SEARCH

SEARCH BY CITATION

Keywords:

  • advanced breast cancer;
  • survival;
  • radiation;
  • disparities

Abstract

  1. Top of page
  2. Abstract
  3. METHODS AND MATERIALS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

BACKGROUND:

The authors previously identified racial/ethnic disparities in the use of radiation therapy (RT) in patients with advanced breast cancer (BC). They hypothesized that disparities in the use of RT were associated with survival differences favoring white patients.

METHODS:

The authors used the Surveillance, Epidemiology, and End Results database to identify white, black, Hispanic, and Asian patients with BC associated with ≥10 metastatic lymph nodes diagnosed between 1988 and 2005. Multivariate analyses of overall survival (OS) and disease-specific survival (DSS) assessed age, sex, race, tumor size, histology, estrogen receptor status, progesterone receptor status, RT, and type of surgery. The authors further stratified for use of RT and type of surgery. Risk of mortality was reported as hazard ratios (HRs) with 95% confidence intervals (CIs).

RESULTS:

Of 15,895 patients with advanced BC, 12,653 met entry criteria. On multivariate analysis, RT was associated with a decreased risk of all-cause (HR, 0.78; 95% CI 0.74-0.83; P < .001) and disease-specific (HR, 0.81; 95% CI, 0.76-0.86; P < .001) mortality; black race was associated with an increased risk of all-cause (HR, 1.54; 95% CI, 1.42-1.68; P < .001) and disease-specific (HR, 1.53; 95% CI, 1.39-1.68; P < .001) mortality. After stratifying by type of surgery and use of RT, blacks demonstrated poorer survival than their white counterparts, regardless of surgery type or receipt of RT.

CONCLUSIONS:

Only black patients had poorer OS and DSS relative to whites. When stratified by type of surgery and use of RT, blacks continued to demonstrate poorer survival. This survival disparity is unlikely to be because of lack of RT. Cancer 2012;. © 2011 American Cancer Society.

Radiation therapy (RT) is currently indicated for all patients with breast cancer (BC) undergoing breast conservation.1, 2 Current guidelines also advise RT after mastectomy for locally advanced primary tumors (≥5 cm) or extensive axillary nodal disease (≥4 lymph nodes).2-6 Postlumpectomy RT for early stage BC has demonstrated a benefit in decreasing rates of local recurrence, and postmastectomy RT improves rates of local recurrence and survival.2, 4, 6-8 RT is therefore indicated for the treatment of advanced BC, regardless of the primary surgical therapy provided. We previously identified lower rates of RT for blacks and Hispanics relative to their white counterparts in a cohort of patients with advanced, American Joint Committee on Cancer stage IIIC BC with ≥10 metastatic axillary lymph nodes.9 Given our previously documented disparities in use of RT, we hypothesized that black and Hispanic patients would have poorer overall survival (OS) and disease-specific survival (DSS) relative to whites.

METHODS AND MATERIALS

  1. Top of page
  2. Abstract
  3. METHODS AND MATERIALS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

The patient population used for this study is identical to 1 previously reported.9 The Surveillance, Epidemiology, and End Results (SEER) database of the National Cancer Institute was used to identify all white, black, Hispanic, and Asian patients with invasive ductal carcinoma (IDC), invasive lobular carcinoma (ILC), or mixed ductal/lobular carcinoma of the breast associated with ≥10 metastatic lymph nodes diagnosed between 1988 and 2005. We chose to examine only BC patients with ≥10 lymph node metastases, rather than specifically those with ≥4 lymph node metastases to obtain a more uniform, homogeneous population. The geographic scope of the current SEER registry has been reported previously.9, 10 SEER registries routinely collect data on patient demographics, primary tumor site, tumor morphology, stage at diagnosis, first course of treatment, and follow-up vital status. These data are de-identified and therefore exempt from institutional review board review.

To best define potential survival differences among racial/ethnic groups, we only included those racial/ethnic groups with significant patient numbers: white, black, Hispanic, and Asian. Further exclusions were made if the use of RT was unknown, or in cases where patients had distant metastases or the diagnosis of advanced BC was documented only by death certificate or autopsy. The final sample included 12,653 patients. We compared differences among racial/ethnic groups using chi-square testing for categorical variables and proportions.

Survival was calculated as the number of completed months between the date of diagnosis and whichever occurred first: date of death, date last known to be alive, or December 31, 2004. The endpoints for the present study were OS and DSS. Patients who were lost to follow-up or survived beyond December 31, 2004 were coded as censored observations. Similarly, for calculations of DSS, patients who died from nonbreast-related causes were coded as censored observations.

Patient, tumor, and treatment factors with known or potential prognostic importance were examined by univariate analysis for their influence on OS and DSS using the Kaplan-Meier method. Statistical differences among or between survival curves were assessed via the log-rank test. Variables subjected to univariate analysis included age, sex, race (white, black, Hispanic, Asian), tumor size, histology (IDC, ILC, mixed ductal/lobular carcinoma), estrogen receptor (ER) status (positive, negative, equivocal, unknown), progesterone receptor (PR) status (positive, negative, equivocal, unknown), RT (yes vs no), and type of surgery (mastectomy vs lumpectomy). Age and tumor size were treated as categorical variables in the univariate analyses using the median split technique. Patients were categorized as being at or younger than the median age (≤56 years) or older (>56 years). Similarly, patients were categorized as having tumors at or below the median size (≤32 mm) or larger (>32 mm).

We reasoned that patients would be more likely to receive RT if they received lumpectomy as surgical treatment, as RT is a recognized component of breast conservation therapy.11 Postmastectomy RT guidelines have been adopted more recently. We therefore constructed 2 additional logistic regression models stratifying patients according to the type of surgical therapy received (mastectomy vs lumpectomy). To address statistical interactions between the use of RT and type of surgery, we constructed multiple Cox proportional hazards models. Model A was an unstratified multivariate model of our entire population incorporating all variables assessed by univariate analysis, regardless of statistical significance. The covariates age and tumor size were analyzed as continuous variables in the multivariate model. Models B through E were stratified according to the type of surgery received (mastectomy vs lumpectomy) and whether RT was used. The goal of this stratification was to better assess the true role of race/ethnicity on OS and DSS while accounting for these interactions. Models B and C examined patients undergoing mastectomy with and without RT, respectively. Similarly, models D and E examined patients undergoing lumpectomy with and without RT, respectively. All analyses were conducted using STATA version 10 (StataCorp, College Station, Tex).

RESULTS

  1. Top of page
  2. Abstract
  3. METHODS AND MATERIALS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

Of 15,895 patients with advanced BC identified in the SEER registry, 12,653 met study entry criteria. The demographics of the study population have been reported previously.9 Patient characteristics, tumor factors, and treatments received according to race/ethnicity are included in Table 1. Significant differences were noted for all covariates among the different racial/ethnic groups.

Table 1. Patient Characteristics, Tumor Factors, and Treatment Received for the Study Population According to Race/Ethnicity
CharacteristicWhiteBlackHispanicAsianP
  1. Abbreviations: ER, estrogen receptor; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; MDLC, mixed ductal/lobular carcinoma; PR, progesterone receptor.

Age, y (%)    <.001
 <564072 (44.7)724 (55.5)844 (64.0)531 (57.2) 
 A≥565030 (55.3)581 (44.5)474 (36.0)397 (42.8) 
Sex, No. (%)    <.001
 Male80 (0.9)26 (2)6 (0.5)3 (0.3) 
 Female9022 (99.1)1279 (98)1312 (99.5)925 (99.7) 
Tumor size, mm (%)    <.001
 ≤324437 (52.2)561 (46.0)600 (49.1)405 (46.2) 
 >324059 (47.8)660 (54.0)621 (50.9)472 (53.8) 
Histology, No. (%)    <.001
 IDC6587 (72.4)1098 (84.1)1059 (80.4)779 (83.9) 
 ILC1536 (16.9)106 (8.1)131 (9.9)60 (6.5) 
 MDLC979 (10.8)101 (7.8)128 (9.7)89 (9.6) 
ER status, No. (%)    <.001
 Positive4734 (52.0)552 (42.3)637 (48.3)462 (49.8) 
 Negative1764 (19.4)350 (26.8)292 (22.2)252 (27.1) 
 Equivocal60 (0.6)10 (0.8)5 (0.4)8 (0.9) 
 Unknown2544 (28.0)393 (30.1)384 (29.1)206 (22.2) 
PR status, No. (%)    <.001
 Positive3851 (42.3)426 (32.6)532 (40.4)408 (44) 
 Negative2524 (27.7)439 (33.7)375 (28.4)304 (32.8) 
 Equivocal74 (0.8)14 (1.1)7 (0.5)3 (0.3) 
 Unknown2653 (29.2)426 (32.6)404 (30.7)213 (22.9) 
Surgery, No. (%)    .04
 Mastectomy7413 (81.4)1048 (80.3)1039 (78.8)775 (83.5) 
 Lumpectomy1666 (18.3)251 (19.2)276 (21)148 (16) 
 Unknown23 (0.3)6 (0.5)3 (0.2)5 (0.5) 
Radiation, No. (%)    <.001
 No radiation4407 (48.4)708 (54.2)677 (51.4)430 (46.3) 
 Radiation4695 (51.6)597 (45.8)641 (48.6)498 (53.7) 

The median age of patients was 56 years, with a range of 21 to 97 years. Less than 1% (n = 115) of our patients were men. The majority of patients were white (n = 9102; 72%). The remainder were Hispanic (n = 1318; 10.4%), black (n = 1305; 10.3%), and Asian (n = 928; 7.3%). Tumor histology was IDC, ILC, or mixed ductal/lobular carcinoma for 75.3%, 14.5%, and 10.2% of patients, respectively. ER and PR status were unknown for 27.9% and 29.2% of patients, respectively. ER status was positive, negative, or equivocal for 50.5%, 21%, and 0.6% of patients, respectively. Similarly, PR status was positive, negative, or equivocal for 41.2%, 28.8%, and 0.8% of patients, respectively. The majority of patients underwent mastectomy (81.2%) for primary BC treatment; 18.5% received lumpectomy, and the type of surgical procedure was unknown for 0.3%. Approximately half of the patients received radiation (50.8%), and half did not (49.2%).

Ten-year rates of OS and DSS for the entire study population were 32.9% and 42.3% (median survival 68 and 86 months, respectively) (Table 2). Patients receiving RT had a 10-year OS of 38.8% compared with 27.9% for those not receiving RT. DSS rates for those receiving and not receiving RT were 47.2% and 37.9%, respectively. Ten-year OS rates for white, black, Hispanic, and Asian patients were 33.9%, 21.2%, 35.8%, and 36.3%, respectively. Ten-year DSS rates for white, black, Hispanic, and Asian patients were 43.6%, 30.2%, 44.6%, and 43.2%, respectively. We further examined rates of 10-year survival according to both race/ethnicity and use of RT (Table 2). Use of RT made the largest contribution to OS and DSS in Asian and white patients. The addition of RT improved OS and DSS in Asians by 14.1% and 13.6%, respectively. Similarly, RT improved OS and DSS in whites by 11.5% and 10.1%, respectively.

Table 2. Ten-Year OS and DSS According to Race/Ethnicity and Use of RT
Group10-Year OS, %Δ, %10-Year DSS, %Δ, %
  1. Abbreviations: DSS, disease-specific survival; OS, overall survival; RT, radiation therapy.

  2. The effect of RT on 10-year survival rates is estimated by the incremental change in survival rate (Δ) with the addition of RT.

Total population32.9 42.3 
White + radiation40.111.548.910.1
White + no radiation28.6 38.8 
Black + radiation23.43.731.32.0
Black + no radiation19.7 29.3 
Hispanic + radiation39.67.446.23.3
Hispanic + no radiation32.2 42.9 
Asian + radiation43.414.150.013.6
Asian + no radiation29.3 36.4 

Age, tumor size, histology, ER and PR status, race/ethnicity, surgery, and RT were all significant factors influencing OS and DSS on univariate analysis at the P < .001 level. Although there were very few male patients in our population, sex was the only factor not significantly associated with OS or DSS.

A summary of the multivariate model of the overall patient population and those models stratifying patients according to type of surgery and use of RT are included in Tables 3 through 7. Model A (Table 3) included the entire population of 12,653 patients and was unstratified. In this model, black race was associated with increased rates of all-cause mortality (hazard ratio [HR], 1.54; 95% confidence interval [CI] 1.42-1.68; P < .001) and breast-specific mortality (HR, 1.53; 95% CI, 1.39-1.68; P < .001); use of RT was associated with a decreased risk of all-cause mortality (HR, 0.78; 95% CI, 0.74-0.83; P < .001) and breast-specific mortality (HR, 0.81; 95% CI, 0.76-0.86). Additional factors influencing OS and DSS are noted in Table 3.

Table 3. Cox Regression Models of Overall Survival and Disease-Specific Survival: Total Population (N=12,653)
CharacteristicOverall SurvivalDisease-Specific Survival
HR95% CIPHR95% CIP
  • Abbreviations: CI, confidence interval; ER, estrogen receptor; HR, hazard ratio; MDLC, mixed ductal/lobular carcinoma; PR, progesterone receptor; RT, radiation therapy.

  • a

    Statistically significant.

Age1.021.02-1.02<.001a1.011.01-1.01<.001a
Sex      
 Female
 Male1.110.85-1.44.500.950.68-1.33<.77
Tumor size1.001.00-1.00<.001a1.001.00<.001a
Histology      
 Ductal
 Lobular0.820.75-0.89<.001a0.780.71-0.86<.001a
 MDLC0.830.75-0.91<.001a0.830.74-0.92<.001a
ER      
 Positive
 Negative1.691.55-1.84<.001a1.811.65-1.99<.001a
 Equivocal1.280.96-1.71.101.491.10-2.03.011a
 Unknown1.631.29-2.06<.001a1.631.26-2.12<.001a
PR      
 Positive
 Negative1.291.19-1.41<.001a1.331.21-1.46<.001a
 Equivocal1.711.31-2.24<.001a1.911.43-2.55<.001a
 Unknown0.820.65-1.04<.100.890.69-1.15<.38
Race/ethnicity      
 White
 Black1.541.42-1.68<.001a1.531.39-1.68<.001a
 Hispanic1.010.91-1.12.800.980.88-1.10<.78
 Asian0.980.88-1.09.750.980.88-1.11<.80
Surgery      
 Mastectomy
 Lumpectomy0.800.74-0.87<.001a0.790.72-0.86<.001a
 Unknown1.350.93-1.96.151.490.99-2.23.051
RT      
 No
 Yes0.780.74-0.83<.001a0.810.76-0.86<.001a
Table 4. Cox Regression Models of Overall Survival and Disease-Specific Survival: Patients Who Received Mastectomy With Radiation Therapy (n=4914)
CharacteristicHR95% CIPHR95% CIP
  • Abbreviations: CI, confidence interval; ER, estrogen receptor; HR, hazard ratio; PR, progesterone receptor.

  • a

    Statistically significant.

Age1.021.02-1.02<.001a1.011.01-1.02<.001a
Sex      
 Female
 Male1.250.82-1.92.301.070.63-1.81<.81
Tumor size1.001.00-1.00<.001a1.001.00-1.01<.001a
Histology      
 Ductal
 Lobular0.770.67-0.88<.001a0.710.60-0.83<.001a
 Mixed0.770.65-0.92.004a0.760.63-0.93.007a
ER      
 Positive
 Negative1.631.42-1.88<.001a1.811.55-2.11<.001a
 Equivocal0.960.55-1.66.871.170.66-2.08<.60
 Unknown1.581.03-2.43.04a1.661.03-2.69<.04a
PR      
 Positive
 Negative1.301.14-1.49<.001a1.361.17-1.58<.001a
 Equivocal1.500.94-2.40<.091.671.00-2.78<.05a
 Unknown0.920.60-1.41<.710.960.60-1.54<.88
Race/ethnicity      
 White
 Black1.761.51-2.04<.001a1.751.49-2.05<.001a
 Hispanic1.070.91-1.27.451.050.88-1.26<.59
 Asian0.950.80-1.14.590.960.79-1.16<.69
Table 5. Cox Regression Models of Overall Survival and Disease-Specific Survival: Patients Who Received Mastectomy Without Radiation Therapy (n=5361)
CharacteristicHR95% CIPHR95% CIP
  • Abbreviations: CI, confidence interval; ER, estrogen receptor; HR, hazard ratio; PR, progesterone receptor.

  • a

    Statistically significant.

Age1.021.02-1.02<.001a1.011.00-1.01<.001a
Sex      
 Female
 Male1.040.73-1.46.850.950.62-1.47<.83
Tumor size1.001.00-1.00<.001a1.001.00-1.00<.001a
Histology      
 Ductal
 Lobular0.810.72-0.91<.001a0.790.69-0.90.001a
 Mixed0.820.71-0.94.004a0.810.70-0.95.01a
ER      
 Positive
 Negative1.651.46-1.87<.001a1.651.46-1.87<.001a
 Equivocal1.571.10-2.29.02a1.571.10-2.29.02a
 Unknown1.551.14-2.12.005a1.551.14-2.12.005a
PR      
 Positive
 Negative1.301.15-1.46<.001a1.291.13-1.48<.001a
 Equivocal1.841.25-2.70.0021.971.29-3.00.002
 Unknown0.800.592-1.10<.170.870.62-1.24<.45
Race/ethnicity      
 White
 Black1.421.27-1.60<.001a1.381.21-1.57<.001a
 Hispanic0.990.86-1.13.850.910.78-1.07<.27
 Asian0.970.84-1.13.750.980.83-1.16<.86
Table 6. Cox Regression Models of Overall Survival and Disease-Specific Survival: Patients Who Received Lumpectomy With Radiation Therapy (n=1504)
CharacteristicHR95% CIPHR95% CIP
  • Abbreviations: CI, confidence interval; ER, estrogen receptor; HR, hazard ratio; NA, not applicable due to low sample size; PR, progesterone receptor.

  • a

    Statistically significant.

Age1.021.01-1.03<0.001a1.011.00-1.02<.05a
Sex      
 Female
 Male1.220.17-8.85.85NANANA
Tumor size1.011.00-1.02<.001a1.011.00-1.02<.001a
Histology      
 Ductal
 Lobular0.970.69-1.35.900.930.63-1.37<.71
 Mixed1.070.74-1.54.751.110.74-1.68<.61
ER      
 Positive
 Negative1.911.46-2.51<.001a1.811.34-2.45<.001a
 Equivocal2.030.73-5.65.182.110.74-6.01<.17
 Unknown1.310.54-3.20.551.370.51-3.70<.54
PR      
 Positive
 Negative1.391.06-1.81.0151.551.15-2.08.004a
 Equivocal1.040.32-3.35.951.190.36-3.93<.78
 Unknown1.180.49-2.84<.711.270.48-3.38<.63
Race/ethnicity      
 White
 Black1.461.09-1.97.02a1.491.08-2.07<.02a
 Hispanic0.880.63-1.23.451.020.63-1.23.45
 Asian1.010.70-1.47.950.930.61-1.44<.76
Table 7. Cox Regression Models of Overall Survival and Disease-Specific Survival: Patients Who Received Lumpectomy Without Radiation Therapy (n=837)
CharacteristicHR95% CIPHR95% CIP
  • Abbreviations: CI, confidence interval; ER, estrogen receptor; HR, hazard ratio; NA, not applicable due to low sample size; PR, progesterone receptor.

  • a

    Statistically significant.

Age1.021.01-1.03<.001a1.011.00-1.03.005a
Sex      
 Female
 Male0.810.11-5.85.85NANANA
Tumor size1.011.00-1.02<.001a1.011.01-1.02<.001a
Histology      
 Ductal
 Lobular0.810.52-1.26.400.770.47-1.27<.31
 Mixed0.740.49-1.11.150.800.51-1.24.32
ER      
 Positive
 Negative1.751.24-2.46<.001a1.811.25-2.64.002a
 Equivocal0.810.28-2.33.701.000.35-2.90.99
 Unknown3.301.26-8.58.02a2.751.01-7.51<.05a
PR      
 Positive
 Negative1.250.89-1.75<.211.230.84-1.78<.29
 Equivocal1.630.64-4.18<.311.990.76-5.23.16
 Unknown0.440.17-1.11<.090.500.19-1.34.17
Race/ethnicity      
 White
 Black1.621.14-2.30.007a1.631.12-2.38.01a
 Hispanic1.070.71-1.60.801.050.68-1.62<.84
 Asian1.070.68-1.69.801.080.66-1.77.76

Additional stratified multivariate models were constructed. Model B included 4914 patients who underwent mastectomy and received RT (Table 4). Black patients had a 76% (HR, 1.76; 95% CI, 1.51-2.04; P < .001) and 75% (HR, 1.75; 95% CI, 1.49-2.05; P < .001) increased risk of all-cause and breast-specific mortality, respectively. Additional factors influencing OS and DSS are listed in Table 4.

Model C included 5361 patients who received mastectomy without RT (Table 5). Relative to whites, black patients had a 42% (HR, 1.42; 95% CI, 1.27-1.60; P < .001) and 38% increased risk of all-cause and breast-specific mortality, respectively. Additional factors influencing OS and DSS are listed in Table 5.

Model D included 1504 patients who received lumpectomy with RT (Table 6). Black patients demonstrated a 46% (HR, 1.46; 95% CI, 1.09-1.97; P = .02) and 49% (HR, 1.49; 95% CI, 1.08-2.67; P < .02) increased risk of death because of any cause and breast-related death, respectively. Additional factors influencing OS and DSS are listed in Table 6.

Model E included 837 patients who received lumpectomy without RT (Table 7). Relative to whites, black patients had a 62% (HR, 1.62; 95% CI, 1.14-2.30; P = .007) and 63% (HR, 1.63; 95% CI, 1.12-2.38; P = .01) increased risk of all-cause and breast-related mortality, respectively. Additional factors influencing OS and DSS are listed in Table 7.

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS AND MATERIALS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

We have previously reported on racial/ethnic disparities in the use of RT for patients with advanced BC, defined as those with ≥10 axillary lymph node metastases.9 Our current study was designed to test whether the previously noted disparities in treatment9 were significant enough to influence OS and DSS.9 To answer this question, we asked whether race/ethnicity itself was an important prognostic factor, then separately evaluated the prognostic importance of RT. On the basis of our prior data, we expected the most significant survival differences related to the use of RT to occur in black and Hispanic patients who underwent mastectomy.

When we analyzed our entire cohort in the multivariate model (Model A), both race/ethnicity and RT were significant prognostic factors. Being black was associated with a 54% increase, whereas the use of RT was associated with a 22% decrease in the risk of death from any cause. Predicted DSS rates paralleled the OS findings. Interestingly, use of lumpectomy was also associated with a decreased risk of mortality in our analyses. This could be attributable to the finding that patients who received lumpectomy had smaller tumors than those who received mastectomy or that lumpectomy patients were more likely to receive RT or adjuvant chemotherapy. To control for possible interaction between the type of surgery and use of RT, we stratified our models based on the type of surgery received and whether RT was used. This allowed us to assess more specifically the role of race/ethnicity on survival when controlling for the type of surgery and use of RT.

Our findings were unexpected. Results from our prior study led us to hypothesize that Hispanics, particularly those undergoing mastectomy without RT, would have poorer OS than their white counterparts. Hispanics did not demonstrate OS or DSS significantly different from their white counterparts. This is illustrative of the previously reported Hispanic paradox, whereby Hispanics demonstrate better than expected outcomes despite poorer access to optimal treatment, and high rates of socioeconomic stressors and comorbidities.12 Black patients, conversely, demonstrated poorer OS and DSS than their white counterparts whether treated with lumpectomy or mastectomy and irrespective of the use of RT. Risk of death from any cause for black patients was increased by 54% relative to their white counterparts. Breast-specific mortality was increased by 53% in black patients relative to their white counterparts. Unexpectedly, black patients undergoing mastectomy without RT demonstrated a lower hazard ratio for death than black patients undergoing mastectomy with RT. One possible explanation for this is that black patients receiving postmastectomy RT may be more predisposed to potentially life-threatening complications of this treatment, such as stroke, cardiac disease, and pulmonary toxicity.13 A similar pattern was not seen for black patients who underwent lumpectomy, however. If the racial/ethnic disparities in the use of RT truly influenced survival, we would anticipate that the poor OS in black and Hispanic patients in the overall cohort (Model A) would be ameliorated once these patients were stratified according to the receipt of RT.

Table 2, which details the 10-year survival rates of patients dependent on race/ethnicity and use of RT, showed that black patients had the smallest differential OS and DSS rates between patients receiving and not receiving RT (Δ = 3.7% and 2%, respectively). We may then conservatively assume that black patients who did not receive RT could at best receive a 3.7% OS and 2% DSS benefit with the addition of RT. By contrast, the differential OS rates for white, Hispanic, and Asian patients were 11.5%, 7.4%, and 14.1%, respectively; the differential DSS rates were 10.1%, 3.3%, and 13.6%, respectively, indicating that the addition of RT to these subpopulations would add a greater survival benefit than to black patients. Black patients seem less responsive to RT than their nonblack counterparts. There are several possible explanations for these findings. Although it is possible that black patients have a biologically different form of breast cancer that is less responsive to RT, it is also possible that black patients received less systemic therapy than their nonblack counterparts, regardless of their RT status. Griggs et al have previously documented that black women are less likely than their white counterparts to receive standard systemic therapy,14 and these findings are even more pronounced if they are obese.15

Poorer OS rates in black patients may be because of other comorbidities.13 Tammemagi et al followed a cohort of 906 patients (264 black and 642 white) in Detroit, Michigan older than 10 years and demonstrated a worse rate of OS (HR, 1.34; 95% CI, 1.11-1.62), BC-specific survival (HR, 1.47; 95% CI, 1.08-2.00), and competing cause survival (HR, 1.27; 95% CI, 1.00-1.63) in black patients. After adjusting for comorbidities, the competing cause survival HR for black versus white patients was not significant (HR, 1.06; 95% CI, 0.83-1.36). Another explanation is that black patients may be experiencing more adverse side effects of RT than their white counterparts.16 Unfortunately, the SEER data does not allow us insight into individual patient comorbidities. We were, however, able to provide insight into rates of DSS, which paralleled OS.

The shortcomings of using registry data like that of the SEER are well known.17 Although we attempted to determine whether disparities in the use of RT are associated with disparities in survival, we know that a cause-effect relationship cannot be ascertained from this analysis. Although using SEER data allowed us to capture a large cohort of patients and analyze real world treatment and survival scenarios, all factors in the decision-making process for these patients cannot be clarified in this database. For example, comorbidities, which may be a relative contraindication for RT for some patients, are not documented in SEER. In addition, socioeconomic status (SES) cannot be determined by SEER. Others have accounted for SES by determining the median income of a patient's resident Zip Code through SEER-Medicare linkage. Although this may have been an important factor in our analysis, it is an aggregate measure that may not accurately reflect an individual's SES. Second, this method would require us to eliminate patients younger than the age of 65 years and would significantly decrease our cohort size. Our study is further complicated because we make no direct measurement of the use of systemic therapy among our study patients. SEER data do not include information on use of hormonal therapy or chemotherapy, whether given in the adjuvant or neoadjuvant setting. Theoretically, the survival differences noted in the current study could be attributable to differential use of systemic therapy among these racial/ethnic populations. Prior studies have suggested that when racial/ethnic minority populations receive treatment equivalent to their white counterparts, equivalent survival may be anticipated.18 Whereas ER and PR status was known for the majority of our patients, HER-2/neu overexpression status is not recorded in the SEER database and was unknown. Finally, we have reported on the use of RT as an essentially binary function: RT was received, yes or no. In reality, the quantity and quality of RT received may also be important, as it pertains to the overall dose given, anatomic distribution, and use of boosts. The dose of RT received and the fields used are unknown; these data are not available in SEER.

Conclusions

Despite a decreased utility of RT in black and Hispanic patients described previously,9 only black patients had a poorer survival relative to white patients. In the black population, this survival disadvantage persisted even after stratifying for type of surgery and RT use, indicating black race as an independent variable of survival in this population apart from RT use disparities. This analysis opposes the hypothesis that RT disparities might account for poorer rates of OS and DSS in black and Hispanic BC patients. Future interventions to address BC survival disparities should focus on the black population.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. METHODS AND MATERIALS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

Supported by grant number UL1 RR024146 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NCRR or NIH.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS AND MATERIALS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES
  • 1
    NIH consensus conference. Treatment of early-stage breast cancer. JAMA. 1991; 265: 391-395.
  • 2
    Clarke M, Collins R, Darby S, et al. Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomised trials. Lancet. 2005; 366: 2087-2106.
  • 3
    Eifel P, Axelson JA, Costa J, et al. National Institutes of Health Consensus Development Conference Statement: adjuvant therapy for breast cancer, November 1-3, 2000. J Natl Cancer Inst. 2001; 93: 979-989.
  • 4
    Overgaard M, Hansen PS, Overgaard J, et al. Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive adjuvant chemotherapy. Danish Breast Cancer Cooperative Group 82b Trial. N Engl J Med. 1997; 337: 949-955.
  • 5
    Recht A, Edge SB, Solin LJ, et al. Postmastectomy radiotherapy: clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol. 2001; 19: 1539-1569.
  • 6
    Ragaz J, Jackson SM, Le N, et al. Adjuvant radiotherapy and chemotherapy in node-positive premenopausal women with breast cancer. N Engl J Med. 1997; 337: 956-962.
  • 7
    Overgaard M, Jensen MB, Overgaard J, et al. Postoperative radiotherapy in high-risk postmenopausal breast-cancer patients given adjuvant tamoxifen: Danish Breast Cancer Cooperative Group DBCG 82c randomised trial. Lancet. 1999; 353: 1641-1648.
  • 8
    Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a randomized trial comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment of invasive breast cancer. N Engl J Med. 2002; 347: 1233-1241.
  • 9
    Martinez SR, Beal SH, Chen SL, et al. Disparities in the use of radiation therapy in patients with local-regionally advanced breast cancer. Int J Radiat Oncol Biol Phys. 2010; 78: 787-792.
  • 10
    Chen SL, Martinez SR. The survival impact of the choice of surgical procedure after ipsilateral breast cancer recurrence. Am J Surg. 2008; 196: 495-499.
  • 11
    Jagsi R, Abrahamse P, Morrow M, et al. Patterns and correlates of adjuvant radiotherapy receipt after lumpectomy and after mastectomy for breast cancer. J Clin Oncol. 2010; 28: 2396-2403.
  • 12
    Franzini L, Ribble J, Keddie A. Understanding the Hispanic paradox. Ethn Dis. 2001; 11: 496-518.
  • 13
    Tammemagi C, Nerenz D, Neslund-Dudas C, Feldkamp C, Nathanson D. Comorbidity and survival disparities among black and white patients with breast cancer. JAMA. 2005; 294: 1765-1772.
  • 14
    Griggs JJ, Culakova E, Sorbero ME, et al. Social and racial differences in selection of breast cancer adjuvant chemotherapy regimens. J Clin Oncol. 2007; 25: 2522-2527.
  • 15
    Griggs JJ, Sorbero ME, Lyman GH. Undertreatment of obese women receiving breast cancer chemotherapy. Arch Intern Med. 2005; 165: 1267-1273.
  • 16
    Ryan J, Bole C, Hickok J, et al. Post-treatment skin reactions reported by cancer patients differ by race, not by treatment or expectations. Br J Cancer. 2007; 97: 14-21.
  • 17
    Nathan H, Pawlik TM. Limitations of claims and registry data in surgical oncology research. Ann Surg Oncol. 2008; 15: 415-423.
  • 18
    Du W, Simon MS. Racial disparities in treatment and survival of women with stage I-III breast cancer at a large academic medical center in metropolitan Detroit. Breast Cancer Res Treat. 2005; 91: 243-248.