Long-term outcomes and clinicopathologic differences of African-American versus white patients treated with breast conservation therapy for early-stage breast cancer

Authors

  • Meena S. Moran MD,

    Corresponding author
    1. Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, Connecticut
    • Department of Therapeutic Radiology, Yale University School of Medicine, 333 Cedar Street, PO Box 208040, New Haven, CT 06520-8040===

    Search for more papers by this author
    • Fax: (203) 785-4622

  • Qifeng Yang MD, PhD,

    1. Department of Breast Surgery, Qilu Hospital, Shandong University, Jinan, People's Republic of China
    2. Department of Radiation Oncology, UMDNJ-Robert Wood Johnson School of Medicine, the Cancer Institute of New Jersey, New Brunswick, New Jersey
    Search for more papers by this author
  • Lyndsay N. Harris MD,

    1. Department of Medical Oncology, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • Beth Jones PhD, MPH,

    1. Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • David P. Tuck MD,

    1. Department of Pathology, Yale University School of Medicine, New Haven, Connecticut
    Search for more papers by this author
  • Bruce G. Haffty MD

    1. Department of Radiation Oncology, UMDNJ-Robert Wood Johnson School of Medicine, the Cancer Institute of New Jersey, New Brunswick, New Jersey
    Search for more papers by this author

Abstract

BACKGROUND.

African–American (AA) and white patients with early–stage disease who were treated with breast conservation therapy (BCT) were examined to detect differences in clinicopathologic features and outcomes as a function of race.

METHODS.

Clinical data from the charts of 2164 white and 207 AA patients treated with BCT, and p53 expression status on 444 patients (from an existing tissue database), were analyzed to detect differences between the 2 cohorts.

RESULTS.

The median follow-up was 7 years. There were no differences in the method of tumor detection, lymph nodes excised, surgical margin status, or chemotherapy/radiotherapy delivered, reflecting similar screening and treatment policies for AA women in the study community. Despite this, AA patient presented at a younger age, with higher T and N classifications, and more estrogen and progesterone negative and “triple negative” tumors (all P values <.016). Tumors in AA patients were p53 positive more often than tumors in white patients (P = .0003). At 10 years, AA patients had a higher rate of distant metastasis (20% vs 17%; P = .042), lymph node recurrence (6% vs 2%; P = .004), and breast recurrence (17% vs 13%; P = .036). There was no difference in overall survival between the 2 groups. On multivariate analysis, only lymph node recurrence (risk ratio of 3.140; 95% confidence interval, 1.396-7.063 [P = .0057]) remained significantly higher among AA women.

CONCLUSIONS.

In this cohort of uniformly treated patients, the authors found the expected clinicopathologic differences, but race was not found to be an independent predictor of local recurrence for AA patients when other confounding variables were taken into account in the multivariate model. These findings suggest that BCT is a reasonable option for appropriately selected AA patients. To the authors' knowledge, this is the largest study addressing outcomes after BCT for AA women published to date. Cancer 2008. © 2008 American Cancer Society.

Breast cancer is the most commonly diagnosed solid tumor among women in the US and the second leading cause of death for women with cancer. Among the estimated 178,480 newly diagnosed breast cancer cases in 2007, 19,010 are expected to occur in African American (AA) women.1 There is increasing evidence that breast cancer in the AA population demonstrates unique clinical behavior compared with the white breast cancer population. The overall incidence of breast cancer in AA is lower than in white women, yet paradoxically, the breast cancer mortality rates are higher.1‒3 In addition, AA patients present at a younger age1, 4, 5 with a later stage of disease at diagnosis,5‒7 and, within each stage, worse 5-year survival rates.4 Whether these differences are because of disparities in socioeconomic status,8‒26 utilization or access to screening and/or treatment,9, 22, 26‒28 or underlying biologic differences6, 29‒37 is poorly understood.

Breast conservation therapy (BCT) is recognized as a conventional treatment option for early-stage breast cancer. Although multiple randomized, prospective studies have shown the equivalence of BCT to mastectomy for early-stage breast cancer,38, 39 local-regional control as a function of race has not been analyzed prospectively. The clinical behavior of breast cancers in the AA compared with white populations suggests that AAs have biologically more aggressive disease and warrants an in-depth look at the appropriateness of BCT for AA patients. In addition, AA breast cancer patients present at a younger age, and there are data suggesting that younger women have worse overall outcomes compared with their older counterparts.40‒42 Although studies have documented that AA patients undergo BCT less frequently than mastectomy43 and that radiation is omitted after conservative surgery more often in the AA population,44, 45 there are a paucity of data regarding local outcomes for AA women after breast-conserving surgery and radiotherapy to guide the physician and patient in the decision-making process of BCT versus mastectomy for early-stage breast cancer.

This study was undertaken to analyze clinicopathologic and outcome parameters between AA and white patients treated with BCT at our institution to detect racial disparities in treatment and outcomes, in an effort to determine the appropriateness of BCT for AA patients with early-stage breast cancer.

MATERIALS AND METHODS

Between 1975 and 2003, 207 AA and 2164 white women with early-stage breast cancer were treated with BCT at the radiation therapy facilities of Yale University School of Medicine. A small number (n = 45) of other races, including Asian, Native American, and Hispanic, were excluded from this comparison of AA and white patients. After obtaining Yale Institutional Review Board approval, a database was created to further analyze disease parameters, treatments, and outcomes. Race information was obtained from the radiation therapy medical records as self-reported by patients, based on preconsultation questionnaires. All patients were treated with breast-conserving surgery, which consisted of lumpectomy, quadrantectomy, or partial mastectomy. Because of the era in which they were treated, the surgical procedure for evaluation of the axilla for the vast majority of patients consisted of a complete axillary lymph node dissection. Adjuvant radiation was delivered in all patients to the whole breast to a median dose of 48 grays (Gy) with a cone-down, to a total median dose of 64 Gy. Regional lymph node radiation was delivered as clinically indicated and has been previously described.46 Adjuvant systemic therapy was delivered at the discretion of the treating physician. All available patient data, including clinicopathologic features, estrogen receptor (ER) and progesterone receptor (PR) status, HER-2/neu status, and treatment and outcome parameters were included in the analysis.

In addition, p53 expression was analyzed as a function of race on patients from our existing Tissue Micro-Array (TMA) database. This tissue database was constructed to represent approximately 20% of our patient population, with careful selection of cases to reflect the known clinical and pathologic features of our large departmental breast database (percentages of patients by race and histology in both the TMA and departmental database were similarly represented). The TMAs were constructed from tumor sections of hematoxylin and eosin–stained slides of the archived paraffin blocks selected by an experienced pathologist. Two 0.6-mm cores were extracted using a Tissue Microarrayer, resulting in 888 cores. This validated technique of 2-fold redundancy has been previously described and correlates to whole section staining.47 An anti-p53 antibody with a dilution of 1:20 was used (clone D07; DAKO, Carpentaria, Calif). For immunohistochemical evaluation of p53, nuclear labeling of tumor cells was classified as either negative (≤10% of stained tumor cells) or positive (>10% of tumor cells stained). To minimize interobserver variability and differences in the laboratory testing methods, the p53 testing was standardized so that all tissue samples were tested in a single laboratory with results read by 1 pathologist.

The data were analyzed to detect differences between the 2 racial groups and correlated with patient outcomes. Breast recurrence-free survival was defined by time to local failure in the treated breast. Lymph node recurrence-free survival was time to regional failure (axilla, supraclavicular fossa, infraclavicular fossa, or internal mammary lymph node). Distant metastasis-free survival was defined by time to disease failure outside the local or regional area. All events were calculated from the time of initial diagnosis to the time of the event. Comparison of clinical and pathologic characteristics between the AA and white groups was evaluated using chi-square analysis. The outcome endpoints were calculated using standard life-table methods, and the differences were compared using Cox regression models.

The outcome parameters (breast recurrence, lymph node recurrence, distant disease-free survival, and overall survival) were analyzed by multivariate analysis incorporating age at diagnosis, T classification, lymph node status, margin status, ER, PR, and triple–negative status in the regression model.

RESULTS

As of September 2006, the median follow-up time was 7 years for both cohorts. The AA group constituted 9% of our patient population treated with BCT. The distribution of AA in the BCT cohort did not differ significantly from the distribution of AA women in the overall breast cancer population at our institution during this time interval (10% AA in the overall breast cancer population at our institution, based on hospital tumor registry data).

As outlined in Table 1, the method of tumor detection did not differ significantly by racial group. The AA and white groups both presented with equivalent frequency of nonpalpable, mammographically detected breast cancers (45% each). There were no significant differences in frequency of tumors detected by physical examination alone, those that were palpable/mammographically detected, or those that were palpable/mammographically negative in the AA versus white groups (P = .165).

Table 1. Screening and Treatment Variables
Screening/Treatment ParameterAfrican AmericanWhiteP
  1. PE indicates physical examination; +, positive; −, negative; CMF, cyclophosphamide, methotrexate, and 5-fluorouracil; CMFV, cyclophosphamide, methotrexate, 5-fluorouracil, and vincristine; CAF, cyclophosphamide, doxorubicin, and 5-fluorouracil; AC, doxorubicin and cyclophosphamide; ACT, doxorubicin, cyclophosphamide, and paclitaxel; High dose, high-dose chemotherapy and stem cell/bone marrow transplantation; RT, radiation therapy; Gy, grey.

Method of detection, %  .165
 PE alone1413 
 Mammography alone4545 
 PE+/mammography +3632 
 PE+/mammography−59 
 Unknown0.00.2 
Lymph nodes removed, n8.47.9.45
Positive surgical margin, %129.26
Chemotherapy delivered, n (%)  .11
 CMF40 (53)266 (56) 
 CMFV0 (0)13 (3) 
 CAF9 (12)50 (11) 
 AC13 (17)89 (19) 
 ACT10 (13)23 (5) 
 High dose0 (0)6 (1) 
 Other3 (4)26 (5) 
Tamoxifen in receptor-positive patients, n (%)  .45
 Yes18 (26)220 (22) 
 No50 (74)778 (78) 
Median RT dose to tumor bed, Gy6464.99

Evaluation of the pathology reports from the surgical specimens revealed that the number of lymph nodes removed on axillary lymph node dissection did not differ significantly by race (P = .45). The percentage of patients with positive surgical margins also did not differ between the 2 groups (P = .26). For adjuvant treatment, the median total radiation dose delivered for the 2 cohorts did not differ (64 Gy for each; P = .99). The frequency of the various regimens used (cyclophosphamide, doxorubicin, and 5-fluorouracil; cyclophosphamide, methotrexate, and 5-fluorouracil; cyclophosphamide, methotrexate, 5-fluorouracil, and vincristine; doxorubicin and cyclophosphamide; doxorubicin, cyclophosphamide, and paclitaxel, etc.) did not differ between the 2 groups (P = .11). The percentage of patients who were hormone receptor positive and received tamoxifen did not differ significantly between the 2 cohorts (P = .45). The treatment parameters by race are highlighted in Table 1.

Despite these similarities in screening and treatment, there were differences noted in the clinical presentation and pathologic features in the AA versus white populations. A comparison of the clinical and pathologic characteristics analyzed by racial group is summarized in Table 2. Statistically significant differences were found in the age of presentation, with a higher percentage of AA patients presenting at age ≤40 years (P = .0158). The AA patients had a higher T stage classification (P = .0001) and a higher N stage classification (P = .0001) at the time of presentation. In addition, the AA cohort had more ER-negative tumors (54% vs 36%; P = .0001), PR-negative tumors (58% vs 47%; P = .0097), and ‘triple–negative’ tumors (ER/PR/HER-2/neu negative) compared with the white cohort (21% vs 8%; P < .0001). The expression of HER-2/neu did not differ significantly for the 2 cohorts (28% vs 27%; P = .8832). As expected, given the difference in receptor status, the overall percentage of patients receiving tamoxifen was significantly less for the AA cohort (27% vs 33%; P = .048), but the percentage of patients who were hormone receptor positive who received tamoxifen did not differ significantly (26% vs 22%; P = .45), as stated above. There were no differences in family history, tumor histology, or HER-2/neu status noted between the 2 groups.

Table 2. Clinical and Pathologic Features of Breast Cancer by Race
FeaturesRace Distribution
African American, n (%)White, n (%)P
  • *

    Triple negative indicates negative expression of estrogen receptor, progesterone receptor, and HER-2/neu.

Age, y  .0158
 ≤4038 (20)264 (12) 
 >40169 (80)1900 (88) 
Pathologic status  .0001
 T1129 (68)1505 (82) 
 T262 (32)333 (18) 
Lymph node status  .0001
 N0121 (75)1176 (79) 
 N133 (21)295 (20) 
 N27 (4)10 (1) 
Family history  .3484
 Positive65 (34)698 (37) 
 Negative129 (66)1183 (63) 
Histology  .1056
 Infiltrating ductal147 (41)1517 (44) 
 Infiltrating lobular5 (2)136 (6) 
 Tubular2 (1)46 (2) 
 Medullary6 (3)36 (2) 
 Mucinous6 (3)35 (2) 
 Other (including mixed histologies)45 (22)394 (18) 
Estrogen status  .0001
 Positive69 (45)1004 (64) 
 Negative84 (54)572 (36) 
Progesterone status  .0097
 Positive62 (42)751 (53) 
 Negative87 (58)670 (47) 
HER-2/neu status  .8832
 Positive19 (28)112 (27) 
 Negative49 (72)305 (73) 
Triple negative*  <.0001
 Yes21 (21)76 (8) 
 No78 (79)892 (92) 
Surgical margin status   
 Positive20 (12)132 (9).2603
 Negative143 (88)1284 (91) 

From our existing TMA database, a total of 444 patients were tested for p53, of which 388 were white and 56 were AA. The AA population had a significantly higher frequency of p53-positive tumors than the white cohort (32% vs 13%; P = .0003). Representative positive and negative cases are depicted in Figure 1. The frequency of p53 by race is presented in Table 3.

Figure 1.

Representative p53-positive and p53-negative slides are shown. The left core was positive for p53, whereas the right core was negative.

Table 3. p53 Expression From Tissue Microarray Database as a Function of Race (P = .0003)
 African American, n (%)White, n (%)Total
  1. The percentage is given as a fraction of the total number of patients tested in each cohort.

p53 negative38 (68%)336 (86%)374
p53 positive18 (32%)52 (13%)70
Total56388444 (100%)

By univariate analysis at 10 years (Table 4), AA patients had higher rates of distant metastasis (20% vs 17%; P = .0423), lymph node recurrence (6% vs 2%; P = .0043), and breast recurrence (17% vs.13%; P = .0365). There was no difference noted in overall survival (81% vs 78%; P = .7731). When these outcome parameters were analyzed using Cox regression multivariate analysis taking age, T classification, lymph node status, margin status, ER, PR, and triple–negative status into account (Table 5), only lymph node recurrence remained higher in the AA cohort (risk ratio of 3.148; 95% confidence interval, 1.047-9.467 [P = .0412]). Race was not an independent predictor of breast recurrence, distant metastasis, or overall survival by multivariate analysis when taking into account the other relevant prognostic features. Figure 2 shows breast recurrence-free survival, lymph node recurrence-free survival, distant metastasis-free survival, and overall survival for the 2 populations truncated at 10 years.

Figure 2.

Ten-year outcomes by race are shown. (Top left) Breast recurrence-free survival. (Top right) Lymph node recurrence-free survival. (Bottom left) Distant metastasis-free survival. (Bottom right) Overall survival. AA indicates African American; C, Caucasian/White.

Table 4. Ten-year Outcomes as a Function of Race
OutcomeRace
African American, %White, %P
  1. NS indicates not significant.

Breast recurrence-free survival8387.0365
Lymph node recurrence-free survival9498.0043
Distant disease-free survival8083.0423
Overall survival8178NS
Table 5. Results of Multivariate Cox Regression Analysis
VariablesBreast RecurrenceDistant MetastasisLymph Node Metastasis
Hazards Ratio (95% CI)PHazards Ratio (95% CI)PHazards Ratio (95% CI)P
  • 95% CI indicates 95% confidence interval; ER, estrogen receptor; PR, progesterone receptor.

  • *

    Triple negative indicates negative expression of ER, PR, and HER-2/neu.

Race .3979 .6544 .0412
 White1.0 (referent) 1.0 (referent) 1.0 (referent) 
 African American1.369 (0.661-2.835) 1.145 (0.633-2.070) 3.148 (1.047-9.467) 
Age, y .0228 .0919 .0040
 ≤401.861 (1.090-3.178) 1.516 (0.934-2.458) 4.585 (1.624-12.944) 
 >401.0 (referent) 1.0 (referent) 1.0 (referent) 
Tumor classification .0420 <.0001 .0095
 T11.0 (referent) 1.0 (referent) 1.0 (referent) 
 T21.687 (1.019-2.793) 2.776 (1.862-4.138) 3.473 (1.356-8.899) 
Lymph node status .6632 .0024 .2325
 Negative1.0 (referent) 1.0 (referent) 1.0 (referent) 
 Positive1.025 (0.916-1.147) 1.161 (1.055-1.279) 1.167 (0.906-1.504) 
Surgical margin .0395 .1037 .8531
 Negative1.0 (referent) 1.0 (referent) 1.0 (referent) 
 Positive1.114 (1.005-1.235) 1.078 (0.985-1.180) 1.022 (0.813-1.284) 
ER .7835 .0055 .6267
 Negative1.0 (referent) 1.0 (referent) 1.0 (referent) 
 Positive1.139 (0.451-2.879) 0.367 (0.181-0.745) 0.637 (0.103-3.928) 
PR .0713 .5418 .5920
 Negative1.0 (referent) 1.0 (referent) 1.0 (referent) 
 Positive0.508 (0.244-1.061) 0.816 (0.425-1.568) 1.629 (0.274-9.690) 
Triple negative* .8550 .9664 .4431
 Yes1.0 (referent) 1.0 (referent) 1.0 (referent) 
 No1.098 (0.403-2.990) 0.985 (0.486-1.995) 0.491 (0.080-3.020) 

Given the significant difference in lymph node recurrence rate for the 2 cohorts, details regarding the lymph node recurrences were further explored and are summarized in Table 6. There were a total of 44 lymph node recurrences in all, 8 in the AA group and 36 in the white group. Of all patients who experienced lymph node recurrences, only 1 patient underwent sentinel lymph node biopsy. The remainder underwent either a full axillary lymph node dissection, radiotherapy to the axilla, or both. Of note, the patient who underwent sentinel lymph node biopsy had 3 lymph nodes removed with no pathologic lymph node involvement; her lymph node recurrence was an axillary recurrence outside the radiated field. Eleven patients (1 AA and 10 white patients) underwent radiation to the axilla instead of axillary surgical exploration; for these patients, the median dose delivered to the axilla was 46 Gy (range, 45 Gy-50 Gy). Their recurrences varied from in-field axillary recurrence (1 AA and 6 white patients) to in-field supraclavicular recurrences (0 AA and 4 white patients). Three patients underwent axillary lymph node dissection and radiotherapy for multiple positive lymph nodes (all had >4 lymph nodes positive); all 3 were white and experienced in-field recurrences in the supraclavicular fossa (median dose delivered to supraclavicular fossa and axilla was 46 Gy).

Table 6. Lymph Node Recurrences in African American Compared With White Patients
 African American, n (%)White, n (%)Total
IFOFIFOF
  1. IF indicates “in-field” recurrence (lymph node recurrence in a previously radiated area); OF, “out-of-field” recurrence (no radiation treatment previously received at the site of recurrence); SC, supraclavicular; IC, infraclavicular; IM, internal mammary.

  2. Percentages are expressed as a fraction of the total number of patients in each cohort (207 African American patients and 2164 white patients).

Axillary recurrence2 (1)2 (1)7 (0.3)6 (0.3)17
SC recurrence2 (1)1 (0.5)9 (0.4)2 (0.1)14
IC recurrence0 (0)0 (0)0 (0)1 (0.01)1
IM recurrence1 (0.5)0 (0)0 (0)0 (0)1
Unclear11
Total recurrences5316944

DISCUSSION

To the best of our knowledge, this is the largest investigation with the longest follow-up evaluating locoregional outcomes of AA patients who have undergone breast–conserving surgery and radiotherapy for early-stage breast cancer published to date. It is notable that we found no apparent disparities in methods of tumor detection, numbers of lymph nodes removed, positive surgical margin rates, radiation doses delivered, type of chemotherapy delivered, or use of hormone therapy in receptor-positive patients, suggesting that our 2 racially distinct cohorts received similar screening and treatment at our institution. Despite these apparent similarities, we were able to demonstrate significant clinical and pathologic differences between the 2 groups. The differences we demonstrated in our AA cohort compare favorably to other publications on the topic, suggesting that our AA population is representative of larger AA breast cancer populations and not distinct to our geographic location. The percentage of patients diagnosed before age 40 years in our study was higher in the AA group and corresponds to the national trend reported by the American Cancer Society and other large databases.1 As generally reported, we also demonstrated higher T classification and N classification in our AA cohort, which parallels the findings of the National Surgical Adjuvant Breast and Bowel Project,48 the American College of Surgeons,49 the American Cancer Society,1 and the Surveillance, Epidemiology, and End Results program of the National Cancer Institute.50 The highly significant differences in hormone receptor status in our AA population are comparable to the findings of multiple publications,6, 50, 51 including the Women's Health Initiative,52 a large longitudinal study with patients recruited from >40 institutions across the country. Lastly, the higher incidence of the triple–negative phenotype noted in our AA cohort is consistent with reports originating from geographically distinct databases such as the California Cancer Registry, the Carolina Breast Cancer Study, and others.53‒55

Although there is convincing evidence that AA breast cancers are clinically and biologically more aggressive, there are limited data available examining long-term outcomes after BCT for the AA population as a whole. Whether the more aggressive clinicopathologic features of AA breast cancers translate into higher breast recurrence rates after BCT remains an unanswered question; local-regional outcomes data from prospective studies reporting outcomes after BCT for AA patients are lacking, and although a few retrospective studies have attempted to address this issue, they were limited by their small sample sizes.55‒60 In 1992, Pierce et al reported on outcomes after BCT for 75 AA and 615 white patients. Despite the small numbers of patients in their series, the authors found a higher rate of lymph node recurrence in AA patients and no difference in local control,56 which is consistent with our findings. An additional few small series have attempted to speak to the issue of locoregional recurrence for AA patients after BCT; none have found a significant difference in local recurrence. Table 7 shows a comparison of the studies addressing locoregional control after BCT.

Table 7. Studies of Outcomes of Breast Conservation Therapy in African American Patients
StudyAfrican American Patients, nWhite Patients, nMedian Follow-up, yFindings
  • LR indicate local recurrence; NR, lymph node recurrence; NS, not significant; NA, not available; DFS, disease-free survival; BCT, breast conservation therapy; MRM, modified radical mastectomy.

  • *

    Subset of patients in study received BCT.

Pierce et al 199256756154.7No difference in LR (5% vs 6%). Higher rate of NR for African American patients (16% vs 4%; P = .001)
Connor et al 200057712044.4LR rate of 13% in African American vs 4% in white patients (P = NS)
Burri et al 20025872NA5.0LR rate of 12% at 10 y, comparable to historic controls
Burri et al 2004591021625.0No difference in LR or DFS at 5 y
Newman et al 199960*42NA5.7Similar LR rate and survival in patients treated with BCT and MRM
Nicolaou et al 199961413016.1Did not address LR. NR rate was higher in African American vs white patients (19% vs 1%; P < .0001)
Current study20721647.0LR rate of 17% vs 13% at 10 y. P = NS by multivariate analysis. NR rate was higher in African American vs white patients (6% vs 2%; P = .0043)

It is important to note that although we found a significant difference in local control for the AA and white groups on univariate analysis, the magnitude of the difference was small (17% vs 13%), and race was not found to be an independent predictor of local recurrence on multivariate analysis when we incorporated the relevant variables into the regression model. These results suggest that, although underlying biologic/molecular differences between AA and white patients exist, local control after BCT does not appear to be compromised for AA patients when taking into account the relevant factors that can affect local control. AA women should not be considered poor candidates for BCT; the decision should be individualized based on clinical and pathologic features, and not race alone. Patients should be selected for BCT based on careful assessment of factors that influence local control, such as surgical margin status.

In addition, although the regional recurrence rate for the AA group remained significantly higher on multivariate analysis, this finding did not translate into worse distant recurrence-free survival or overall survival. Although we attempted to discern patterns of failure (in-field or out-of-field) for the patients who experienced lymph node recurrence based on the location of the recurrence and treatment delivered, our absolute number of lymph node recurrences was small (8 in AA patients and 36 in white patients), and therefore no definitive conclusions can be drawn regarding regional treatment based on these data. It is important to note that all patients with lymph node recurrence appeared to be appropriately treated based on our institutional policies (ie, all lymph node–positive patients underwent a full lymph node dissection, some with the addition of axillary radiotherapy; adjuvant radiation to the axilla was delivered to those who did not have any surgical exploration of the axilla and all lymph node–positive patients received supraclavicular radiation). Clinicians should be aware of the slightly higher risk of regional recurrence after BCT in the AA population, with careful scrutiny of lymph node basins to optimize regional control. We would recommend careful assessment for adequate surgical lymph node treatment (sufficient number of lymph nodes removed at dissection, full lymph node dissection after a positive sentinel lymph node biopsy), meticulous computed tomography treatment planning for adequate dose delivery to the regional lymph nodes (supraclavicular, axilla, and internal mammary, as indicated), and consideration of a boost to the lymph node regions, if appropriate.

We found that the percentage of patients in our existing TMA database with overexpression of p53 was significantly higher for the AA cohort. In a normal cell cycle, any cellular stresses that can potentially lead to cell cancerization can activate the p53 pathway by either stopping the cell cycle to repair the lesion(s) or switching on the programmed cell death pathways (apoptosis), forcing the damaged cells to “commit suicide.” In this way, the p53 protein protects the genome by preventing cancer formation, and the loss of p53 is a known predictor of worse prognosis in the general breast cancer population. To our knowledge, there are a paucity of data regarding the expression of p53 in AA breast cancer patients. This study adds to the limited literature on this topic by expanding on the small numbers of AA patients tested for p53 expression and supports the existing data62‒64 suggesting a higher prevalence of p53 overexpression in AA versus white patients. In addition, to our knowledge, the current study is the first to report p53 expression and clinical outcomes on AA patients treated with BCT.

We recognize the major limitation of our study: it is retrospective in nature, with all the inherent flaws of a retrospective study. In addition to the potential flaws of selection bias and retrospective data collection, ours is a single-institution series, which is limited in the numbers of patients (in comparison to larger multi-institutional databases). Although our patients were not drawn from a large geographic area, the multidisciplinary, comprehensive cancer care setting allowed for similar classification of patients into prognostic (and therefore, treatment) categories to ensure homogeneous treatment policies for the 2 cohorts. In addition, although race was classified by patient self-identification and is therefore relatively reliable, subgroups of ethnicity were not considered (ie, AA from African vs Caribbean vs European lineage, etc). It is important to note that we included all conservatively treated AA and white patients at our institution during the time span of the study, and that the racial distribution of patients did not differ from those in our overall institutional tumor registry database, thereby minimizing selection biases.

Another major shortcoming of the current study is the limited information regarding chemotherapy that was available in our radiation therapy departmental records; although the regimen delivered was available for each patient, details regarding dose reductions, adherence, and compliance issues are a potential confounder and will be further explored in a future study. Given that all other studies attempting to show race as an independent predictor for local control after BCT were also retrospective and had significantly smaller sample numbers, we underscore the need for clinicians to encourage their AA breast cancer patients to enroll in studies to improve the understanding of the ethnicity-based differences in breast cancer. Clearly, additional prospective and retrospective studies are required to further analyze prognostic variables for AA patients undergoing BCT. Future studies addressing disparities as well as investigations of molecular and genetic profiling are warranted.

In summary, AA women with early-stage breast cancer, after receiving apparently similar screening and treatment to that of their white counterparts, appear to have equivalent local and survival outcomes after BCT. As expected, the AA cohort as a whole had more aggressive clinical and pathologic features, which may in part contribute to the higher rate of regional recurrence. BCT should continue be offered to AA patients with early-stage breast cancer based on individual clinical and tumor characteristics, with emphasis on surgical and radiation treatment factors to optimize regional control, and should not be based on race alone. Additional studies of the underlying molecular and genetic differences in these racial populations may help to further identify prognostic factors for regional recurrence.

Ancillary