Fax: (215) 955-2340
Differences in breast carcinoma characteristics in newly diagnosed African–American and Caucasian patients
A single-institution compilation compared with the National Cancer Institute's Surveillance, Epidemiology, and end results database
Article first published online: 9 JUL 2007
Copyright © 2007 American Cancer Society
Volume 110, Issue 4, pages 876–884, 15 August 2007
How to Cite
Morris, G. J., Naidu, S., Topham, A. K., Guiles, F., Xu, Y., McCue, P., Schwartz, G. F., Park, P. K., Rosenberg, A. L., Brill, K. and Mitchell, E. P. (2007), Differences in breast carcinoma characteristics in newly diagnosed African–American and Caucasian patients. Cancer, 110: 876–884. doi: 10.1002/cncr.22836
- Issue published online: 2 AUG 2007
- Article first published online: 9 JUL 2007
- Manuscript Accepted: 16 APR 2007
- Manuscript Revised: 22 MAR 2007
- Manuscript Received: 24 JAN 2007
- breast cancer;
- African American;
- basal phenotype;
Breast carcinomas in African–American patients appear to be more aggressive than in Caucasian patients due to multifactorial differences.
The authors compiled pathology data from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) database regarding stage, histologic grade, and estrogen receptor (ER) expression in breast carcinomas diagnosed in 197,274 African–American and Caucasian patients between 1990 and 2000, and the same information, along with nuclear grade, Ki-67, c-erb-B2, and p53 expression, in 2230 African–American and Caucasian patients diagnosed at Thomas Jefferson University Hospital between 1995 and 2002. Immunohistochemical markers were assayed in paraffin-embedded, formalin-fixed tissue stained with hematoxylin and eosin using antibodies to these proteins, with differences in expression analyzed by the chisquare test.
In both databases, more African–American patients presented with advanced stage tumors and higher histologic (P < .001) and nuclear grade (P < .001) than Caucasian patients. African–American patients had less ER positivity (51.9% vs 63.1%; P < .001) but significantly higher Ki-67 (42.4% vs 28.7%; P < .001) and p53 expression (19.4% vs 13.1%; P < .05) than Caucasian patients with all stages of disease. In addition, the basal or “triple-negative” breast cancer phenotype was more common in African–American patients than in Caucasian patients (20.8% vs 10.4%; P < .0001), and was associated with higher histologic and nuclear grade (P < .0001).
African–American patients with breast carcinomas are more likely than Caucasian patients to present with tumors that are of a later stage and higher grade, with higher Ki-67 expression and more ER negativity, thereby highlighting a greater need for early screening among African–American women. Molecular studies that may explain these differences, and correlations with survival, have been proposed to identify therapeutic targets. Cancer 2007. © 2007 American Cancer Society.
Invasive breast carcinoma is the most commonly diagnosed cancer in women in all racial groups in the U.S. Breast carcinoma is the second leading cause of cancer mortality in women after lung cancer, but is the leading cause of cancer mortality in African–American women.1 Over the last several decades, a paradoxical correlation has been observed among African–American women in terms of higher breast cancer mortality rates despite a lower incidence of breast cancer diagnoses compared with Caucasian women.2 Although the age-adjusted incidence rates for breast carcinoma are lower in African–American women compared with Caucasian women (95.4 per 100,000 vs 111.8 per 100,000), age-adjusted breast cancer mortality rates are poorer in African–American women (31.4 per 100,000 vs 27.0 per 100,000),3 with the risk of death from breast cancer being 67% higher for African–American women compared with that for Caucasian women.4 In addition, a greater proportion of breast cancers are diagnosed in African–American women of younger age groups compared with Caucasian women (33% of African–American women age <50 years vs 25% of Caucasian women).4 Invasive breast carcinomas diagnosed in African–American patients are described as more aggressive compared with those in Caucasian patients. Several studies entered into the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) database have confirmed that not only are African–American women more likely to be diagnosed at later stages of the disease,5 most notably stages IIIB and higher,4, 6–8 but they are also more likely to present with higher grade tumors.6 African–American women are more likely to have more aggressive breast cancer pathologies such as inflammatory, medullary, or papillary carcinomas, and are less likely to have lobular or tubular carcinomas than Caucasian women.4 Furthermore, African–American women are less likely to have positive estrogen receptors (ERs) and progesterone receptors (PRs) than Caucasian women.3, 9, 10 Multivariate Cox proportional hazards analyses have shown that race remains a significant independent risk factor for mortality, with African–American race reported to be a prognostic indicator of worse outcome in breast cancer.11
Racial differences in breast carcinoma mortality rates are believed to likely be multifactorial in nature.4, 12 Multiple variables can affect the stage at and the timing of diagnosis, including lower utilization of mammography among African–American women (thereby delaying the diagnosis), cultural beliefs, and socioeconomic factors.13–18 Other variables noted among African–American women that are believed to be associated with a poorer prognosis and higher stage tumors may include diet, increased body mass index and obesity, increased parity, an earlier age at first full-term pregnancy,19 and a lower incidence of breastfeeding,20 all of which may affect the duration of circulating estrogen levels. Treatment course also may have an effect on outcome in African–American women with breast cancer. A recent study has demonstrated an excess of psychosocial barriers leading to frequent modifications of adjuvant chemotherapy administration in African–American patients.21 Another study demonstrated that African–American patients were more likely to receive fewer cycles of treatment than expected. Of those patients who terminated treatment early, survival was poorer; of those whose treatment was delayed (due to low white blood cell counts) but completed, survival was not found to be reduced.22 More delays in the initiation of radiation therapy in African–American women also may be associated with poorer survival.23 Minority women, including African–Americans, may be less likely to receive necessary adjuvant treatments (radiation therapy, chemotherapy, and/or hormonal therapy) due to differences in comorbities and a lack of insurance despite equivalent rates of oncologic consultation when compared with Caucasian patients.24, 25 Despite these explanations, a study stemming from survival data from the Cancer And Leukemia Group B (CALGB) trial 8541, in which patients with stage II breast cancer received chemotherapy with a combination of cyclophosphamide, doxorubicin, and 5-fluorouracil (CAF), found that African–American women were more likely to have higher rate of disease recurrence due to other prognostic features such as tumor size and receptor status on univariate analysis only; after multivariate analysis for dose of chemotherapy, the number of lymph nodes, hormone receptor status, and age, race was not found to be a significant prognostic indicator.26 However, an analysis of CALGB trials 9342 and 9840 in patients with metastatic breast cancer demonstrated that African–American patients still have shorter survival despite achieving a similar response from chemotherapy.27 These data highlight the importance of studying the pathologic characteristics of breast cancers in African–American women in addition to demographic differences.
Intrinsic genotypic variables associated with more aggressive breast carcinomas have been investigated in both African–American and Caucasian women. Founder mutations may exist in tumors in both African–American and native African women.2, 28 In addition, numerous BRCA mutations have been reported among African–American breast cancer patients29, 30; however, knowledge and awareness of genetic testing for breast cancer is reported to be lower among African–Americans.31, 32 As mentioned, it has been found that in African–American women, these carcinomas are more likely to be negative for ER expression2, 19, 33, 34; have overexpression of cell-cycle regulators cyclin E, p16, and p53; have a lower expression of cyclin D134; have alterations in p5335, 36; have alterations in c-met (stem-cell factor/hepatocyte-growth factor receptor)35; have polymorphisms in the UGT1A1 gene37; and have polymorphisms in cytochrome P-450-1A1, with heightened risk if coupled with cigarette smoking.38 With the exception of the analysis of the SEER database (135,424 cases in 2000) by Joslyn and West,4 many of the above-mentioned studies have been performed in smaller study populations. Therefore, access to larger databases is important in confirming these disparities.
Genomic analyses of breast cancer phenotypes have subclassified breast cancers into 4 important categories, each with different clinical behavior. These include the luminal A (ER-positive and/or PR-positive, HER-2-negative), luminal B (ER-positive and/or PR-positive, HER-2-positive), HER-2-positive (ER-negative and PR-negative), and basal-like (ER-negative, PR-negative, HER-2-negative, CK5/6-positive, and/or HER-1-positive) phenotypes, and, in this order, have increasingly aggressive behavior and worse prognosis.39, 40 These molecular breast cancer subtypes may respond differently to chemotherapy,41 with the basal-like or c-erb-B2/HER-2-positive phenotype having a greater sensitivity to regimens in the preoperative setting that contain paclitaxel and doxorubicin. African–American women are more likely to have hormone receptor-negative breast cancers than their Caucasian counterparts.39, 40, 42, 43 In addition, Carey et al, from the Carolina Breast Cancer Study registry, reported that African–American women, particularly those who are premenopausal, have a higher likelihood of developing the basal-like (“triple-negative”) phenotype of breast cancer than Caucasian women.40 Consistent with previous data, these tumors were also found to be associated with more p53 mutations, and a higher mitotic index, nuclear pleomorphism, and tumor grade.40 These explanations may contribute to the disparity in prognosis observed between African–American and Caucasian women, and provide hypothesis-generating information toward the development of therapeutic interventional strategies.
We hypothesized that breast carcinomas diagnosed in African–American women in a large population of women presenting with breast carcinomas at a single institution in Philadelphia had more aggressive tumor characteristics than those diagnosed in Caucasian women, which is reflective of data compiled in the national SEER database. In addition, we hypothesized that the basal-like (“triple-negative”) phenotype is more prevalent in the African–American population compared with Caucasian breast cancer patients, and that additional disparities in the expression of cell cycle-related proteins, including p53, may explain differences in aggressive characteristics as well.
MATERIALS AND METHODS
As part of an Institutional Review Board -approved retrospective registry study with waiver of consent, we compiled information from the Thomas Jefferson University Hospital (TJUH) breast carcinoma registry regarding tumor stage; histologic tumor grade; and ER and PR positivity; as well as nuclear grade; Ki-67 proliferation index; and expression of c-erb-B2, p53, p21, and bcl-2 in 2230 African–American and Caucasian patients diagnosed between 1995 and 2002. As a reference comparison [Classification of Malignant Tumors (TNM) of the American Joint Committee on Cancer (AJCC)], we compiled available data regarding tumor stage, histologic tumor grade, and hormone receptor status from the National Cancer Institute's SEER database of breast carcinomas diagnosed in 197,274 African–American and Caucasian patients between 1990 and 2000. In TJUH specimens, paraffin-embedded, formalin-fixed tissue was stained with hematoxylin and eosin. Immunohistochemical markers were assayed using antibodies to the above-mentioned proteins. Differences in expression were analyzed using the chi-square test. Immunohistochemical markers were assayed for according to a previously published method.44 Formaldehyde-fixed and paraffin-embedded tissues were used. Representative sections of the tumors were stained with hematoxylin and eosin, periodic acid–Schiff stain, and trichrome stain, and were subsequently stained with antibodies to ER-A (Dako Corporation, Carpinteria, Calif.), PR (Dako), and c-erb-B2 (Dako). In earlier submitted specimens (up to 2001), tumor sections were also stained with antibodies to Ki-67 (Dako), p53 (Dako), p21 (Dako), and bcl-2 (Dako), with results expressed as percentages per 100 cells counted. A positive result was considered for >20% expression for Ki-67, p52, and p21 staining; c-erb-B2 and bcl-2 staining was expressed as 1+ (weak membranous staining), 2+ (moderate membranous staining), or 3+ (strong membranous staining). Statistical analyses were performed using the chi-square test as per a previously published method.45
Distribution by stage in patients in the SEER database as well as those diagnosed and treated at TJUH between 1995 and 2002 is shown in Table 1. We found that more African–American patients presented at the time of diagnosis with advanced stage tumors (stages III/IV) in both databases compared with those with early-stage tumors (stages I/II) (P < .01). In addition, Table 1 shows comparisons of stage distribution and ER positivity among patients in both databases in which African–American patients with early-stage tumors had a significantly lower ER positivity in the TJUH database compared with Caucasian patients (60.8% vs 66.8% in stage I and 44.4% vs 60.2% in stage II; P < .001), and demonstrated a trend toward lower ER positivity in African–American patients with advanced stage tumors (53.6% vs 52.1% in stage III and 33.3% vs 56.4% in stage IV, compared with Caucasian patients with advanced stage tumors; P = .8257). PR positivity, however, was not found to be significantly different by stage between African–American and Caucasian patients (P = .1858 for early stage disease and P = .5474 for advanced stage disease; data not shown). Tables 2 and 3 tabulate comparisons of stage distribution with histologic grade in both databases and with nuclear grade in the TJUH database. We found that African–American patients were more likely to present with a higher histologic grade in the SEER database (P < .001). This was also confirmed as a statistically significant finding in African–American patients with early-stage tumors in the TJUH database (P = .001), with a trend toward more African–American patients with advanced tumors and a higher histologic grade (P = .2069) (Table 2). In the TJUH database, we found that African–American patients also had a trend toward a higher nuclear grade (grade 3/4) than Caucasian patients; this finding was statistically significant in patients with early-stage tumors (50.0% vs 31.1%; P < .001) (Table 3). In addition, expressions of Ki-67, p53, c-erb-B2, p21, and bcl-2 were compiled between African–American and Caucasian patients by stage (early vs advanced stage) from the TJUH registry; data were not available from the SEER database. We found that African–American patients with unilateral tumors of all stages had less ER positivity than Caucasian patients (Table 4). In addition, these African–American patients with all-stage unilateral tumors were found to have significantly higher Ki-67 expression (42.2% vs 28.7%; P < .001) and p53 expression (19.4% vs 13.1%, P < .005) than Caucasian patients with similarly staged tumors (Table 4). There was an additional trend toward increased Ki-67 expression in African–American patients with advanced stage tumors (50.0% vs 47.2%; P = .9521), which was not statistically significant (data not shown). Expression of c-erb-B2, p21, and bcl-2 was not found to be significantly different between African–American and Caucasian patients in all stages of disease (data not shown).
|NCI SEER database (1995–2002) N = 197,274||TJUH database (1995–2002) N = 2230|
|No. (%)||% ER+||No. (%)||% ER+|
|Stage I||5724 (30.6)||55.8||120 (42.4)||60.8|
|Stage II||7524 (40.2)||46.8||108 (38.2)||44.4|
|Stage III||1976 (10.6)||39.2||28 (9.9)||53.6|
|Stage IV||1358 (7.3)||36.7||15 (5.3)||33.3|
|Stage I||78,706 (44.1)||71.3||1075 (55.2)||66.8|
|Stage II||63,619 (35.6)||64.8||628 (32.3)||60.2|
|Stage III||10,980 (6.1)||55.3||140 (7.2)||52.1|
|Stage IV||7581 (4.2)||48.0||39 (2.0)||56.4|
|NCI SEER database (1995–2002)||TJUH database (1995–2002)|
|Grade 1, %||Grade 2, %||Grade 3, %||Undifferentiated, %||Grade 1, %||Grade 2, %||Grade 3, %||Undifferentiated, %|
|Low grade (1–2)||High grade (3–4)||P|
|N = 100 with Stage I/II (early)||50.0||50.0||<.0010|
|N = 14 with Stage III/IV (advanced)||50.0||50.0||.4113|
|N = 673 with Stage I/II (early)||68.9||31.1||<.001|
|N = 63 with Stage III/IV (advanced)||61.9||38.1||.4113|
|ER+ (%)||Ki-67+ (%)||p53+ (%)|
|African–American||147 (51.9)||120 (42.2)||55 (19.4)|
|Caucasian||1228 (63.1)||559 (28.7)||255 (13.1)|
Using the available data from both registries, we also investigated whether these data correlated with survival. Figure 1 shows survival analysis in Kaplan–Meier estimates in patients with early-stage tumors. Figure 1A shows that survival was decreased in African–American patients compared with Caucasian patients in the National Cancer Institute SEER database, with a parallel trend toward decreased survival in African–American patients observed in the TJUH registry shown in Figure 1B. However, more specifically, we found a more significant difference in survival in African–American patients with Ki-67-positive, early-stage tumors compared with those with Ki-67-negative, early-stage tumors (P < .05) (Fig. 1C), which is similar to differences in Caucasians with Ki-67-positive versus those with Ki-67-negative tumors (P < .05) (Fig. 1D). We further examined the distribution of specific phenotypes, specifically the triple-negative (basal-like, negative for ER, PR, and HER-2) phenotype in the TJUH registry and found a higher incidence of this aggressive subtype in African–American compared with Caucasian patients (20.8% vs 10.4%; P < .0001) (Table 5). This triple-negative phenotype was associated with higher histologic (73.1% grade 3/basal vs 41.3% grade 3/nonbasal; P < .0001) and nuclear (63.5% grade 3/basal vs 31.5% grade 3/nonbasal; P < .0001) grade in all stages in both races (data not shown), and indeed was found to correlate with worsened survival in both races compared with triple-positive phenotypes (Fig. 2 A and 2B). Survival for patients with basal phenotype tumors was not found to be significantly different between races (data not shown).
|Basal-like||59 (20.8)||203 (10.4)||262|
|HER-2+/ER−||11 (3.9)||93 (4.8)||104|
|Luminal A||125 (44.2)||1054 (54.1)||1179|
|Luminal B||26 (9.2)||165 (8.5)||191|
|Unclassified||62 (21.9)||432 (22.2)||494|
|Total||283 (100)||1947 (100)||2230|
In the current study, preliminary information from breast cancer patients entered into our single-institution tumor registry database appeared to correlate with findings extracted from the National Cancer Institute SEER database. African–American patients with invasive breast carcinomas were more likely to present with later stage and higher grade tumors, and higher expression of Ki-67, and were less likely to have ER positivity than Caucasian patients. We found in both the TJUH database and SEER database that, stage for stage, African–American patients were significantly more likely to have higher grade breast carcinomas and were significantly less likely to have ER positivity than Caucasian patients. These data are consistent with other large population-based studies of racial differences in breast carcinoma presentation at the time of diagnosis.4, 6–9, 20 Our study also illustrates that data available in the TJUH registry reflects national trends in demographics, and that further studies from this registry may be utilized as consistent parallel indicators. To our knowledge, the current study also evaluated the largest number of patients presented to date. We plan to open the current study to a network of hospitals in the TJUH system to continue our study of these disparate characteristics and confirm our results in a large population.
Due to these disparate presentations of breast carcinoma characteristics between African–American and Caucasian patients, we examined more specific expressions of proteins previously reported to have different levels of expression in these populations,36 including the Ki-67 proliferation index, tumor suppressor protein p53, cell cycle protein p21, antiapoptotic protein bcl-2, and c-erb-B2 (HER-2/neu) membrane positivity. As described above, we found that African–American patients with all stages of tumors had a significantly higher nuclear grade, Ki-67 proliferation index, and p53 expression than Caucasian women, but expression of p21, bcl-2, and c-erb-B2 was not found to be significantly different. We found that Ki-67 proliferation indexes are higher in African–American patients, correlating with worsened survival. These data indicate that growth signals may differ in aggressive biologic phenotypes of breast cancer, and raises questions as to what drives those signal(s) to explain their disparate behavior and survival rates.
Comprehensive genomic expression profiling studies that compare mRNA arrays from tumors matched with regard to clinical characteristics but from patients of different ethnic backgrounds will become extremely valuable in defining [a] unique set(s) of genes that may drive tumor growth in 1 patient population more than another.46 As mentioned by Carey et al, the basal or “triple-negative” phenotype of breast cancer as identified through genomic profiling is more prevalent in African–American women compared with Caucasian women, and is associated with more aggressive biology, higher tumor grade, and a worse prognosis.39, 40, 42 We found similarly that African–American women in our tumor registry study had a higher likelihood of triple negativity correlating with a poorer prognosis in both our single-institution registry and the national SEER database. Additional microarray analyses between samples from African–American patients matched with breast cancer phenotypes but compared with Caucasian subjects are proposed to identify differences at the mRNA level.47, 48 This important addition to our study will provide valuable information with which to explain disparities in proliferation and survival, and will drive hypotheses toward targeting specific sequences for aggressive therapy and early treatment intervention. Clinical trials for the treatment of basal-like breast cancers currently are underway to evaluate the response to platinum-containing regimens in the setting of disease recurrence, as well as the incorporation of therapy targeted toward angiogenesis.
Data from our current study will be utilized to heighten awareness and to further institute aggressive and specific treatment regimens for different breast cancer phenotypes.49 As previously mentioned, several studies have noted the possible underutilization of breast screening techniques among African-American women of lower socioeconomic status,2, 13–19 and these data may heighten the need to employ these techniques to increase detection rates in this high-risk population. Although some studies have suggested similar response rates to treatment,27, 50 other studies suggest increased rates of treatment failure among African–American patients.2 With the knowledge that basal-type and c-erb-B2-positive subtypes of breast cancer may respond better to certain chemotherapy regimens (as observed in the preoperative setting), the higher propensity for the former subtype in African–American women with breast cancer should result in the prompt initiation of aggressive and specific chemotherapy regimens once these patients are diagnosed.41 The significance of the Ki-67 proliferation index is worthy of further explanation. Differences in cell cycle proteins may warrant the incorporation of cell cycle inhibitors and/or cyclin-dependent kinase inhibitors into treatment regimens.
The results of the current study demonstrate that our tumor registry data from TJUH reflect data in larger databases such as the National Cancer Institute SEER database. In addition, the current study reflects the national trends toward the importance of the early screening, diagnosis, and treatment of breast cancer in the African–American population because, compared with Caucasian patients, this population is more likely to present at a higher stage of disease, with a higher tumor grade and more aggressive biology. It is encouraging that conditional survival (ie, survival predicted after patients have already survived several years after the diagnosis) is particularly heightened in the African–American population.51 However, early intervention once disease is detected will be a key component in improving breast cancer-related survival in this population. More future studies identifying differences in gene expression will help to target and enhance treatments more specifically.
We wish to thank the following physicians who also contributed toward the diagnosis and care of these patients. From the Department of Surgery: Donna Barbot, MD; Barbara Cavanaugh, MD; Herbert Cohn, MD; Emily C. Conant, MD; Steven C. Copit, MD; Jerome Cotler, MD; Paul G. Curcillo, MD; Dion Farria, MD; Stephen Feig, MD; James W. Fox IV, MD; Diane Gillum, MD; Enrique Guttin, MD; Eric Hume, MD; John C. Kairys, MD; Kris R. Kaulback, MD; Marvin Krane, MD; Sun H. Lee, MD; Robert G. McCairns, MD; Stanton B. Miller, MD; John H. Moore, MD; Melvin Moses, MD; Ernest L. Rosato, MD; Francis Rosato, MD; Jennifer L. Sabol, MD; Ed Sauter, MD; Robert Solit, MD; and Susan Trevisan, MD. From the Department of Obstetrics/Gynecology: Lorraine C. King, MD. From the Department of Radiation Oncology: Lydia Komarnicky, MD (currently at Hahnemann University Hospital, Philadelphia, Penn). From the Department of Radiology: Susan DeWyngaert, MD; Valerie Gilliam, MD; Frank Paolantonio, MD; Catherine W. Piccoli, MD; and Annina N. Wilkes, MD. From the Department of Pathology: Harvey Bellin, MD; Corrado Minimo, MD; and Roland Schwarting, MD. Above all, we thank the patients whose breast tumor characteristics were analyzed in the above databases
- 2Breast cancer in minority women. In: SingletarySE, RobbGL, HortobagyiGN, editors. Advanced Therapy of Breast Disease. 2nd ed. Hamilton, Ontario, Canada: BC Decker, Inc, 2004: 713–726..
- 11African–American ethnicity, socio-economic status, and breast cancer survival. A meta-analysis of 14 studies involving over 10,000 African–American and 40,000 white American patients with carcinoma of the breast. Cancer. 2002; 94: 2844–2854., , , et al.
- 15Black women receive less mammography even with similar use of primary care. Ann Intern Med. 1996; 125: 151–156., , , et al.
- 17Factors affecting breast cancer screening practices among African American women in Western New York. Breast Cancer Res Treat. 2005; 94 ( suppl 1): S157. Abstract 3093., , , et al.
- 19Racial differences in breast cancer of lower socio-economic status women. 27th Annual San Antonio Breast Cancer Symp. 2004. Abstract 2076., , , et al.
- 23Delays and receipt of adjuvant chemotherapy and radiation therapy in breast cancer patients: a population-based study. Breast Cancer Res Treat. 2005; 94( Suppl 1) S155. Abstract 3089., , , et al.
- 24Breast cancer treatment and survival in African Americans and Caucasians. Breast Cancer Res Treat. 2005; 94( suppl 1): S158. Abstract 3096., , , et al.
- 27Understanding racial differences in outcome from metastatic breast cancer (MBC): a pooled analysis of Cancer and Leukemia Group B (CALGB) 9342 and 9840. Breast Cancer Res Treat. 2005; 94 ( suppl 1): S158. Abstract 3097., , , et al.
- 34A comparison of clinicopathological features and molecular markers in British and Nigerian women with breast cancer. Breast Cancer Res Treat. 2005; 94( suppl 1): S160. Abstract 3101., , , et al.
- 39Beyond anatomic staging: are we ready to take the leap to molecular classification? American Society of Clinical Oncology 2005 Educational Book. 2005., , , et al.
- 42Race and the poor prognosis basal breast tumor (BBT) phenotype in the population-based Carolina Breast Cancer Study (CBCS). J Clin Oncol. 2004; 22( suppl): 14S. Abstract 9510., , , et al.
- 43“Intrinsic gene expression” subtypes correlated with grade and morphometric parameters reveal a high proportion of aggressive basal-like tumors among black women of African ancestry. J Clin Oncol. 2004; 22( suppl): 14S. Abstract 9509., , , et al.
- 45Epidemiology in Medicine. 1st ed Boston: Little, Brown and Company: 1987..
- 51Impact of ethnicity on conditional survival of breast cancer patients: analysis from the SEER database. Breast Cancer Res Treat. 2005; 94( suppl 1): S156. Abstract 3090., , , et al.