Racial/ethnic differences in survival rates in a population-based series of men with breast carcinoma


  • The opinions expressed herein do not necessarily reflect the view or policies of the U.S. Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.

  • This article is a US Government work and, as such, is in the public domain in the United States of America.



A rare occurrence, about 1500 men in the United States develop breast carcinoma each year. Little is known about survival patterns at the population level, particularly about racial/ethnic variation.


Using data from the Surveillance, Epidemiology, and End Results Program, we examined survival rates in 1979 men diagnosed with primary invasive breast carcinoma between 1973 and 1997. Race was defined as non-Hispanic white, non-Hispanic black, and other race/ethnicity (predominantly Asian/Pacific Islander and Hispanic). The two outcomes were all-cause and breast carcinoma- specific mortality. Survival curves were drawn using Kaplan–Meier estimates and Cox regression was used to estimate the risk of death with hazard ratios and 95% confidence intervals. For both outcomes, the racial/ethnic survival curves differed significantly when the log rank test was used. Therefore, separate models were run for each racial/ethnic group. Covariates included age, stage, histology, surgery, radiation therapy, and year of diagnosis. Estrogen and progesterone receptor status were available for 616 men.


Survival rates differed significantly by race/ethnicity. Overall, 5-year survival rates were 66% for whites, 57% for blacks, and 75% for men of other race/ethnicity. Blacks presented with more advanced disease. By stage, whites and blacks had worse survival rates compared with men of other race/ethnicity. The effects of prognostic factors such as age, surgery type, and radiation were similar, but not always significant, for all groups. Diagnosis year and estrogen receptor status did not affect survival.


Survival following male breast carcinoma differed by race/ethnicity, whereas the prognostic factors associated with survival were similar. Cancer 2002;94:2836–43. Published 2002 American Cancer Society.

DOI 10.1002/cncr.10521

Male breast carcinoma (MBC) is diagnosed each year in approximately 1500 men in the United States and is associated with 400 deaths. 1 Because of its rarity, MBC has not generated much research and its clinical management has been guided by research on the disease in females. To assess the prognosis of MBC, clinicians also rely on data from case-series of MBC and retrospective reviews of groups of males, but these studies often are hampered by a small population and reflect the experience of single institutions. Two early studies that examined survival after breast carcinoma did not find any differences between black and white men, but other nonwhite groups were not included. 2, 3 Little is known about population-level patterns of survival after MBC, particularly its variation by race/ethnicity. Yet in females, survival patterns show significant racial/ethnic variation. Compared with whites, black women have shorter survival rates, Hispanic women have comparable survival rates, and Asian/Pacific Islander women have longer survival rates. 4–6 As the published information on racial/ethnic differences in MBC survival rates is so sparse and patterns have never been examined in population-based data, we examined 2468 cases of male invasive breast carcinoma identified since 1973 through the population-based Surveillance, Epidemiology and End Results (SEER) Program. Specifically, we asked: 1) Do survival rates after MBC vary by race/ethnicity? 2) What factors affect survival rates? 3) Do these factors vary by race/ethnicity?


Study Population

Study data were derived from the SEER Program, a consortium of 11 population- based cancer registries supported by the National Cancer Institute. Nine registries have been monitoring cancer incidence by active surveillance since 1973. In 1992, registries in Los Angeles and the San Jose-Monterey region joined SEER, increasing the coverage to approximately 14% of the U.S. population.

This study identified all males with pathologically confirmed, invasive breast carcinoma diagnosed in the nine SEER regions operating from 1973 through 1997 (N = 2468). We excluded men with previous primary cancers (to eliminate treatment-related breast carcinoma; n = 311), sarcomas (n = 6), those with unknown stage at diagnosis, race/ ethnicity, or survival time (n =135), and those diagnosed at autopsy or identified by death certificate (n = 37). Therefore, our final sample included 1979 men.

Measurement of Variables

SEER routinely collects medical record data on demographics (including race/ ethnicity), tumor characteristics, stage at diagnosis, date of diagnosis, and treatment in the first 4 months following diagnosis. It obtains date of death or censoring and cause of death from National Death Index for Vital Statistics.

Using the International Classification of Disease (ICD)-O coding system, tumors were classified histologically as ductal/other (8000–8020, 8070–8231, 8400–8502, 8504, 8510), papillary (8050, 8260, 8503), mucinous (8480, 8481), lobular (8520–8522), Paget (8540–8543), or inflammatory (8530). 7 The 14 cases of inflammatory breast carcinoma and 33 cases of Paget disease were combined into the ductal/other group after analyses showed no differences in survival rates among the three groups.

MBC was classified by stage using the SEER historic stage variable, as this staging system was used consistently during the 25-year study period. Stage at diagnosis was defined as local (confined entirely to the breast), regional (extended into surrounding organs/tissues or into the regional lymph nodes), or distant (spread to remote parts of the body). 8 Staging systems for SEER and the American Joint Committee on Cancer (AJCC) overlap. In general, SEER local stage corresponds to the AJCC Stage I, SEER regional stage corresponds to AJCC Stages II and III, and SEER distant stage approximates AJCC Stage IV. 9

SEER data for the first course of treatment consisted of the type of surgery and radiotherapy (yes/no). Mastectomies were categorized as partial/ lumpectomy, simple, modified, radical modified, and not otherwise specified. The mastectomy, not otherwise specified category, was used mostly in the 1970s before detailed surgery data were collected routinely. Radiation therapy was categorized into any/ none. Specifics of chemotherapy and hormonal therapy were not available.

For cases diagnosed after 1988, SEER retained tumor size and the number of positive lymph nodes as individual variables, in addition to the composite historical stage variable.Tumor grade was also collected. Estrogen (ER) and progesterone receptor status (PR) have been collected since 1990. Therefore, analyses using these variables were limited to patients diagnosed from 1993 to 1997 (n = 616) when data for these variables were complete. ER and PR status were classified as positive, negative, and unknown, with borderline results included in the unknown group.

In the SEER registries, patient follow-up after diagnosis is undertaken routinely through record linkages with death certificates, voting and driver's license records, as well as through information from hospital tumor registries, physicians, and obituary reviews. For this study, survival time in months through December 31,1997 was calculated from the diagnosis date to either the date of death or the date of last follow-up. For cause-specific survival analyses, only deaths with ICD, 9th Revision, codes 1740–1749, were classified as due to breast carcinoma, and males who died of other causes or of unknown cause were treated as censored observations. Results did not differ when the 51 men with unknown cause of death were excluded.

Statistical Analysis

The outcomes of interest were overall mortality and cause-specific (death from breast carcinoma) mortality. Chi-square and exact tests were used to compare the demographic and clinical characteristics associated with survival by race/ethnicity. Survival curves were produced for each outcome using Kaplan– Meier product-limit estimates. Cox proportional hazard regression was used for multivariable modeling, with log-log plots to determine the validity of the proportional hazards assumptions. For both outcomes, the survival curves differed significantly by race/ethnicity using the log rank test (P < 0.002). Therefore, separate Cox proportional hazards models were used to assess the relative contributions of stage, age at diagnosis, histology, year of diagnosis, and type of surgery to death from any cause and from breast carcinoma. Due to the few deaths from breast carcinoma in men of other race/ethnicity, we were unable to present a multivariate model for this outcome. All analyses were performed using SAS statistical software. 10


Among the 1979 males included in the survival analysis, whites comprised 82%, blacks represented 11%, and men of other race/ethnicity accounted for 7% of the total sample. This last group was predominantly Asian/Pacific Islanders (46%) and Hispanics (39%). As preliminary analyses did not show any differences in survival between Hispanics and Asian/Pacific Islanders, we combined the groups to improve statistical power. Table 1 presents characteristics of the study subjects by race. The age of the subjects ranged from 10 to 103 years. Seventeen males were diagnosed before the age of 35 years and 27 males were diagnosed after the age of 90 years. Whites were slightly, but significantly, older at diagnosis. They had a mean age of 65.4 years, compared with age 63.3 years for blacks and age 63.7 years for the other racial/ethnic group. Blacks were significantly more likely to present with regional or distant disease than either of the other racial/ethnic groups. Men of other race/ethnicity were more likely to have lobular carcinoma, more likely to have partial mastectomies/lumpectomies, and less likely to have mastectomies of unspecified type. Blacks and men in the other racial/ethnic group were significantly more likely to have been diagnosed in the later years of the study period. This is consistent with Los Angeles and San Jose-Monterey joining SEER in the last 5 years of the study, two regions with large nonwhite populations.

Table 1. Characteristics of Male Breast Carcinoma Patients at Time of Diagnosis by Race, SEER, 1973–1997a
CharacteristicWhite (N = 1613)Black (N = 226)Other (N = 140)P value
  • SEER: Surveillance, Epidemiology, and End Results Program.

  • a

    Percentages may not total 100 due to rounding.

Mean age (standard deviation)65.4(12.6)63.3(12.8)63.7(13.9)0.03
Age at diagnosis (yrs)
Stage at diagnosis
Surgery (mastectomy) type
 None4(< 1)3(1)0(0)0.00
 Modified radical845(52)110(49)83(59)
 Radical19(1)4(2)1(< 1)
 Not otherwise specified476(30)64(28)26(19)
Unknown surgery48(3)8(4)1(< 1)
Radiation therapy
Year of diagnosis
Estrogen receptor status for subset diagnosed 1993–1997
Progesterone receptor status for subset diagnosed 1993–1997

During the follow-up period, 992 subjects died (50%). Breast carcinoma was listed as the underlying cause for 386 of these subjects (40%). These deaths varied significantly by race/ethnicity. Overall, 51% of whites died of any cause, compared with 58% of blacks dying of any cause and 33% of men of other race/ethnicity. Similar proportions of whites and blacks died of breast carcinoma (20%), whereas only 10% of men of other race/ethnicity died of breast carcinoma. For all stages combined, 5-year all-cause survival rates were 66% for whites, 57% for blacks, and 75% for men of other race/ethnicity. For breast carcinoma-specific survival rates, the 5-year survival rates were higher: 83%, 72%, and 89%.

Figure 1 shows the unadjusted all-cause and breast carcinoma-specific survival curves for males diagnosed with local disease. Survival estimates were consistently higher for men of other race/ethnicity, whereas whites and blacks had similar poorer survival. For men with regional stage cancer (Fig. 2), survival was poorer for blacks than for whites. Five- year survival rates for death caused by breast carcinoma were 78% for whites, 65% for blacks, and 83% for men of other race/ethnicity with 10- year survival rates of 59%, 38%, and 75%, respectively. Among men diagnosed with distant stage disease (Fig. 3), survival remained higher for the other racial/ethnic group, although the number of men was small (n = 13).

Figure 1.

Kaplan–Meier survival estimates, local stage disease.

Figure 2.

Kaplan–Meier survival estimates, regional stage disease.

Figure 3.

Kaplan-Meier survival estimates, distant stage disease.

Table 2 shows that stage at diagnosis was a significant predictor of death from any cause for all racial/ethnic groups and it was the only significant predictor for men of other race/ethnicity after accounting for other factors. Increasing age was significantly associated with all-cause mortality for blacks and whites. Histology was a significant predictor of survival among blacks, for whom lobular breast carcinoma elevated the risk of death compared with ductal/other breast carcinoma. Surgery also altered survival significantly. Compared with men with modified radical mastectomies, those undergoing surgery other than a radical mastectomy were at an increased risk of death, although these findings were not statistically significant for the other racial/ethnic group. Radiation therapy did not affect survival.

Table 2. Multivariate Proportional Hazards Model Results for All-Cause Mortality by Race
All-cause death HR (95% CI)All-cause death HR (95% CI)All-cause death HR (95% CI)
  • HR: hazard ratio; CI: confidence interval.

  • a

    The only patient who had radical mastectomy was included in the not otherwise specified category.

 Local (reference)1.0 ——1.0——1.0
 Regional2.8 (1.6, 4.9)2.5 (1.6, 3.9)3.5 (1.5,8.1)
 Distant7.9 (3.6, 17.6)5.7 (3.2, 10.4)9.6 (2.3,31.9)
Age at diagnosis (yrs)
 10–490.4 (0.3, 0.5)0.5 (0.3, 0.9)1.5 (0.5,4.8)
 50–640.6 (0.5, 0.8)0.8 (0.5, 1.3)0.6 (0.3,1.5)
 65–74 (reference)1.0 ——1.01.0 ——
 75+1.8 (1.5–2.1)1.6 (0.9, 2.9)1.9 (0.8,4.8)
 Ductal/other(reference)1.0 ——1.0 ——1.0——
 Papillary0.4 (0.1, 1.4)0.6 (0.23, 1.6)2.2 (0.3,19.4)
 Mucinous1.7 (0.6, 4.5)1.9 (0.6, 6.4)0.9 (0.1,7.7)
 Lobular1.0 (0.4,2.4)16.7 (3.6, 78.1)1.1 (0.1,11.6)
Diagnosis year
 1973–19771.6 (0.4,2.1)1.6 (0.2,2.7)1.6 (0.1,6.3)
 1978–19820.8 (0.3,1.9)1.7 (0.2,2.2)1.7 (0.1,6.5)
 1983–19871.6 (0.8,1.8)1.8 (0.3,1.6)1.8 (0.1,3.4)
 1988–19921.0 (0.6,1.5)1.5 (0.8,2.8)1.9 (0.5,3.9)
 1993–1997(reference)1.0 ——1.0 ——1.0 ——
Surgery (mastectomy type)
 Partial/lumpectomy1.9 (1.4,2.5)2.2 (1.1, 4.4)1.10 (0.6,5.8)
 Simple1.6 (1.1,2.2)1.9 (0.5,3.0)1.11 (0.6,17.2)
 Modified (reference)1.0 ——1.0 ——1.0 ——
 Radical1.0 (0.5,1.9)2.6 (0.8,4.3)0.0a ——
 Not otherwise specified1.9 (1.1,3.1)2.7 (0.5,5.2)2.6 (0.2,9.1)
 None/unknown3.6 (2.2,6.0)2.8 (1.1,6.9)3.2 (0.2,51.1)
 Receiving treatment1.2 (0.9,1.4)1.4 (0.9,2.2)0.5 (0.2,1.4)

For death from breast carcinoma (Table 3), younger age was not protective, but the increased risk of death remained for whites who were 75 years or older.The impact of increasing stage on the risk of death was marked, particularly for blacks. For blacks with regional stage disease, the risk of death from breast carcinoma was 24-fold higher than that for blacks with local disease, although the small sample size led to wide confidence intervals (CIs) around this estimate. The impact of lobular histology on death persisted for blacks, although based on very small numbers, and mucinous histology became a significant predictor. Among blacks and whites, men undergoing partial mastectomies/lumpectomies were more likely to die of breast carcinoma than men who underwent modified mastectomies. For men with no surgery or unknown surgery status, the risk of death was tripled. Although radiation therapy was not associated with mortality for either outcome in any racial/ethnic group, for whites it was of borderline significance (cause-specific hazard ratio [HR] 1.3, P = 0.05). Diagnosis year was not significantly associated with mortality rates in this study.

Table 3. Multivariate Proportional Hazards Model Results for Breast Carcinoma- Specific Mortality by Race
Breast carcinoma death HR (95% CI)Breast carcinoma death HR (95% CI)
  1. HR: hazard ratio; CI: confidence interval.

 Local (reference)1.01.0
 Regional3.9 (2.9, 5.3)12.7 (4.2, 38.1)
 Distant2.8 (8.6,19.1)24.4 (7.3, 89.1)
Age at diagnosis (yrs)
 10–491.0 (0.7, 1.5)1.0 (0.5, 2.3)
 50–641.2 (0.9, 1.6)1.1 (0.6, 2.1)
 65–74 (reference)1.0 ——1.0 ——
 75+1.5 (1.1, 2.1)1.0 (0.5, 2.5)
 Ductal/other(reference)1.0 ——1.0 ——
 Papillary0.4 (0.1, 1.4)1.1 (0.1, 9.0)
 Mucinous1.7 (0.6, 4.5)5.6 (1.0, 30.6)
 Lobular1.0 (0.4, 2.4)44.7 (4.2, 472.4)
Diagnosis year
 1973–19770.9 (0.4, 2.1)0.6 (0.1,3.0)
 1978–19820.8 (0.3,1.9)1.6 (0.1,2.3)
 1983–19871.2 (0.8,1.8)1.7 (0.4,2.4)
 1988–19921.0 (0.6,1.5)1.1 (0.4,2.5)
 1993–1997 (reference)1.0 ——1.0 ——
Surgery (mastectomy type)
 Partial/lumpectomy1.6 (1.2,3.0)3.9 (1.6,9.7)
 Simple1.1 (0.6,2.2)2.3 (0.7,7.5)
 Modified (reference)1.0 ——1.0 ——
 Radical0.9 (0.3,2.9)1.6 (0.2,12.6)
 Not otherwise1.6 (0.7,3.5)1.8 (0.4,7.7)
 None/unknown2.8 (1.3, 6.1)3.5 (1.2,10.8)
 Receiving treatment1.3 (0.9,1.7)1.8 (0.9,3.2)

ER status was not related to either survival end point for any racial/ethnic category. However, white men with PR-positive tumors were less likely to die of breast carcinoma compared with men with PR-negative tumors (HR = 0.3, 95% CI 0.1–1.1, P = 0.07). PR status had no effect on blacks or on men of other race/ethnicity (data not shown). Grade was not significantly associated with either outcome in any of the racial/ethnic groups (data not shown).


Using national population-based data, we examined racial/ethnic variation in survival following MBC, expanding on previous work by characterizing the survival experiences of blacks and men of other race/ethnicity. Although our findings for whites were consistent with those published by others, 11– 18 we found that blacks had survival similar to those of whites and men of other race/ethnicity had substantially increased survival. Although survival differs by race/ethnicity, the prognostic factors were similar, suggesting uncontrolled confounding. Stage was the only indicator significantly associated with survival for men of other race/ethnicity, whereas stage, age, and type of surgery were important for blacks and whites. Rare histologic types were associated with decreased survival in blacks only.

As most survival studies do not report results by race, we cannot directly compare our results with those findings. However, our data for whites support the survival figures reported in clinical series that do not define race. For local disease, our 5- and 10-year all-cause crude survival rates of 76% and 57% in whites concur with the 57–100% (5-year) and the 39– 76% (10-year) ranges reported by others. 11–18 Survival rates following regional and distant stage disease in our study were in the middle of the published ranges (5-year survival rates from 16% to 73% and 10-year survival rates from 10% to 50%). 12–14, 16, 18, 19 The ranges are broad, as sample size, definition of stage, and age of the patients vary among studies.

Small samples have prevented most studies from examining breast carcinoma- specific deaths. Our results are in the high end of the range reported elsewhere. 13, 20, 21 Although error in the underlying cause of death can bias the results, this is unlikely to happen in studies of MBC because the rarity of the disease should result in a careful coding of the diagnosis. The sample was restricted to men with breast carcinoma as the first primary site, eliminating the possibility of metastases to the breast from other sites such as the prostate.

The unadjusted survival rates showed that men of other race/ethnicity have the highest survival rates, whites have intermediate survival rates, and blacks have the lowest survival rates of the three groups. After adjustment for stage, blacks experienced survival rates similar to rates experienced by whites except for regional stage disease, at which they were more likely than whites to die of breast carcinoma. Two studies found no difference in survival after MBC between blacks and whites, although only one adjusted for stage. 2, 3 In a study of MBC deaths in the United States between 1949 and 1958, age-specific death rates were higher for nonwhites than for whites in every age group until age 80. 22 Our result of similar black–white survival curves for all but regional stage differs from that seen in females with breast carcinoma. Black women have consistently lower survival rates than white women have, across all stages 4, 5, 23, 24 although the gap narrows after adjustment for other factors. 25, 26

Published reports of survival rates in males of race/ethnicities other than white or black are few. According to a study of U.S. and Japanese males diagnosed with breast carcinoma (all stages) between 1951 and 1970, mortality rates for Japanese men were only one fouth that of U.S. men (0.9 per 100,000 vs. 3.4 per 100,000). 27 In our study, 46% of the men of other race/ethnicity were Asian/Pacific Islanders, and most of these were Japanese (27 of 64 Asian/Pacific Islanders). The better survival rates for males of the other race/ethnicity group mirror those seen in Asian/Pacific Islander females, 4, 6 suggesting a protective influence unrelated to gender. It may be that genetics, body size, or lifestyle factors such as nutrition moderate both breast carcinoma incidence and survival rates in nonwhites and nonblacks.

Increasing age has been associated with increased mortality rates in some studies. 2, 11, 21, 27 Other studies did not report this finding. 3, 14 When death from breast carcinoma was the outcome (eliminating competing causes of death), increasing age was a significant predictor in white men older than 74 years. Older males may not be treated as aggressively as younger males, due to comorbid conditions or to the belief that older persons cannot tolerate treatment well. A multicenter study of 335 males with breast carcinoma found that being older than age 65 years was a significant predictor for all-cause deaths, but not for disease-specific deaths. 20 Keller, 3 however, reported higher disease-specific survival rates in older people than in younger black and white males with breast carcinoma diagnosed between 1958 and 1963 (n = 181). Differences in study design, such as different criteria for classifying breast carcinoma death, or different analytic methods (e.g., multivariate analysis vs. unadjusted, age-specific mortality rates) may explain why our results differed.

We did not find any differences in survival between patients with ductal carcinoma and with either inflammatory breast carcinoma (known to be very aggressive and fatal in females) 28, 29 or Paget disease, reported to have very poor survival in men. 16 Although lobular and mucinous tumors are associated with better survival in females, a strong increased risk of death was associated with lobular (both outcomes) or mucinous (breast carcinoma-specific death) tumors, albeit only in blacks and based on a small sample. 24 Lobular carcinoma is rare in males, as men usually do not develop breast lobules. Males who develop lobular carcinoma may have other physical differences that alter survival and the lobular carcinoma may be the result of or a marker for another biologic process. Alternatively, the histologic differences may be due to misclassification. Although the histologic distribution of tumors in our study is similar to that reported elsewhere, 19, 30 our results may include misclassification due to the subjective nature of histologic classification. These data were abstracted from hospital pathology records and not reviewed by pathologists.

For both survival outcomes, men treated with either lumpectomy/partial mastectomy or no surgery/unknown types of surgery were at an increased risk of death. It may be, however, that men who underwent less aggressive surgery had comorbid conditions or advanced age that prevented them from undergoing more aggressive surgery. Although we have no data on comorbidity in this study, the men undergoing lumpectomy/partial mastectomy or unknown/no surgery were significantly older than those undergoing other mastectomies. Older men may not be considered good candidates for more aggressive surgery. Two studies of older men reported that lumpectomy was less successful at controlling local spread than other types of mastectomies. 11, 13 Given that the male breast has much less tissue than the female breast, it may be more difficult to perform lumpectomy/partial mastectomy in men.

Unfortunately, these data showed no improvement in survival over 25 years. Although we did not have consistent data on adjuvant treatments, which have changed over the study period, we included the year of diagnosis in the model to reflect changing treatment patterns. A lack of improvement in survival over time has also been reported by others for the periods 1933–1983 and 1955–1996. 11, 19 However, improvement over time was reported in studies from Sweden, Italy, and Wisconsin. 31–33 Collecting detailed data to understand the role of adjuvant treatment may elucidate the differences in survival rates and should be a focus of future research. In particular, investigating the relationships between ER/PR status and the type(s) of treatment received may help us to understand whether adapting female adjuvant treatments to males is beneficial.

Although these data suggest differences in survival rates among men of different race/ethnicities, the differences may be due to factors reflected in race/ ethnicity, rather than genetics. For example, the differences in black–white survival rates in females are partially explained by socioeconomic differences. 25, 26 Research on females with breast carcinoma has suggested that cultural factors are important in treatment selection and adherence, which may in turn, influence survival rates. 34, 35 Unfortunately, our data cannot address differential access to or acceptance of adjuvant treatment. Treatment may be altered by comorbidity, and the prevalence of comorbidity is higher among blacks than among other racial/ ethnic groups. 36 The better survival rates among those of other race/ethnicity may reflect a “healthy immigrant” effect. For example, men who emigrate to the United States may be in better overall health and are better able to survive breast carcinoma. Future investigations exploring the socioeconomic status, comorbidity, and detailed treatment data may explain the racial/ethnic differences.

Our study used a large sample for a rare disease. The sample was obtained from a national population-based cancer surveillance program, thereby minimizing selection bias. The size of the sample allowed us to examine death from breast carcinoma, rather than being limited us to death from any cause, and to investigate variations by race. Use of the SEER historical stage allowed us to keep a consistent staging system over the 25 years. The major limitation of this study was a lack of consistent data on chemotherapy, hormone therapy, and socioeconomic status. Data on ER and PR were not collected until 1990, so we could only examine its impact in a subset of the population.

To summarize, men of other race/ethnicity have consistently better survival than blacks or whites, which is similar to the data reported for females. Blacks and whites have similar survival, except for regional stage disease. Stage is a strong factor, particularly in breast carcinoma-specific death. ER and PR status do not alter survival. Unfortunately, there has been little improvement in outcome over the last 25 years.