) and Puerto Rico (17.3%).[53,61]

Case-control[69,70,73-75] and cohort analyses[71,72] were reported in six Caribbean countries, and common risk factors were identified: older age, late menarche, late parity, null parity, body size, family history, low DNA repair capacity, and low education level. However, the study in Jamaica noted that 54% of patients with BC did not have these common risk factors.[72] Moreover, a higher percentage of cases were diagnosed in women between the ages of 45 and 54 years in Guadeloupe and Martinique.[71]

BC outcomes and predictive factors.

Factors associated with BC outcomes were reported in four Caribbean countries (Cuba, French Guiana, Jamaica, and Trinidad). In accordance with the literature, negative HRS, advanced stage at diagnosis, positive lymph node status, and histologic type (infiltrating ductal carcinoma) were associated with poor prognosis or survival.[76-82] Survival rates for patients with BC in Trinidad and Guyana were lower as compared with Caribbean-born patients with BC living in the United States (Brooklyn, NY).[78] In Jamaica, one study reported that clinical staging was inadequate for the management of stage I BC and highlighted a need for consistent pathologic staging as well as use of other prognostic predictors to improve clinical outcomes in BC.[82]

Behavioral risk factors.

Seven Caribbean countries contributed to six publications on behavioral factors relevant to BC.[83-88] Screening was associated with younger age and church attendance in Grenada[84] and general practitioner visits and provider communication in Guadeloupe and Martinique and the US Virgin Islands.[85,88] In Grenada, perceived susceptibility to BC and perceived benefit of screening were more likely to be associated with women who frequently attended church.[84] The only intervention study was conducted in Jamaica, where a theory-based educational intervention positively influenced knowledge of BC risk factors, symptoms, and types of screening and increased screening rates in screening-naïve women.[86] The two behavioral studies involving BC cases were conducted in Cuba and Puerto Rico; in Cuba, family history of BC was perceived as a risk factor, but the concept of genetic risk had little meaning[83]; in Puerto Rico, patients with BC had low physical activity, despite access to exercise equipment and facilities.[87]

Barriers to early detection and BC treatment.

Seven Caribbean studies yielded eight publications related to barriers to early detection and treatment.[89-96] Lack of knowledge of signs and symptoms was highlighted in Haiti,[91] and low frequency of breast self-examination, infrequent clinical breast examinations, and limited access to mammography were reported in Trinidad.[95] Health care cost was seen as a barrier to treatment in Haiti[92] and a barrier to screening in Trinidad.[95] In Trinidad, knowledge of family history as a risk factor for BC was common (76.8%), but knowledge of other risk factors was limited.[96] In contrast, in Barbados and Grenada, university students identified multiple BC risk factors, such as genetics, stress, smoking, and diet.[89] Similarly, among BC survivors in Guadeloupe and Martinique, stress, genetic causes, and poor diet were the most frequently cited risk factors.[90] In Puerto Rico, two studies respectively reported that underserved women diagnosed with BC were treated comparably to other patients[93] and that referral by a physician was an important factor related to mammogram compliance.[94]

Genetic risk factors and molecular biomarkers.

Examinations of genetic and molecular biomarkers among Caribbean women were limited in comparison with US and European populations.[97-103] BRCA1/2 mutations were the only genetic risk factors evaluated, with the highest prevalence reported in the Bahamas (approximately 27%).[97,99] A Bahamian study showed that genetic testing of unaffected family members of patients with BC with BRCA1/2 mutations might be more effective if investigators use the proband as an intermediary to communicate with them.[98] In addition, the only molecular biomarker investigated in outcomes of patients with BC was reported from Cuba, where a strong expression of interleukin-10 in tumors was associated with tumor markers of poor prognosis.[101]

Cost analysis.

The only Caribbean study on cost was a cost analysis of labor market productivity loss in Puerto Rico as a result of premature mortality from cancer. The study showed that death resulting from BC was included among cancer-related mortality causes that contributed the most to productivity loss.[104]

DISCUSSION

Our literature review includes 92 publications of BC burden and research in native Caribbean women from 35 Caribbean nations over a span of 43 years and underlines the lack of precise field data as well as the limited research scope. The countries with the most research publications were countries with long-standing cancer registries and local scientific teams.

Most publications focused on describing cancer trends and clinicopathologic features (Fig 3), which were in line with GLOBOCAN estimates,[2] placing BC as one of the top cancer sites for incidence and mortality among Caribbean women. However, several factors contribute to the weaknesses of these publications, such as the limited number of cancer registries with high-quality data and the rare focus on the impact of social and environmental factors on epidemiologic trends.[2,106]

Moreover, availability of cancer screening programs and access to these programs play key roles in BC trends. Access to cancer screening programs varies not only according to an individual’s geographic residence but also among countries. A recent review of Caribbean cancer screening programs revealed that for 12 countries that have BC screening services available, mammography is not available in the public sector for more than half of them.[107] It is then plausible that Caribbean women with private insurance and those with higher socioeconomic status may have better access to mammography. Furthermore, although traveling abroad to other countries for BC screening is unlikely, for those who can afford it, traveling abroad to other Caribbean countries (eg, Trinidad, Martinique, or Puerto Rico), the United States, the United Kingdom, or France for confirmation of a BC diagnosis or for treatment may be an option. All these factors may influence epidemiologic trends in BC. Therefore, BC incidence reported in Caribbean publications may have been underreported and subject to ascertainment bias caused by variability in data quality, completeness, socioeconomic status, BC screening access, and guidelines.

In terms of clinicopathologic status, few Caribbean studies described late stage at diagnosis, and a majority reported on ER-positive and/or PR-positive BC, with triple-negative BC averaging approximately 15% of all cases.[53,61] As previously mentioned, these reports of cinicopathologic characteristics may have been influenced by limited access to and uptake of BC screening in the region. Furthermore, the availability of or access to HRS testing in each country may also have had an impact. Immunohistochemistry (IHC) staining for BC receptors is part of the standard care in a few Caribbean countries. For others, IHC is not performed routinely except in the private sector, and for patients in Martinique and Guadeloupe, IHC is performed primarily in mainland France. Resolution of indeterminate HER2 results usually requires additional testing via in situ hybridization, and in these instances, samples are typically sent to the United States for testing. Therefore, some Caribbean countries may not experience the prognostic benefit of hormone receptor profiling, and although Caribbean publications report differences in the prevalence of ER and PR status among countries, these data may not necessarily reflect true biologic differences among Caribbean populations. These limitations should also be considered when reviewing published data from the Caribbean.

Although HRS and more advanced cancer stages are linked to poorer survival, as described by the few Caribbean studies on the subject,[53,61,77] there are still multiple issues pertaining to cancer outcome and survival that have yet to be fully examined. Survival studies have almost exclusively been led in Trinidad, and virtually no information exists for other countries. Furthermore, apart from the Dominguez et al[30] study describing disability-adjusted life-years in Cuban women, quality of life of Caribbean BC survivors after diagnosis, during treatment, or after returning to work has rarely been explored. Usual treatment regimens have not been described in detail, nor has access to such regimens. Behavioral factors, which play a major role in compliance with good health practices and screening, are also not well documented, nor is the impact of socioeconomic disparities within and among nations on health outcomes.

Age-specific incidence presented in this review further suggests that research studies are needed in the Caribbean to examine the contribution of hereditary versus sporadic BC. In the limited studies on molecular and genetic factors, mutations in BRCA1/2 genes were shown to influence early BC onset.[97,103] However, there is an entire panel of highly penetrant (PALB2, TP53, CDH1, and STK11) and moderately penetrant germline (CHEK2, BRIP1, RAD51, and ATM) pathogenic variants and other genetic markers that are currently being examined in genome-wide association studies in populations other than those of the Caribbean.[108-110] A comprehensive approach toward gene mutation and expression profiling in native Caribbean women must be adopted to characterize specific differences and improve cancer prognosis and outcome.

Key findings from Caribbean publications confirm similarities between the region and other countries, such as the United States. However, these publications also suggest within-region differences, and we have previously described plausible explanations such as differences in ethnic mixes, health care systems and economies, cancer registration, screening use.[110-112] Our review also demonstrates that too little research has been conducted in the Caribbean to provide certified evidence of real or artifactual differences. Our review has highlighted significant research gaps in behavioral, molecular, genetic, and epidemiologic investigations of BC in native Caribbean women.

These disparities among nations, which might be linked to the geographic heterogeneity in BC incidence and mortality in the Caribbean, call for specifically tailored research and cancer control interventions that can only be addressed if the regional research portfolio is expanded to include more large-scale, targeted epidemiologic investigations with diverse research scopes. However, a more diverse research focus also requires adequate expertise and resources (human and infrastructural), which are not always readily available. Only those countries with strong local research teams (university campuses), strong public health policies in terms of cancer control and prevention, and reliable cancer registration and/or screening infrastructures seem able to publish or implement more diverse research studies. This is strongly supported by data shown in Figure 4. In those countries where research gaps exist or that experience challenges in diversifying their research scopes, improved cancer registration, capacity building, resource optimization and sustainability, and collaboration and communication among investigators will be essential in increasing research output and findings.

Today, a regional common unified strategy is necessary to address research needs among nations. Additional solutions lie in developing a strategic plan that includes setting standards for reinforced, sustainable, diverse, and efficient population research. This implies identifying intra- and intercountry research specialties, needs, and priorities as well as existing tools, resources, stakeholders, and collaborators.

To that end, research consortia, such as the African Caribbean Cancer Consortium,[113] represent a rich network of collaborators (existing and aspiring cancer registries, research and medical teams, governmental and institutional policymakers, and other cancer advocates and research teams at the local, regional, and international levels) and can leverage transdisciplinary expertise as a solution in addressing identified research gaps and promoting more diverse research opportunities in the Caribbean. We will promote research enhancement activities and training and establish research resources (eg, cancer and control cohorts with linked epidemiologic, clinical, and molecular and biomarker data). Regional investigations of BC risk and outcomes as well as tailored strategies for cancer prevention interventions in Caribbean women can be achieved.