High prevalence of BRCA1 and BRCA2 mutations in unselected Nigerian breast cancer patients†
Conflicts of interest: None of the authors of this manuscript has any financial or personal relationship with any outside entity that could inappropriately influence this work.
Inherited mutations in the BRCA1 and BRCA2 genes are the strongest genetic predictors of breast cancer and are the primary causes of familial breast/ovarian cancer syndrome. The frequency, spectrum and penetrance of mutant BRCA1/BRCA2 alleles have been determined for several populations, but little information is available for populations of African ancestry, who suffer a disproportionate burden of early onset breast cancer. We have performed complete sequence analysis of all BRCA1 and BRCA2 exons and intron–exon boundaries for 434 Nigerian breast cancer patients from the University College Hospital in Ibadan, Nigeria. In contrast to previous suggestions that BRCA1/BRCA2 mutation frequencies are low or undetectable in African American populations, we find that Nigerian breast cancer patients have an exceptionally high frequency of BRCA1 and BRCA2 mutations (7.1 and 3.9%, respectively). Sixteen different BRCA1 mutations were detected, seven of which have never been reported previously, while thirteen different BRCA2 mutations were seen, six of which were previously unreported. Thus, our data support enrichment for genetic risk factors in this relatively young cohort. To improve breast cancer outcomes, we suggest that family-based models of risk assessment and genetic counseling coupled with interventions to reduce breast cancer risk should be broadly disseminated in Nigeria and other underserved and understudied populations.
Breast cancer patients of African ancestry have increased likelihood of early onset of disease and high frequency of hormone receptor negative tumors. While overall breast cancer incidence is lower among populations of African ancestry than other groups, mortality rates are higher. These high mortality rates are observed in both African and African American populations for reasons that are poorly understood and yet remain understudied.1–5
BRCA1 (MIM 113705) and BRCA2 (MIM 600185) are the two most commonly mutated tumor suppressor genes associated with early onset and familial forms of breast cancer.6–9 Tumors associated with germline BRCA1 mutations are more likely to be ER negative, PR negative, HER2/neu negative (triple-negative) and further classified as basal-like based on expression of EGFR or Cytokeratin 5/6. These tumors have no defined therapeutic target and can be aggressive when refractory to current combination chemotherapy regimes. In contrast, tumors in BRCA2 mutation carriers are usually ER positive and amenable to treatment with hormonal therapies.10–12 Given the clinical implications of these mutations and the promise of prevention in further reducing morbidity and mortality associated with breast cancer, the American Society of Clinical Oncology has promoted integration of genetic testing as part of a comprehensive risk assessment and prevention plan.
To date, studies of population-specific risks attributable to inherited BRCA1/2 mutations have focused on populations with strong founder effects (reviewed in Fackenthal and Olopade13). This approach allows assessment of mutation frequencies, spectra and penetrance estimates with minimal “noise” from the genetic diversity typical of nonfounder populations. However, results from such studies may not necessarily apply to populations with different genetic histories, especially ethnically mixed populations typical of urban clinical settings. Several studies have reported low or undetectable frequencies of BRCA1/2 mutations in African American cohorts.14–21 One study showed a 1.3% prevalence of BRCA1 mutations in African American breast cancer patients, but 16.7% in African American patients selected for extremely early age of onset (<35 years old).22 To determine whether these results reflect the frequency and spectrum of BRCA1/2 mutations in an African population without non-African genetic admixture, we have performed complete BRCA1/2 sequence analysis of 434 breast cancer cases from University College Hospital in Ibadan, Nigeria. Our goal was to evaluate the burden of disease attributable to inherited mutations in BRCA1 and BRCA2 in a tertiary hospital that can inform policy regarding breast cancer screening in resource poor African countries. To our knowledge, this is the first comprehensive analysis of BRCA1 and BRCA2 mutations in an African country and underscores the need for resource poor countries to “leap frog” by investing in genome-based technologies to spur development while reducing health inequities.
Material and Methods
The study setting and design have been described in detail elsewhere.23 Briefly, breast cancer cases were identified through the Surgical Oncology and Radiotherapy units of the University College Hospital (UCH), Ibadan, Nigeria. This hospital directly serves a population of more than 3 million people and additional patients are referred from other hospitals in the region. Most of the breast cancer patients diagnosed in this region would be seen at UCH. All consecutive female breast cancer patients aged 18 and above, with a clinical diagnosis of invasive breast cancers between March 1998 and July 2006, were eligible and invited to participate in the study. Additionally, eight male breast cancer cases were included. Trained nurse interviewers obtained informed consent, administered a structured questionnaire and obtained blood samples for genotyping. Nearly 96% of eligible patients approached agreed to participate in the study. Ethnicity was self-reported.
The Institutional Review Boards of the University of Ibadan and the University of Chicago approved this study.
Genomic DNA was prepared from peripheral blood from breast cancer patients using Puregene™ DNA isolation kit (Qiagen, Germantown, Maryland, USA) according to the manufacturer's instructions.24
PCR fragments were sequenced directly using a fluorescent dye dideoxy chain terminator (BigDye 3.1) reaction and products were analyzed on an Applied Biosystems 3730XL 96-capillary DNA sequencer (both from Applied Biosystems, Foster City, CA). Primer sequences and PCR conditions are available upon request. PCR products included all sequence information from coding exons, plus flanking intronic sequence extending beyond intron/exon boundaries. In some cases, exonic sequences were too long to be covered by single PCR products, so several overlapping fragments were used. As a result, we sequenced 33 and 45 PCR-amplified genomic DNA fragments representing the BRCA1 and BRCA2 genes, respectively.
In this study, the analyses focused only on mutations that are presumed to be disease-associated. We calculated the mutation prevalence and exact 95% confidence interval using a Binomial distribution. Fisher's exact test was used to determine whether mutation prevalence varied by age, ethnicity and family history.
In total, 434 consecutively enrolled breast cancer patients were included in this analysis. They were not selected for age of onset, family history or any other criterion that would enrich for BRCA1/2 mutation carriers. Demographic characteristics, family history information and other known risk factors for these patients are shown in Table 1. The mean age at diagnosis was 46.5 ± 11.5 years old and 10% of patients reported a positive family history of breast cancer. As shown in Table 1, the patient population has a low profile of known breast cancer risk factors, including late age at menarche, high parity, long duration of lactation and low prevalence of hormone contraceptive use.
Table 1. Demographic characteristics and selected risk factors of 434 breast cancer patients in Nigeria
Complete sequence analysis for all 434 patients revealed 31 BRCA1 mutation carriers, giving a prevalence rate of 7.1% (95% CI: 4.9–10.0%) (Table 2). These mutations included nonsense, missense, frame-shifting and splice-site mutations (Table 3). Note, in Tables 3–5, allele names use Breast Cancer Information Core (BIC) nomenclature (http://research.nhgri.nih.gov/bic/). We also include a single large-scale rearrangement that deletes BRCA1 exon 21 that we reported previously.25 Additionally, we identified 17 BRCA2 mutation carriers (3.9%, 95% CI: 2.3–6.2%). These included one nonsense mutation and six frame-shifting mutations (Tables 2 and 3). All mutation carriers are female. Table 2 presents mutation prevalence according to demographic factors and family history of breast cancer. BRCA1 mutation carriers had a mean age of 43.5 ± 10.7 years, BRCA2 mutations carriers had a mean age of 45.6 ± 9.8 years and noncarriers had a mean age of 46.8 ± 11.7 years. There was a higher proportion of BRCA1 carriers in the “under 50 years” age category than in the “over 50 years old” category (p = 0.02). Patients with a family history of breast cancer were two-fold more likely than patients without a family history to carry a mutation in BRCA1 or BRCA2 (RR = 2.04, 95% CI = 1.06–3.93). Notably, 7 out of the 28 patients (25%) with positive family history and age of onset before 50 years were BRCA1 mutation carriers. Still, as with other populations, the majority of mutation carriers reported no family history, and the majority of patients with family history carried no mutations. Mutation prevalence did not vary by ethnicity in this cohort.
Table 2. Prevalence of BRCA1 and BRCA2 mutation carriers in Nigerian breast cancer patients
Table 3. Deleterious BRCA1 and BRCA2 mutations in the Nigerian cohort
The overall BRCA1/2 mutation rate in our cohort is 11% (95% CI: 8.1–13.9%), which we believe is larger than the reported rates for any breast cancer cohort from a nonfounder population unselected for family history or age of onset.26, 27 However, this rate is still lower than those reported for unselected cohorts from founder populations, including the Bahamas (23% BRCA1 mutation carriers)28 and nonfounder populations selected for early-onset disease, such as the United States (17% BRCA1 or BRCA2 mutation carriers).16
Table 3 lists the deleterious mutations of BRCA1/2. Of the 29 distinct BRCA1/2 mutant alleles identified, 11 were recurring mutations, occurring in two or more subjects (eight in BRCA1 and three in BRCA2). Thus, 62.5% of all mutation carriers in our cohort carried a recurrent mutation.
In addition to deleterious mutations, we observed 46 variants of uncertain significance (VUS) in BRCA1 (Table 4) and 96 VUS in BRCA2 (Table 5). These frequencies are consistent with the relative sizes of the BRCA1 and BRCA2 cDNA sequences (5.7 and 11 kb, respectively). Note the frequency of highly recurrent (≥10%) VUS is higher in BRCA1 (9/46, 20%) than BRCA2 (9/96, 9%).
Table 4. Variants of uncertain significance (VUS) in BRCA1
Table 5. Variants of uncertain significance (VUS) in BRCA2
In this study, we have determined the prevalence of BRCA1/2 mutations in a cohort of Nigerian (predominantly Yoruba) breast cancer patients unselected for age of onset or family history. Surprisingly, 31/434 (7.1%) patients carried BRCA1 mutations and 17/434 (3.9%) individuals were BRCA2 mutation carriers. Other reports show consistently lower frequencies of BRCA1 mutations in nonfounder populations, and especially low frequencies in African Americans. For example, Malone et al. examined 1,628 population-based breast cancer cases and showed that BRCA1 mutations occurred in 2.9% of White cases and only 1.4% of Black cases.17 For BRCA2, a 3.9% frequency is slightly higher than the 2.6% frequency seen among African Americans by Malone et al.17 Thus, the 11% prevalence observed in this previously understudied population underscores the need for further research into the true population prevalence of BRCA mutations in diverse cancer populations of African ancestry.
The difference between mutation frequencies in African Americans and our Nigerian cohort could be because of the younger age of the cohort and overall population structure, or selection bias if cases with the worse severity of disease were the only ones who made it to the hospital given the resource constraints in Nigeria. Differences in mutation frequencies between reports could also be due in part to mutation detection methodologies. We believe the complete sequencing reported here, along with our MLPA analysis reported previously29 are the most accurate methods currently available for mutation detection. However, we note that local sequence environments placed constraints on primer design in two regions of BRCA2 such that full-length sequence reads were difficult and underrepresented in the final tally of genotypes. Specifically: (i) the first 203 base pairs from the 5′ end of BRCA2 exon 10 (and the bases immediately flanking the 5′ exon 10/intron boundary) were not determined in the majority of samples sequenced. We note that nine different deleterious mutations in this interval have been reported to the Breast Cancer Information Core (BIC) for 19 different individuals30; (ii) BRCA2 exon 11 cDNA sequence between positions 5,207 and 5,256 were not determined in the majority of samples sequenced. As before, we note that BIC lists a single report of a deleterious mutation in this region. Because of the extreme rarity of deleterious mutations falling within these problematic sequence regions, we expect they are unlikely to contribute significantly to our final tally of deleterious mutations. Additionally, the average sequence failure rates for the BRCA1/2 amplicons were 4.6% per subject for BRCA1 amplicons and 5.9% for BRCA2 amplicons. If a subject had an unknown mutation status at a specific allele, we tallied it as negative for that variant. This conservative handling of missing data would slightly underestimate mutation prevalence by about 5%.
Further adjustments to our population-specific mutation frequencies must account for variants of uncertain clinical significance that could potentially affect BRCA1 or BRCA2 gene function.31 We have also shown our Nigerian cohort has a high frequency of polymorphisms and unclassified variants, consistent with previous observations19 (Tables 4 and 5). These variants include one example of BRCA1 R1699Q and two examples of A1708V, both of which have been suggested as potential deleterious mutations.32 Additionally, our cohort contains three examples of BRCA1 Ser616del and one example of BRCA1 Ala1175del, either of which conceivably could compromise gene function. Thus, our assessment of the BRCA1/2 mutation frequencies in our cohort may be an underestimate. BRCA1 M1783T has also been suggested as a potential deleterious mutation according to the International Agency for Research on Cancer (IARC) (brca.iarc.fr/LOVD). This same organization has also defined several unclassified variants of BRCA1 and BRCA2 as “not pathogenic or of no clinical significance,” as indicated in Tables 4 and 5. Additionally, cooccurrence of an unclassified variant with a deleterious mutation, especially multiple deleterious alleles, can be used as evidence against the clinical significance of the VUS. Several such variants are indicated in Tables 4 and 5, suggesting they are unlikely to contribute to disease risk on their own.
To further assess whether the spectrum of mutations identified in the Nigerian cohort is similar to those found in other populations of African Ancestry, we checked the mutant alleles identified in this study for reports in the Breast Cancer Information Core (BIC) http://research.nhgri.nih.gov/bic/ (Table 3). On September 1, 2010, we found 13/29 (45%) deleterious BRCA1/2 alleles in our Nigerian cohort had not been reported to BIC. We therefore suggest these mutations originated in Africa prior to the African slave trade diaspora, and that some were carried along in the diaspora while others were not represented.
For populations in resource poor settings, there is paucity of research data to guide evidence based cancer control policies. Women of African ancestry have the highest mortality rates for early onset aggressive breast cancer for reasons that are largely unknown and understudied. There exist breast cancer phenotypic similarities between Africans and African Americans that distinguish these groups from other groups. For example, both African and African American breast cancer cases have higher frequencies of triple-negative/basal-like (ER-, PR-, HER2/neu negative) and young onset breast cancers than Whites or other groups.33–38
This study shows that deleterious mutations in BRCA1 and BRCA2 at least in part contribute to population differences in age-specific incidence as well as differential distribution of molecular subtypes of breast cancer. Beyond the high mutation frequency in this unselected cohort, it has been suggested that methylation of the CpG islands in the BRCA1 promoter is correlated with reduced BRCA1 transcription39 and protein levels,40 and other studies show that basal-like tumors show reduced BRCA1 mRNA levels regardless of promoter methylation status.41 This underscores the need for a comprehensive analysis of genes and genomic pathways that are dysregulated leading to the aggressive behavior of breast tumors in women of African ancestry. This study illustrates the opportunity to end global health disparities using rigorous science and technology, which is only now being addressed by the newly launched African Genome Project, a product of the Human Heredity and Health in Africa (H3 Africa) project. We believe our study is the first comprehensive analysis of two major genes that increase susceptibility to breast and ovarian cancer and have significant clinical implications for African populations as envisioned by these initiatives. To eliminate inequities, disseminating information and providing access to genetic services that reduce the burden of chronic noncommunicable diseases like breast cancer in resource poor settings should be elevated to national health priority agenda. Only then can the promise of evidence-based interventions to reduce the disproportionate burden of death from early onset breast cancer in women in the African diaspora be realized.