SEARCH

SEARCH BY CITATION

Keywords:

  • ductal carcinoma in situ;
  • breast cancer;
  • ethnicity;
  • race;
  • health care disparity;
  • recurrence

Abstract

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

BACKGROUND:

The impact of race and ethnicity on the biologic features and outcome variables of women who are diagnosed with preinvasive breast cancer—ductal carcinoma in situ (DCIS)—has not been addressed widely in the published literature.

METHODS:

Patient demographic, clinical, and pathologic features and outcome variables were analyzed with respect to the patient's initial self-reported race/ethnicity among women who received treatment for a diagnosis of pure DCIS from 1996 to 2009.

RESULTS:

Of 1902 patients, 1411 were white (74.2%), 214 were African American (11.3%), 175 were Hispanic (9.1%), and 102 were Asian/Pacific Islander (5.4%). The majority of patients were between ages 41 and 70 years (83%). Patients of Hispanic and Asian/Pacific Islander descent were significantly younger than white and African American patients (P < .001). DCIS size and grade, the presence of necrosis, and the frequency of breast-conserving surgery did not differ significantly between groups. African American patients aged >70 years and Hispanic patients aged <50 years were significantly more likely to have estrogen receptor-positive DCIS than patients of other races in the same age categories (P < .001). Adjuvant radiotherapy and tamoxifen were received significantly less often by white women (P < .001). At a median follow-up of 4.8 years (range, 1-14 years), recurrence rates and the development of contralateral breast cancer did not differ significantly among racial/ethnic groups when stratified by treatments received.

CONCLUSIONS:

There was variation in age at presentation, biologic features, and treatment of DCIS among the different ethnic groups. Additional studies with larger numbers of ethnic minority patients are needed to confirm whether the consistent application of evidence-based treatment practices presents an opportunity for reducing disparities in patients with DCIS. Cancer 2013. © 2012 American Cancer Society.

Ductal carcinoma in situ (DCIS) itself does not affect survival but identifies a group of women—more than 60,000 women per year in the United States—who are at increased risk for developing invasive breast cancer.1 There is unambiguous evidence that African American women with invasive breast cancer present with biologically more aggressive disease and have an increased hazard of mortality from the disease.2-4 However, only a few population-based and small clinical, single-institutional, correlative studies have compared differences in prognosis between African American women and white women with DCIS.5-11 More so, there is a paucity of data involving potential differences in DCIS outcome among Hispanic women and Asian/Pacific Islander women.12-14 The objective of this study was to address potential racial and ethnic disparities among women with DCIS by identifying significant differences in presentation, biologic features, treatments, and outcomes in a large, multiethnic cohort of patients who were treated at The University of Texas M. D. Anderson Cancer Center according to multidisciplinary, evidence-based treatment guidelines.

MATERIALS AND METHODS

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

After we obtained approval from the Institutional Review Board of The University of Texas M. D. Anderson Cancer Center, we used the M. D. Anderson Breast Cancer Management System (BCMS) database to identify patients who had a diagnosis of pure DCIS and underwent surgery between January 1996 and July 2009. We excluded patients who did not report their race (n = 22) and patients who had not reached at least 1 year of follow-up (n = 231). The final study analysis consisted of 1902 patients. The BCMS database contains detailed information on patient characteristics (race/ethnicity, age, menopausal status, breast and ovarian cancer family history), tumor characteristics (clinical and pathologic data, including estrogen receptor [ER] status in patients beginning in 2003, and nuclear grade), endocrine treatment, surgery type (segmental mastectomy, mastectomy, reconstruction), radiotherapy, tamoxifen use, recurrence status, and survival. A variable was categorized as unknown if it was not recorded in the imaging or pathology report, such as mammographic dimension (n = 1,042), nuclear grade (n = 40), and pathologic size (n = 479).

Follow-up information for patients with DCIS is updated every 6 months by direct review of the medical records and linkage to the M. D. Anderson Tumor Registry, which also mails annual follow-up letters to each patient registered and known to be alive to determine their clinical status. The Tumor Registry also checks the Social Security Death Index and the Texas Bureau of Vital Statistics for the status of patients who do not respond to the letters. There were 212 patients lost to follow-up. The racial distribution of the patients who were lost to follow-up was similar to that of the entire study population (white, 72.2%; African American, 13.2%; Hispanic, 8.4%; Asian Pacific Islander, 6.2%).

Statistical Analyses

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

Patient demographic, clinical, and histopathologic features and outcome variables were analyzed with respect to the patient's initial self-reported race/ethnicity. The chi-square test was used to compare groups with respect to categoric variables. The Wilcoxon rank-sum test and the Kruskal-Wallis test were used to examine associations between categoric and ordinal variables. Locoregional recurrence and the development of contralateral breast cancer were determined using the Kaplan-Meier product-limit method and were compared between racial/ethnic groups with the log-rank test. The time to breast cancer event (locoregional recurrence, contralateral breast cancer development, or distant metastases) was calculated from the date of surgery to the date of the first breast cancer event. Patients who were not known to have an event at the date of last contact or who had not experienced a recurrence or death were censored. We used multivariable Cox proportional hazards regression modeling to estimate the hazard ratio (HR) and 95% confidence interval (CI) of epidemiological, clinical, and pathologic characteristics associated with having a breast cancer event. An interaction between race/ethnicity and family history was explored within the context of the multivariable model that contained the interaction term and the main effects terms. All reported P values are 2-sided, and P < .05 was considered statistically significant. Analyses were performed using STATA/IC for Windows (release 11.1; Stata Corporation, College Station, Tex) and STATISTICA (release 9.0; StatSoft, Inc., Tulsa, Okla).

RESULTS

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

Clinical and Pathologic Characteristics of Ductal Carcinoma In Situ According to Race

The study cohort consisted of 1902 patients who received treatment for DCIS at our institution between 1996 and 2001 (24%), between 2001 and 2005 (44%), and between 2006 and 2009 (32%). Overall, 1411 patients were white (74.2%), 214 were African American (11.3%), 175 were Hispanic (9.1%), and 102 were Asian/Pacific Islander (5.4%), as indicated in Table 1. The majority of patients were between ages 41 and 70 years (83%). Patients of Hispanic and Asian/Pacific Islander descent were significantly younger (median age, 50 years and 52 years, respectively) than white and African American patients (median age, 55 years and 56 years, respectively; P < .001) (Table 1). The majority of patients were postmenopausal (71.7%), and white patients were more likely than patients of other ethnicities to report having received hormone-replacement therapy (33.3%). A family history of breast and/or ovarian cancer was significantly more common in whites and African Americans than in Hispanics and Asians/Pacific Islanders (P < .001) (Table 1).

Table 1. Patient and Ductal Carcinoma In Situ Characteristics Grouped According to Race: Univariate Analyses (N = 1902)
 No. of Patients (%) 
CharacteristicWhiteAfrican AmericanHispanicAsian/Pacific IslanderPa
  • Abbreviations: ER, estrogen receptor; HRT, hormone-replacement therapy.

  • a

    P values were calculated based on a comparison across all groups.

  • b

    Menopausal status was not reported in 4 patients; 2 patients were perimenopausal.

  • c

    These P values are statistically significant.

  • d

    These P values were calculated using the Kruskal-Wallis test.

  • e

    Initial presenting signs were not available for 57 patients.

  • f

    Mammographic dimensions were not available for 1042 patients.

  • g

    Pathologic size was not available for 479 patients.

  • h

    Nuclear grade was not available for 40 patients.

  • i

    ER status was not assessed in 798 patients.

Total1411 (74.2)214 (11.3)175 (9.2)102 (5.4) 
Menopausal statusb     
Premenopausal375 (26.7)58 (27.1)60 (34.7)38 (37.3) 
Postmenopausal1032 (73.4)156 (72.9)113 (65.3)64 (62.8).023c
HRT, ever used470 (33.3)48 (22.4)36 (20.6)26 (25.5)< .001c
Age, y     
Median [range]55 [22-89]56 [26-88]50 [18-79]52 [33-82]<.001c,d
<50451 (32)68 (31.8)78 (44.6)40 (39) 
50-70798 (56.6)119 (55.6)90 (51.4)57 (56) 
>70162 (11.5)27 (12.6)7 (4)5 (4.9)< .001c
Any family history of breast or ovarian cancer703 (49.8)97 (45.3)60 (34.3)31 (30.4)< .001c
Presence of bilateral breast cancer at diagnosis105 (7.4)14 (6.5)10 (5.7)8 (7.8).825
Initial presenting signse     
Clinical183 (13.4)36 (17.4)29 (17.1)17 (17.2) 
Radiologic1186 (86.6)171 (82.6)141 (82.9)82 (82.8).232
Largest recorded mammographic dimension, n = 860     
Median [range], cmf2.0 [0.01-20]2.5 [0.2-17]2.0 [0.3-13]1.7 [0.2-10.1].143d
Largest recorded pathologic size, n = 1423     
Median [range], cmg1.2 [0.01-18]1.5 [0.1-19]1.5 [0.1-11]1.3 [0.09-10].007c,d
Nuclear gradeh     
1127 (9.2)26 (12.4)17 (9.9)15 (14.7) 
2 or 31252 (90.8)183 (87.6)155 (90.1)87 (85.3).181
Presence of necrosis549 (38.9)72 (33.6)70 (40)41 (40.2).472
ER status: Age group, yi     
Positive619 (78)120 (87)87 (82.9)54 (80.6)< .001
<50196 (31.7)38 (31.7)44 (50.6)21 (38.9) 
50-70365 (59)58 (48.3)40 (46)30 (55.6) 
>7058 (9.4)24 (20)3 (3.4)3 (5.6) 
Negative175 (22)18 (13)18 (17.1)13 (19.4).245
<5047 (26.9)3 (16.6)4 (22.2)2 (15.4) 
50-70101 (57.7)14 (77.8)14 (77.8)10 (76.9) 
>7027 (15.4)1 (5.65)0 (0)1 (7.7) 

The majority of patients presented with an abnormal breast imaging finding as opposed to a clinical symptom (83.1%) (Table 1), and the method of detection did not differ significantly based on race. The greatest median mammographic dimension, which was recorded in African Americans, was 2.5 cm and did not differ significantly from the dimension in whites (2.0 cm), Hispanics (2.0 cm), and Asians/Pacific Islanders (1.7 cm). Although the median pathologic size of tumors in African Americans was significantly greater (1.5 cm) than in whites (1.2 cm) and Asians/Pacific Islanders (1.3 cm), the clinical relevance of this finding probably would not have an impact on therapy or outcome, because the actual numerical difference in size between groups was only 2 or 3 mm (P = .007) (Table 1).

There was no overall significant difference in the presence of necrosis, nuclear grade of DCIS, or ER status among the different races (Table 1). African American, Hispanic, and Asian/Pacific Islander patients were more likely to have ER-positive tumors compared with white patients, but the difference did not reach significance (P = .084). There were significant differences in ER-positive DCIS when patients were stratified according to race and age (P < .001). For example, African American patients aged >70 years and Hispanic patients aged <50 were significantly more likely to have ER-positive DCIS than patients of other races in the same age categories.

Treatment Characteristics of Ductal Carcinoma In Situ Associated With Race

The different types of treatments received by our patient population are summarized in Table 2. There was no significant difference between races/ethnicities in the type of surgery received (P = .626). Among patients who underwent mastectomy, the rate of immediate breast reconstruction was significantly higher in whites (67.4%) than in the other ethnicities (P = .017). The use of contralateral mastectomy as prophylactic therapy among patients with DCIS was infrequent among all ethnicities (6.4%), yet it was significantly more frequent in white patients (8.7%) than in African American (4.2%), Hispanic (6.3%), and Asian/Pacific Islander (2%) patients (P = .013) (Table 2). Overall, among patients who had ER status determined and had ER-positive DCIS (n = 880), 422 patients (48%) received tamoxifen. The receipt of tamoxifen among patients who had ER-positive DCIS was significantly lower among whites than among other ethnicities (white, 43.8%; African American, 53.3%; Hispanic, 62.1%; Asian/Pacific Islander, 61.1%; P = .001). Among patients who underwent breast-conserving therapy (BCT), tamoxifen and/or adjuvant radiotherapy also were received significantly less often in white patients than in patients of other ethnic groups (P < .001) (Table 2).

Table 2. Treatment Characteristics Grouped According to Race: Univariate Analyses (N = 1902)
 No. of Patients (%) 
TreatmentWhiteAfrican AmericanHispanicAsian/Pacific IslanderP for All Groups
  • Abbreviations: BCT, breast-conserving therapy; RT, radiotherapy.

  • a

    These P values were statistically significant.

Surgery     
BCT: Lumpectomy827 (58.6)134 (62.6)108 (61.7)62 (60.8) 
Mastectomy584 (41.4)80 (37.4)67 (38.3)40 (39.2).626
Immediate breast reconstruction in patients who underwent mastectomy399 (67.4)43 (53.8)41 (61.2)20 (50).017a
Contralateral prophylactic mastectomy122 (8.7)9 (4.2)11 (6.3)2 (2).013a
Adjuvant tamoxifen in patients who underwent BCT     
Total324 (39.2)68 (50.8)53 (49.1)39 (62.9)< .001a
Age group, y     
<5096 (29.6)13 (19.1)26 (49.1)20 (51.3) 
50-70200 (61.7)48 (70.6)26 (49)18 (46.2) 
>7028 (8.6)7 (10.3)1 (1.9)1 (2.6)< .001a
Adjuvant RT in patients who underwent BCT     
Total638 (77.2)121 (90.3)91 (84.3)49 (79).003a
Age group, y     
<50176 (27.6)31 (25.6)38 (41.8)22 (44.9) 
50-70387 (60.7)76 (62.8)50 (55)24 (49) 
>7075 (11.7)14 (11.6)3 (3.2)3 (6.1).006a
Adjuvant tamoxifen and adjuvant RT in patients who underwent BCT275 (66.3)66 (85.7)45 (83.3)34 (81)< .001a

Outcomes of Patients With Ductal Carcinoma In Situ According to Race

Overall locoregional recurrence and the development of contralateral breast cancer are summarized in Table 3. The median follow-up for the entire population was 4.8 years (range, 1.0-14.0 years). In total, 40 of 1131 patients who underwent BCT had a locoregional recurrence (DCIS recurrence, 67.5%; ipsilateral invasive carcinoma recurrence, 32.5%), and 6 of 771 patients who underwent mastectomy had a locoregional recurrence (DCIS recurrence, 16.7%; invasive disease, 83.3%). Sixty-five patients (3%) developed a contralateral breast cancer (DCIS, 47.7%; invasive breast cancer, 52.3%) during the study period. The overall 5-year rates of locoregional recurrence among patients who underwent BCT and mastectomy were 3.5% and 1.1%, respectively. The overall 5-year rate of contralateral breast cancer development was 3.8%.

Table 3. Overall Locoregional Recurrence and Development of Contralateral Breast Cancer According to Race
 No. of Patients (%) 
 WhiteAfrican AmericanHispanicAsian/Pacific IslanderP for All Groupsa
  • Abbreviations: BCT, breast-conserving therapy; DCIS, ductal carcinoma in situ;

  • a

    P values were calculated using the Kaplan-Meier method; differences were compared using the log-rank test.

Locoregional recurrence in patients who underwent BCT     
Total30 (3.6)3 (2.2)5 (4.6)2 (3.2).775
DCIS19 (63.3)2 (66.7)4 (80)2 (100) 
Invasive11 (36.7)1 (33.3)1 (20)0 (0) 
Locoregional recurrence in patients who underwent mastectomy     
Total6 (1)0 (0)0 (0)0 (0).601
DCIS1 (16.7)0 (0)0 (0)0 (0) 
Invasive5 (83.3)0 (0)0 (0)0 (0) 
Development of contralateral breast cancer     
Total51 (4.5)8 (4.3)4 (2.7)2 (2.3).725
DCIS25 (49)2 (25)3 (75)1 (50) 
Invasive26 (51)6 (75)1 (25)1 (50) 

In total, 111 breast cancer events were detected during follow-up. Table 4 provides the 5-year rates of any breast cancer event (the development of locoregional recurrence, contralateral breast cancer, or distant metastasis) according to ethnicity and type of surgery. There were no significant differences in the overall or 5-year rates of these events based on race and stratified by receipt of radiation or tamoxifen (Tables 3, 4). Two white patients (0.001%) with an initial diagnosis of pure DCIS developed distant metastases during the study period and died of disease.

Table 4. Five-Year Rates of Breast Cancer Events by Ethnicity and Treatments Receiveda
 Percentage of Patients (No.) 
TreatmentWhiteAfrican AmericanHispanicAsian/Pacific IslanderP for All Groupsb
  • Abbreviations: BCT, breast-conserving therapy; RT, radiotherapy.

  • a

    Breast cancer events at 5 years included locoregional recurrence (n = 28) and the development of contralateral breast cancer (n = 48) or distant metastasis (n = 2). This analysis included only patients who had pure ductal carcinoma in situ at diagnosis and had no history of contralateral, invasive breast cancer.

  • b

    P values were calculated using the Kaplan-Meier method; differences were compared using the log-rank test.

BCT and RT, n = 8434.4 (20)6.2 (6)2.8 (2)5 (2).687
BCT and RT without tamoxifen, n = 4545.8 (15)3.9 (2)5.2 (2)7.7 (1).984
BCT and RT with tamoxifen, n = 3892.3 (5)8.4 (4)0 (0)3.5 (1).189
BCT without RT, n = 20310.5 (14)10 (1)20.9 (2)9.1 (1).959
BCT without RT or tamoxifen, n = 15612 (13)11.1 (1)014.3 (1).823
BCT without RT with tamoxifen, n = 474.8 (1)035.7(2)0.214
Mastectomy, n = 6435.5 (22)3 (2)5 (2)0 (0).569
Mastectomy with tamoxifen, n = 1434 (3)0 (0)6.7 (1)0 (0).819
Mastectomy without tamoxifen, n = 5005.9 (19)4 (2)2.9 (1)0 (0).601

In the adjusted multivariable analysis, significant independent predictors of having a breast cancer event were larger DCIS size (HR, 1.79; P = .030), BCT compared with mastectomy (HR, 2.47; P = .004), not receiving radiotherapy (HR, 2.18; P = .005), and having a family history of breast and/or ovarian cancer (HR, 1.64; P = .018). Race, age, ER status, nuclear grade, and receipt of tamoxifen were not statistically significantly associated with having a breast cancer event (Table 5). There was no significant interaction between ethnicity and family history of breast cancer and the risk of having a breast cancer event.

Table 5. Multivariate Cox Proportional Hazards Analysis of Clinicopathologic and Treatment Factors Associated With Breast Cancer Events
VariableHR (95% CI)P
  • Abbreviations: BCT, breast-conserving therapy; CI, confidence interval; ER, estrogen receptor; HR, hazard ratio; HRT, hormone-replacement therapy; RT, radiotherapy.

  • a

    These P values are statistically significant.

Race  
Caucasian1.00 (Reference) 
African American1.12 (0.59-2.14).733
Hispanic0.97 (0.44-2.13).937
Asian/Pacific Islander0.89 (0.32-2.48).830
Age, y  
<501.00 (Reference) 
≥500.88 (0.47-1.65).686
Menopausal status  
Premenopausal1.00 (Reference) 
Postmenopausal0.68 (0.35-1.29).238
HRT  
No1.00 (Reference) 
Yes1.23 (0.75-2.01).417
Family history of breast and/or ovarian cancer  
No1.00 (Reference) 
Yes1.64 (1.09-2.48).018a
Initial presenting signs  
Clinical1.00 (Reference) 
Radiologic1.08 (0.61-1.93).795
Pathologic size, cm  
<1.51.00 (Reference) 
≥1.51.79 (1.06-3.03).030a
Nuclear grade  
11.00 (Reference) 
2 or 31.21 (0.54-2.68).646
ER status  
Positive1.00 (Reference) 
Negative1.37 (0.69-2.71).364
Surgery  
Mastectomy1.00 (Reference) 
BCT2.47 (2.23-7.01).004a
Adjuvant tamoxifen  
Yes1.00 (Reference) 
No1.59 (0.96-2.63).071
Adjuvant RT  
Yes1.00 (Reference) 
No2.18 (1.26-3.75).005a

DISCUSSION

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

In a large, multiethnic cohort of patients, we evaluated the presentation, pathologic features, treatment, and outcome of DCIS. There was no significant difference between ethnic groups in the type of surgery performed (BCT vs mastectomy), although white patients were more likely to undergo immediate breast reconstruction. It is noteworthy that, when patients were stratified according to age at diagnosis, there were significant differences in the ER status of DCIS by ethnicity, and African American, Hispanic, and Asian/Pacific Islander patients were more likely than white patients to receive adjuvant therapy with tamoxifen and radiotherapy. In multivariable analysis, race was not a statistically significant, independent predictor of having a breast cancer event (locoregional recurrence, contralateral breast cancer, or distant metastasis). However, the small number of breast cancer events in the ethnic minority groups and the relatively short median follow-up limits a definitive conclusion that small disparities do not exist among patients who receive similar treatments for DCIS.

Population-based studies have demonstrated a significant decrease in the proportion of patients who undergo mastectomy for DCIS since the 1990s, with the result that BCT is the most common surgery performed in the United States.15 Although we did not observe a statistically significant difference in the rate of mastectomy between ethnic groups, white patients (41.4%) were more likely to undergo a mastectomy than African American (37.4%), Hispanic (38.3%), and Asian/Pacific Islander (39.3%) patients, and this finding is consistent with prior studies.12, 14 Among patients who undergo mastectomy for DCIS, immediate reconstruction often is considered to have psychological beneficial; however, its use is not widespread, and the rates range from 13.5% to 51%.16-18 In a population of patients who received treatment for DCIS from 2003 to 2007 in 4 Southern California counties, the overall rate of immediate reconstruction was 51%, and white (49.2%) and African American (54.6%) patients had statistically significantly higher rates of breast reconstruction than Hispanic (18.4%) and Asians/Pacific Islander (18.8%) patients.12 The overall rate of immediate breast reconstruction in our patient population was 65.2%, and white (67.4%) and Hispanic (61.2%) patients had the highest rates of immediate reconstruction. We hypothesize that the reasons for the higher use of immediate breast reconstruction in a tertiary care center may include younger age at diagnosis and increased access to plastic surgery expertise.

The percentages of patients with a family history of breast or ovarian carcinoma were significantly higher in white and African American women compared with Hispanic and Asian women in this study. Referral bias likely plays a role in the high proportion of white (49.8%) and African American (45.3%) women who reported a family history of breast or ovarian cancer, because those who are referred to The University of Texas M. D. Anderson Cancer Center are more likely to be high-risk patients. In addition, the incidence of breast cancer in the United States is lowest among Hispanic and Asian women; therefore, these women were less likely to report a family history.19

The current study, to our knowledge, is the first to examine ER receptor status in detail among different racial and ethnic groups. It is well established that African American women are more likely than white women to develop ER-negative, invasive breast cancers, which are associated with a poorer prognosis.4, 5 Therefore, we were interested in determining whether DCIS, as a nonobligate precursor of invasive breast cancer, also exhibits the same ER phenotype of invasive breast cancer in African American women. It is noteworthy that, in the current study, African American women with DCIS exhibited the lowest rate of ER-negative DCIS (13%) among all racial/ethnic groups (Table 1). This finding implies either that DCIS loses ER expression during invasive carcinogenesis or that the invasive carcinoma may originate from a different progenitor cell in African American women. In our study, Hispanic and Asian/Pacific Islander patients also exhibited ER-positive DCIS more frequently than white patients. When stratified by age at diagnosis, patients ages 50 to 70 years had the highest proportion of ER-positive DCIS, which correlates with the national guidelines for the age group recommended for screening mammography.20 Indeed, 89% of our study population ages 50 to 70 years had their DCIS detected through screening mammograms.

A noteworthy finding of our analysis was that the receipt of adjuvant tamoxifen and radiotherapy separately or combined was significantly greater in all ethnicities than in whites. In another hospital-based study, Nassar et al6 reported that white patients (31%) were more likely to receive tamoxifen than African American patients (18%) and were also more likely than African American and Asian patients to receive radiation after BCT. Joslyn21 analyzed women with DCIS in the 1973 to 2000 Surveillance, Epidemiology, and End Results database and observed that African American women were less likely to receive adjuvant radiation therapy than white and Asian women. The increased receipt of adjuvant tamoxifen and radiation by ethnic minority patients in our academic cancer center may reflect a healthier patient population and increased patient and physician awareness and motivation to comply with evidence-based recommendations for breast cancer risk reduction. Indeed, consulting an oncologist for DCIS follow-up care was associated with a 5-fold increase in the receipt of adjuvant endocrine therapy in a recent population-based study of English-speaking or Spanish-speaking Latina and non-Latina white women.13

The treatment goals for DCIS are primarily to prevent DCIS recurrences and to prevent the development of invasive breast cancer. Published studies using the Surveillance, Epidemiology, and End Results database on the outcomes of DCIS reported higher all-cause and breast cancer mortality among African American women than among white woman.5, 6, 21, 22 However, a close analysis of some of those studies demonstrates that the overall chance of death from DCIS is <1%, and the absolute difference in the risk of death between African American women and white women also was <1%.5 We observed no significant differences in the overall or 5-year rates of developing locoregional recurrence, contralateral breast cancer, or distant metastasis based on ethnicity. Similarly, Warren et al23 reported no difference between racial groups and the risk of DCIS or invasive recurrence after adjusting for tumor characteristics. It is possible that the higher rate of receiving adjuvant radiation and tamoxifen in our patient population or the shorter median follow-up compared with other published studies may explain the lack of ethnic disparity in DCIS outcomes.

Although the current study, to our knowledge, represents the largest reported single-institution analysis of the impact of race and ethnicity on the biologic features and outcome of DCIS, it also has limitations that should be considered. First, it was conducted in a single institution; thus, the findings may not be generalized to the rest of the population. However, this also may be considered a strength of the investigation, because individual patient's treatment recommendations were based on multidisciplinary institutional guidelines. The study population represents a mix of self-referred patients undergoing screening mammography at the institution with a new diagnosis of DCIS and patients who were referred specifically by their physicians for the management of DCIS. The effect of referral bias for higher risk patients at our center is demonstrated in our finding that 47% of all patients reported a family history of breast or ovarian cancer, which is higher than the 15% overall prevalence of having a family history of breast cancer documented through the National Health Interview Survey.24 We did not have information on the mutation status of the breast cancer susceptibility genes BRCA1/BRCA2 in the study population, which may contribute to the high prevalence of having a family history of breast cancer. Indeed, previous studies have demonstrated a 3.3% to 27% prevalence of BRCA1/BRCA2 mutations in women diagnosed with DCIS,25, 26 and these patients have an elevated lifetime risk of developing invasive breast cancer.27, 28 Nevertheless, few prior studies have evaluated ethnic differences in DCIS treatment and biologic features among patients with equal access to state-of-the-art breast cancer care. Compared with other hospital-based studies, our cohort consisted of an ethnically diverse patient population and included important information on the receipt of adjuvant tamoxifen. In conclusion, our findings suggest that, in a screened population of women, there is some variation in age at presentation, biologic features, and treatment of DCIS between different ethnic groups. Additional studies that include larger numbers of minority ethnic patients with DCIS are needed to confirm whether the consistent application of evidence-based treatment practices results in similar outcomes among racial/ethnic groups.

Acknowledgements

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

We thank Sunita Patterson of The University of Texas M. D. Anderson Cancer Center's Department of Scientific Publications for providing editorial assistance.

FUNDING SOURCES

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

This research was supported in part by a Cancer Prevention Fellowship supported by National Cancer Institute grant R25E CA56452 (Shine Chang, PhD, principal investigator), by the National Institutes of Health through The University of Texas M. D. Anderson Cancer Center support grant CA016672, and by the National Institute on Minority Health and Health Disparities (grant P60MD000503).

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES

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