1 ORIGINAL ARTICLE Impact of Triple Negative Phenotype on Breast Cancer Prognosis Henry G. Kaplan, MD* and Judith A. Malmgren, PhD à *Swedish Cancer Institute at Swedish Medical Center; HealthStat Consulting Inc.; and à Department of Epidemiology, School of Public Health and Community Medicine, University of Washington, Seattle, Washington n Abstract: Triple negative (TN) [estrogen receptor (ER), progesterone receptor (PgR)] (ER) PgR) Her2 neu)) breast cancer (BC) is an aggressive disease without tumor-specific treatment options. Our objective is to evaluate the relative contribution of combined Her2 neu (Her2) and hormone receptor (HR) status to BC progression. A prospective primary BC cohort of 1550 patients at our institution, stage I IV, from 1998 to 2003 were categorized by HR and Her2 status into ER+ PgR+ Her2) (HR+ Her2)) (n = 1134), ER+ PgR+ Her2+ [triple positive (TP)] (n = 138), ER) PgR) Her2) (TN) (n = 183), and ER) PgR) Her2+ (HR) Her2+) (n = 95). Clinical variables were chart abstracted and vital and disease status updated annually. Log-rank tests and Cox regression analyses were used to assess associations with survival. Patient age ranged from 21 to 88 years and average length of follow-up was 4.24 years. Overall survival at 5 years was 94% (HR+ Her2)), 91% (TP), and 81% (TN and HR) Her2+) (log rank test = 22.22, p < 0.001). Disease-specific survival at 5 years was 98% (HR+ Her2)), 93% (TP), 88% (TN), and 86% (HR) Her2+) (log rank test = 25.85, p < 0.001) and 5-year relapse-free survival was 95% (HR+ Her2)), 89% (TP), 84% (TN), and 76% (HR) Her2+) (log rank test = 20.29, p < 0.001). In a model adjusted for age, race, TNM stage, and treatment using HR+ Her2) patients as the reference group, recurrence risk was 1.98 for TP (95% CI = 1.02 to 3.84), 2.32 for TN (95% CI = 1.32 to 4.08), and 4.25 for HR) Her2+ patients (95% CI = 2.33, 7.75). A hierarchy of BC phenotypes defined by HR and Her2 status exists with progressively worse disease outcomes by category. n Key Words: breast cancer, estrogen receptor, her-2 neu, survival, triple negative Hormone receptor and Her2 neu (Her2) analysis are two characteristics commonly used in current clinical practice to gauge breast cancer prognosis and identify appropriate treatment options. Recent studies have suggested differential survival in breast cancer patients based on the presence of specific combinations of estrogen receptor (ER) and progesterone receptor (PgR) status (1 3). Patients with both ERand PgR-positive status have the best reported outcomes whether treated with tamoxifen or not (1,3). In 1987, Slamon et al. (4) reported a worse prognosis among Her2-positive breast cancers compared with Her2-negative tumors. With targeted treatment in the form of trastuzumab (herceptin) for metastatic disease approved by the FDA in November 1998 and for Address correspondence and reprint requests to: Henry G. Kaplan, MD, Swedish Medical Center, Swedish Cancer Institute, 1221 Madison Street, Seattle, WA 98104, USA, or Supported by the Kaplan Research Fund. Our study is an original research study, part of which was presented as a poster at the San Antonio Breast Cancer Symposium, December ª 2008 Wiley Periodicals, Inc., X/08 The Breast Journal, Volume 14 Number 5, adjuvant treatment in 2006, testing for Her2 has become part of routine clinical practice (5,6). In a cohort study of breast cancer patients from North Carolina oversampled for premenopausal and African American women, Carey et al. (7) found the shortest survival among Her2+ ER) and basal-like subtypes compared with other HR and growth factor receptor subtypes. Gago et al. (8) found a lower disease-free survival rate among hormone receptor (HR)-positive Her2 neu-positive patients compared with HR-positive Her2 neu-negative patients, all treated with tamoxifen. More recent studies have identified a prognostic concordance between gene markers, other tumor markers, and triple negative (TN) status, indicating a distinct biologic nature of the TN phenotype (9 11). Our objective is to identify distinct subgroups of primary breast cancer using ER, PgR, and Her2 neu status and to compare presentation, treatment, and outcomes in a cohort of breast cancer patients prior to widespread use of trastuzumab for metastatic disease or adjuvant chemotherapy. In particular, we are interested in differential survival and recurrence between TN patients, ER) PgR) Her2 neu), and ER) PgR)
2 Survival in Triple Negative Breast Cancer 457 Her2 neu+ patients compared with their ER+ PgR+ counterparts, correcting for presentation characteristics to determine if phenotypic subtype carries a differential risk of it s own. METHODS A breast cancer registry of all patients seen at our institution was created in 1990 containing detailed information on diagnosis, staging, surgery, chemotherapy, radiation therapy, tumor markers, and follow-up status. Registry follow-up is updated on an annual basis by a certified cancer registrar with information on recurrence, subsequent treatment, and vital status, current through Vital and disease status information is obtained from chart review if the patient is still seen at the Institution or through physiciandirected follow-up letter if follow-up care is provided elsewhere. Patients not under the care of a managing physician are contacted by mail using an Institutional Review Board (IRB) approved letter from their diagnosing physician requesting annual follow-up information. If no response is received, the institution s cancer registry and the Seattle-Puget Sound Surveillance Epidemiology and End Results cancer registry are reviewed for vital and disease status. Only IRB approved methods were used for patient follow-up and data was input and stored in a password protected HIPAA compliant database. All analyses were conducted using de-identified data as per IRB and HIPAA guidelines. This project was reviewed and approved by the IRB at our community-based regional cancer center. The cohort of 2338 patients TNM [primary tumor (T), regional node (N), distant metastasis (M)] breast cancer stage I IV [Americal Joint Committee on Cancer 6 (AJCC 6)], excluding inflammatory breast cancers (n = 65), was diagnosed between 1998 and 2003 and updated annually for recurrence and or vital status, current as of (12). Her2 neu tumor marker testing was carried out on the majority of the patient population in 1998 (64%) and by 2001, 94% of all breast cancer patients had Her2 neu testing conducted on them. Out of the 2338 patients, 2019 in our cohort were tested for Her2 neu (86%). Her2 neu testing on tissue samples was carried out at various local pathology laboratories by immunohistochemistry and the results were abstracted by chart review for entry into the registry. Her2 neu tests were chart reported in three ways and are as follows: (i) 0, 1+ = negative, 2+ = indeterminate, and 3+ = positive; (ii) as percentages <10% = negative = 0)1+, 10% to 30% = equivocal = 2+, and >30% = positive=3+; and (iii) as negative, weakly positive, and positive. Patients with Her2 neu test results reported as weak-positive, 2+, or indeterminate without fluorescence in situ hybridization (FISH) testing conducted for confirmation were excluded from the analysis (n = 168) leaving a cohort of 1851 patients. FISH testing was recorded in the chart on (7.1%) of these patients. ER and PgR testing was conducted at various local laboratories with a single report format of positive, negative, or borderline test results. Of the 1851 Her2 neu tested patients, 60 were excluded because they were not tested for HR status or had ER or PgR test results reported as borderline in the chart. We restricted our comparative and survival analysis to the cohort of HR concordant (ER+ PgR+ and ER) PgR)) pairs with conclusive Her2 neu tests (n = 1550). In survival analysis using Kaplan Meier plots and log rank pairwise category comparisons, ER+ PgR), ER) PgR+, and ER) PgR) were all significantly different from ER+ PgR+. However, the ER+ PgR), ER) PgR+, and ER) PgR) groups were not significantly different from each other (data not presented). Therefore to evaluate the independent effect of Her2 neu status in combination with HR status, we excluded discordant ER PgR pairs, ER+ PgR) (n = 224), and ER) PgR+ (n = 17) from the analysis. The resulting cohort consisted of females aged years diagnosed with primary invasive breast cancer, with a minimum of 2 years follow-up (n = 1550). Bivariate analyses of significant differences by HR her2-neu status were performed using Pearson chi-square tests for categorical variables and t-tests for continuous variables (F statistic). For ordinal categorical variables in 2 by k table analysis by detection method, a chi-square test for trend was used with one degree of freedom. All p-values were two-sided. The Kaplan Meier method was used to create survival curves and calculate 5-year survival rates. Overall survival (OS) was defined as time to death from any cause, disease-specific survival (DSS) was defined as time to death from breast cancer and relapse-free survival (RFS) was defined as the time to first recurrence of breast cancer, local, regional, distant, or combination. DSS is defined as the percentage of people in a study who have survived a particular disease since diagnosis or treatment. For the disease-specific and RFS analysis, stage IV tumors were excluded. Breast cancer specific death was assigned as the endpoint if
3 458 kaplan and malmgren the patient died without ever being cancer-free or if disease had recurred and was present at death. Patients who died of other causes without breast cancer present at death were not counted as a breast cancer specific death. Cox regression analysis was used to estimate crude and adjusted hazard ratios and their 95% confidence intervals (CIs) for OS, DSS, and RFS. The Cox model regression by HR Her2 category was run unadjusted and adjusted for age, race, treatment (adjuvant chemotherapy and or radiation), and AJCC 6 TNM stage. spss version (SPSS Inc., Chicago, IL, USA) was used for all statistical analysis (13). Of the Stage I II patients, 14% of the TP patients (n = 19) and 19% of the HR) Her2+ patients (n = 18) (total = 37) received adjuvant trastuzumab treatment. Five of these patients were part of experimental protocols and the rest were not. Analysis was conducted with and without these patients in the equations to assess the impact of trastuzumab on outcome. RESULTS In our cohort of 1550 women diagnosed with breast cancer between 1998 and 2003, the mean age was years with a range from 21 to 88 years. Eightyeight percent were white, 2% were black, 7% were Asian, and 3% were other or unknown. Sixty percent were diagnosed between 2001 and The majority of patients were ER+ PgR+ Her2) (HR+ Her2)) (n = 1134, 73%), 138 patients were ER+ PgR+ Her2+ (triple positive, TP) (9%), 183 were ER) PgR) Her2) (TN) (12%), and 95 patients were ER) PgR) Her2+ (HR) Her2+) (6%). Less than half of women aged 21 to 39 years had HR+ Her2) tumors (44.5%) with 14% TP, 27% TN, and 15% HR) Her2+ marker status tumors (Pearson v 2 = 86.27, p < 0.001). Conversely, the majority of women aged years had HR+ Her2) tumors (81%) (Fig. 1). The cohort was predominantly white with the highest percentage of black women in the TN group (7%) (Pearson v 2 = 34.79, p < 0.001). Results are presented in Table 1. Significantly more of the TP (51%), TN (61%), and HR) Her2+ (57%) women had patient-detected breast cancer compared with the HR+ Her2) (35%) group (Pearson v 2 = 66.16, p < 0.001). More of the HR+ Her2) patients had lobular and ductal lobular carcinoma than the other three groups. The TN and HR) Her2+ women had significantly more unusual histologic types with six metaplastic tumors in the TN Figure 1. HR Her2 status by age. group. The TN group had the largest mean tumor size (2.79 cm, 95% CI = 2.43, 3.15) compared with HR+ Her2) (1.88 cm, 95% CI = 1.79, 1.97), TP (2.50 cm, 95% CI = 2.14, 2.85), and HR) Her2+ (2.65 cm, 95% CI = 2.23, 3.07) (F statistic = 20.44, p < 0.001). HR+ Her2) patients had the highest percentage of tumors less than or equal to 2 cm (73%), with 57% in the TP group, 51% in the TN group, and 47% in the HR) Her2+ group (Table 1). The HR+ Her2) group had the largest percentage of TNM stage I cases (55%) with progressively fewer in the TP (43%), TN (39%), and HR) Her2+ (30%) groups (Table 1). The TN group had the most patients with high histologic and nuclear grade and high Bloom-Richardson scores, 92%, 80%, and 86%, respectively, followed by the HR) Her2+ group, TP group, and HR+ Her2) group. The HR) Her2+ group had the highest mean number of positive lymph nodes, 2.17, followed by TN, 1.50, TP, 1.37, and HR+ Her2), 1.15 (F statistic = 3.32, p = 0.019). The HR+ Her2) and the TN groups were equally unlikely to have one or more positive lymph nodes, 36% and 35% respectively, with a higher percentage of one or more positive nodes in the TP (44%) and HR) Her2+ (53%) groups. Triple positive (TP) (41%), TN (33%), and HR) Her2+ (36%) women were significantly more likely to have modified radical mastectomies than the HR+ Her2) women (27%) (Pearson v 2 = 23.90, p = 0.004). Thirty-seven percent of the HR+ Her2) women received both radiation and chemotherapy treatment compared with the TP (61%), the TN (73%), and the HR) Her2+ patients (84%) (Pearson v 2 = , p < 0.001). Vital status at follow-up was significantly different by HR Her2 category. With a minimum of 2 years
4 Survival in Triple Negative Breast Cancer 459 Table 1. Descriptive Characteristics by Hormone Receptor (HR) Her-2 neu Status (n = 1550) HR+ Her2) (col. %) HR+ Her2+ (col. %) HR) Her2) (col. %) HR) Her2+ (col. %) Variable n = 1134 n = 138 n = 183 n =95 Chi-square p-value Race White 1012 (89%) 116 (84%) 156 (85%) 78 (82%) <0.001 Black 14 (1%) 3 (2%) 12 (7%) 2 (2%) Other* 108 (10%) 19 (14%) 15 (8%) 15 (16%) Menopausal status Post 750 (67%) 75 (56%) 91 (51%) 48 (51%) <0.001 Detection method Mammography 636 (57%) 57 (42%) 56 (31%) 31 (33%) <0.001 Patient 395 (35%) 69 (51%) 111 (61%) 53 (57%) Physician 87 (8%) 9 (7%) 15 (8%) 9 (8%) Histology Ductal 860 (76%) 121 (88%) 166 (91%) 87 (92%) <0.001 Lobular 130 (12%) 6 (4%) 5 (3%) 0 (0%) Ductal lobular 84 (7%) 4 (4%) 1 (0.5%) 1 (0.5%) Other mixed types 60 (5%) 7 (5%) 11 (6%) 7 (7%) T stage T1: 2 cm 828 (73%) 78 (57%) 93 (51%) 45 (47%) <0.001 TNM stage (12) I 619 (55%) 59 (43%) 72 (39%) 28 (30%) <0.001 II 398 (35%) 56 (41%) 73 (40%) 45 (47%) III 97 (9%) 19 (14%) 31 (17%) 18 (19%) IV 20 (2%) 4 (3%) 7 (4%) 4 (4%) Histologic grade Grade 1 low 138 (12%) 7 (5%) 1 (0.6%) 3 (3%) <0.001 Grade 2 int 362 (33%) 38 (29%) 13 (7%) 8 (9%) Grade 3 high 615 (55%) 87 (66%) 165 (92%) 79 (88%) Nuclear grade Grade 1 low 127 (11%) 8 (6%) <0.001 Grade 2 int 772 (69%) 67 (50%) 36 (20%) 20 (22%) Grade 3 high 217 (19%) 60 (44%) 141 (80%) 72 (78%) B R score Grade (36%) 14 (12%) 3 (2%) 1 (1%) <0.001 Grade (48%) 61 (52%) 20 (12%) 18 (22%) Grade (16%) 43 (36%) 146 (86%) 64 (77%) int, intermediate. *Other: asian, hispanic, and native american; B R scores, Bloom Richardson scores. Other mixed types: colloid mucinous, tubular, cancer not otherwise specified, ductal carcinoma in situ with microinvasion, papillary, medullary, and metaplastic. Table 2. Outcomes by Hormone Receptor (HR) Her-2 neu Status (n = 1550) HR+ Her2) (col. %) HR+ Her2+ (col. %) HR) Her2) (col. %) HR) Her2+ (col. %) Variable n = 1134 n = 138 n = 183 n =95 Chi-square p-value Outcomes Vital status Alive NED 1038 (92%) 122 (88%) 144 (78%) 74 (78%) <0.001 Alive w this Ca 36 (3%) 4 (3%) 10 (6%) 5 (5%) Alive w other Ca 9 (0.8%) 1 (0.7%) 3 (2%) 1 (1%) Expired NED 23 (2%) 2 (1%) 4 (2%) 2 (2%) Expired w this Ca 22 (2%) 8 (6%) 19 (10%) 13 (14%) Expired w other Ca 6 (0.5%) 1 (0.7%) 4 (2%) 0 Recurrence type None 1074 (96%) 123 (91%) 153 (85%) 72 (77%) <0.001 Local regional 6 (0.5%) 2 (1.5%) 5 (3%) 4 (4%) Distant 39 (3.5%) 10 (7.5%) 23 (13%) 15 (16%) Combination 1 (0.1%) (3%) NED, no evidence of disease; Ca, cancer; col, column.
5 460 kaplan and malmgren follow-up for alive patients, 91% of the HR+ Her2) group were alive no evidence of disease, followed by 88% of the TP, 78% of the TN, and 77% of the HR) Her2+ women. The HR) Her2+ patients had the highest rate of death because of disease (14%), followed by 10% among the TN, 6% among the TP, and 2% among the HR+ Her2). The rate of distant recurrence was 17% in the HR) Her2+ group, 13% in the TN group, 7% in the TP group, and 3% in the HR+ Her2) group. There were fewer cases of local or regional recurrence and overall recurrence was highest in the HR) Her2+ group (24%), followed by 16% in the TN group, 8.4% in the TP group, and 3.3% in the HR+ Her2) group (Pearson v 2 = , p < 0.001). Outcome results are presented in Table 2. For OS using all cause mortality as the endpoint, the HR+ Her2) patients (5-year OS = 94%) and TP patients (5-year OS = 91%) had significantly better survival than the HR) Her2+ (5-year OS = 81%) and TN (5-year OS = 81%) patients (n = 1550). At 5 years, the HR+ (HR+ Her2), TP) and HR) (TN, HR) Her2+) groups were significantly different but there was no significant difference in OS comparing HR+ Her2) to TP and TN to HR) Her2+ groups by pairwise log rank comparisons (Fig. 2). For DSS with mortality from this cancer as the outcome and excluding TNM stage IV patients, DSS was 98% for the HR+ Her2) group, 95% for the TP group, 88% for the TN group, and 86% for the HR) Her2+ group (overall log rank test = 36.92, p < 0.001) (n = 1515). By log rank pairwise comparisons, the HR+ Her2) had significantly better DSS than the TN and HR) Her2+ groups but the same DSS as the TP group (Fig. 2). Relapse-free survival with local, regional, or distant relapse as the outcome and excluding the TNM stage IV patients were the following: HR+ Her2), 95%; TP, 89%; TN, 84%, and HR) Her2+, 76% (Fig. 2) (n = 1515). The HR) Her2+ group had worse RFS than the HR+ Her2) and TP groups but TN patients were only significantly different compared with the HR+ Her2) group (Fig. 2). When adjuvant herceptin-treated patients were removed from the analysis (n = 34), the HR) Her2+ group RFS dropped from 76 to 73%. In unadjusted models, TP status was not associated with OS (HR = 1.42, 95% CI = 0.74, 2.72) or disease-free survival (HR = 2.31, 95% CI = 0.91, 5.83) but was associated with RFS (HR = 2.10, 95% Figure 2. Kaplan Meier plots. CI = 1.11, 3.97) compared to the reference group (HR+ Her2)) (Table 3). TN status was associated with OS (HR = 3.21, 95% CI = 2.02, 5.12), DSS (HR = 4.89, 95% CI = 2.43, 9.83), and RFS (HR = 3.29, 95% CI = 1.99, 5.44) compared to the reference group. HR) Her2+ status was associated
6 Survival in Triple Negative Breast Cancer 461 Table 3. Cox Regression Analysis of Relapse- Free, Disease-Specific, and Overall Survival by Hormone Receptor (HR) Her2 neu Status (n = 1550) Hormone receptor Her2 neu status* Unadjusted hazard ratio (95% CI) p-value Adjusted hazard ratio (95% CI) p-value Overall survival (n = 1550) TP 1.42 (0.74, 2.72) (0.85, 3.22) TN 3.21 (2.02, 5.12) < (2.48, 7.40) <0.001 HR) Her (1.56, 4.96) < (2.06, 8.10) <0.001 Disease-specific survival (n = 1515) TP 2.31 (0.91, 5.83) (0.98, 6.64) TN 4.89 (2.43, 9.83) < (2.11, 10.48) <0.001 HR) Her (2.86, 13.00) < (3.58, 20.68) <0.001 Relapse-free survival (n = 1515) TP 2.10 (1.11, 3.97) (1.02, 3.81) TN 3.29 (1.99, 5.44) < (1.42, 4.23) HR) Her (3.08, 9.04) < (2.47, 8.06) <0.001 TP, triple positive; TN, triple negative. *HR+ Her2) subtype was used as the reference category; model adjusted for age, race, treatment, and TNM stage. Excluding TNM stage IV patients. with OS (HR = 2.78, 95% CI = 1.56, 4.96), DSS (HR = 6.10, 95% CI = 2.86, 13.00), and RFS (HR = 5.28, 95% CI = 3.08, 9.04) compared to the reference group (Table 3). In Cox regression analysis age, race, treatment (adjuvant chemotherapy, radiation therapy, or combined chemotherapy and radiation therapy), TNM stage (AJCC 6), and HR Her2 neu category, were significant predictors of any cause mortality. In the adjusted model, (i) TP status was not associated with OS (hazard ratio = 1.65, 95% CI = 0.85, 3.22) or DSS (hazard ratio = 2.55, 95% CI = 0.98, 6.64), but was associated with RFS (hazard ratio = 1.98, 95% CI = 1.02, 3.81) compared to HR+ Her2) patients (reference group); (ii) TN status was associated with OS (hazard ratio = 4.29, 95% CI = 2.48, 7.40), DSS (hazard ratio = 4.70, 95% CI = 2.11, 10.48), and RFS (hazard ratio = 2.45, 95% CI = 1.42, 4.23) compared to the reference group; and (iii) HR) Her2+ status was associated with OS (hazard ratio = 4.09, 95% CI = 2.06, 8.10), DSS (hazard ratio = 8.61, 95% CI = 3.58, 20.68), and RFS (hazard ratio = 4.46, 95% CI = 2.47, 8.06) compared with the reference group (Table 3) (Fig. 3). DISCUSSION A linear trend was observed in prognostic characteristics by HR Her2 neu-defined phenotypic subtype with a higher percentage of favorable characteristics in the HR+ Her2) and progressively less favorable Figure 3. Hazard function graphs. profiles in each subsequent phenotypic subtype from HR+ Her2) to HR+ Her2+ to HR) Her2) to HR) Her2+. Variables following this trend included age, stage, histologic, and nuclear grade. A reverse progression was observed by group for treatment with the HR) Her2+ group receiving the most aggressive treatment followed by the TN, TP, and HR+ Her2) group. Likewise for outcomes, the HR+ Her2) group was least likely to die of disease and least likely to have a recurrence followed by the TP, TN, and
7 462 kaplan and malmgren HR) Her2+ groups. OS was significantly better for HR-positive breast cancer [HR+ Her2) and HR+ Her2+ (TP)] than for HR-negative breast cancer (HR) Her2) (TN) and HR) Her2+). The disease-specific and relapse-free risk patterns by HR and Her2 groupings mirrored the severity levels observed by prognostic characteristics with the best survival among the HR+ Her2) group followed by TP, TN, and the HR) Her2+ group. Likewise, the Cox proportional hazards model results display a pattern of progressively worse risk of disease specific mortality and disease recurrence for the TP, TN, and HR) Her2+ groups compared with the HR+ Her2) group adjusted for age, race, treatment, and stage. The strengths of our study include the comprehensive nature of the registry database with patient presentation characteristics, treatment and outcomes, and complete ascertainment of patient status at regular follow-up intervals. The availability of a registry cohort of breast cancer patients with complete follow-up and Her2 neu test results present a unique opportunity to study the presentation characteristics and outcomes associated with Her2 neu-defined disease. Her2/neu testing was done on 94% or more of breast cancer patients from 2000 to 2003 with lower testing rates in 1998 and 1999 when the test first became available. Women in our cohort not tested for Her2/neu were more likely to be ER-positive and TNM stage I (data not presented). The result of a lower rate of testing among ER-positive and stage I patients would be a reduced representation of lower stage, more benign disease in our cohort primarily in the reference group of HR+ Her2) patients. The small admixture of trastuzumab patients was addressed in two ways to correct for potential confounding. One way was to conduct the survival analysis with and without trastuzumabtreated patients, and two to add chemotherapy treatment to the hazard analysis model. Our experience, which encompasses the pre-herceptin era, confirms the findings of others that Her2- positive disease has a worse prognosis than Her2-negative disease (4,14). Recent studies have shown that treatment with trastuzumab appears to eliminate the disparity in outcomes between these two groups (6,15). Our findings indicate that HR status is a more powerful predictor of outcome than Her2 neu status. Better outcomes in the HR-positive patients are partly because of better response to more specific multimodality treatment of disease more often discovered at an early stage by mammography. Her2 status is not as strong a distinguishing prognostic indicator but in combination with HR status, identifies a progressively worse outcome with each change in HR and Her2 neu status in spite of aggressive treatment administration. We have described a hierarchical model of breast cancer phenotypic subtypes based on HR and Her- 2 neu status that has prognostic value for clinical evaluation of patient risk. The differential presentation of more aggressive phenotypes in young women suggests a different biologic mechanism may be at work creating a situation in which patients present with more advanced disease that is less often detected by mammography (16). Hopefully, further study of the biology of more aggressive phenotypes like TN HR Her-2 neu tumors will lead to improved methods of diagnosis and better treatment protocols. Acknowledgments The authors wish to gratefully acknowledge the work of Mary Atwood, the Swedish Cancer Institute s breast cancer registrar for her invaluable assistance in collecting the data for this study. REFERENCES 1. Arpino G, Weiss H, Lee AV, et al. Estrogen receptor-positive, progesterone receptor-negative breast cancer: association with growth factor receptor expression and tamoxifen resistance. J Natl Cancer Inst 2005;97: Grann VR, Troxel AB, Zojwalla NJ, et al. Hormone receptor status and survival in a population-based cohort of patients with breast carcinoma. Cancer 2005;103: Ouyang Y, Li D, Pater JL, et al. The importance of temporal effects in evaluating the prognostic impact of joint ERPR expression in premenopausal women with node-positive breast cancer. Breast Cancer Res Treat 2005;92: Slamon DJ, Clark GM, Wong SG, et al. Human breast cancer: correlation of relapse and survival with amplification of the Her-2 neu oncogene. Science 1987;235: Slamon DJ, Leyland-Jones B, Shak S, et al. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 2001;344: Romond EH, Perez EA, Bryant J, et al. Trastuzumab plus adjuvant chemotherapy for operable HER2-positive breast cancer. N Engl J Med 2005;353: Carey LA, Perou CM, Livasy CA, et al. Race, breast cancer subtypes, and survival in the Carolina breast cancer study. JAMA 2006;295: Gago FE, Fanelli MA, Ciocca DR. Co-expression of steroid hormone receptors (estrogen receptor a and or progesterone receptors) and her2 neu (c-erbb-2) in breast cancer: clinical outcome following tamoxifen-based adjuvant therapy. J Steroid Biochem Mol Biol 2006;98: Haffty BG, Yang Q, Reiss M, et al. Locoregional relapse and distant metastasis in conservatively managed triple negative early stage breast cancer. J Clin Oncol 2006;24:
8 Survival in Triple Negative Breast Cancer Rakha EA, El-Sayed ME, Green AR, et al. Prognostic markers in triple-negative breast cancer. Cancer 2007;109: Fan C, Oh DS, Wessels L, et al. Concordance among geneexpression-based predictors for breast cancer. N Engl J Med 2006;355: American Joint Committee on Cancer Staging, 6th edn. New York, NY: Springer-Verlag, SPSS. SPSS for Windows, Release Chicago, IL, USA: SPSS, Perou CM, Sorlie T, Elsen MB, et al. Molecular portraits of human breast tumours. Nature 2000;406: Piccart-Gebhart MJ, Procter M, Leyland-Jones B, et al. Trastuzumab after adjuvant chemotherapy in HER2-positive breast cancer. N Engl J Med 2005;353: Kaplan HG, Malmgren JA. Disease specific survival in patient detected breast cancer. Clin Breast Cancer 2006;7: