ORIGINAL ARTICLE

Prognostic Value of Breast Cancer Subtypes, Ki-67 Proliferation Index, Age, and Pathologic Tumor Characteristics on Breast Cancer Survival in Caucasian Women N. Lynn Ferguson, MD,*a John Bell, MD,† Robert Heidel, PhD,§ Solomon Lee, DO,* Stuart VanMeter, MD,* Lisa Duncan, MD,* Barbara Munsey, BS,¶ Timothy Panella, MD,‡ and Amila Orucevic, MD, PhD* *University of Tennessee Medical Center at Knoxville, Graduate School of Medicine, Department of Pathology, Knoxville, TN, USA; †University of Tennessee Medical Center at Knoxville, Graduate School of Medicine, Department of Surgery, Knoxville, TN, USA; ‡University of Tennessee Medical Center at Knoxville, Graduate School of Medicine, Department of Internal Medicine, Knoxville, TN, USA; §University of Tennessee Medical Center at Knoxville, Graduate School of Medicine, Knoxville, TN, USA; ¶University of Tennessee Medical Center at Knoxville, Cancer Center, Knoxville, TN, USA

n

Abstract: Estrogen receptor (ER), progesterone receptor (PR), and epidermal growth factor receptor 2 (HER2) status are well-established prognostic markers in breast cancer management. The triple negative breast carcinoma subtype (ER-/ PR-/HER2-) has been associated with worse overall prognosis in comparison with other subtypes in study populations consisting of ethnic minorities and young women. We evaluated the prognostic value of breast cancer subtypes, Ki-67 proliferation index (Ki-67PI), and pathologic tumor characteristics on breast cancer survival in Caucasian women in our institution, where greater than 90% of the total patient population is white. From 628 new invasive breast cancer cases in our data base (2000-late 2004), 593 (94%) were identified in Caucasian women. ER/PR/HER2 breast cancer subtypes were classified based on St. Gallen International Expert Consensus recommendations from 2011. ER/PR/HER2 status and its effect on survival were analyzed using a Kaplan–Meier curve. ER/PR/HER2 status, grade, tumor-node-metastasis status (TNM)/ anatomic stage, and age were analyzed in terms of survival in a multivariate fashion using a Cox regression. Ki-67PI was analyzed between ER/PR/HER2 groups using the Kruskal–Wallis, Mann–Whitney U-tests, and 2 9 5 ANOVA. Our results showed that patients with stage IIB through stage IV breast carcinomas were 2.1–16 times more likely to die than patients with stages IA-B and IIA disease, respectively (95% CI 1.17–3.81 through 9.68–28.03, respectively), irrespective of ER/PR/ HER2 subtype. Similar effect was seen with T2, N2/N3, or M1 tumors in comparison with T1, N0/N1, and M0 tumors. Chances of dying increase approximately 5% for every year increase in age. There was a significant main effect of Ki-67PI between ER/PR/HER2 subtypes, p < .001, but Ki-67PI could not predict survival. In summary, TNM status/anatomic stage of breast carcinomas and age are predictive of survival in our patient population of Caucasian women, but breast carcinoma subtypes and Ki-67 proliferation index are not. n Key Words: Breast cancer subtypes, Caucasian women, clinicopathologic characteristics of breast carcinoma, Ki-67 proliferation index, overall survival

Address correspondence and reprint requests to: Amila Orucevic, MD, PhD, The University of Tennessee Medical Center, Department of Pathology, 1924 Alcoa Highway, Knoxville, TN 37920, USA, or e-mail: aorucevic@ utmck.edu. a The abstract with partial data from this manuscript received a 2012 The Breast Journal Award, and Dr. Ferguson received the third place award upon presentation of our work at the 17th Annual Multidisciplinary Symposium on Breast Disease in Amelia Island, Florida, in February 2012. DOI: 10.1111/tbj.12059 © 2012 Wiley Periodicals, Inc., 1075-122X/13 The Breast Journal, Volume 19 Number 1, 2013 22–30

INTRODUCTION

B

ased on the American Cancer Society Cancer Facts and Figures 2012 data for breast cancer (1), 226,870 new cases of invasive breast cancer are anticipated to occur among women in the United States during 2012. Incidence of breast cancer rates was stable from 2004 to 2008 (most recent available data), and death rates have progressively decreased in

Breast Cancer Prognostic Markers & Survival • 23

women since 1990, with larger decreases in younger women. The decrease in breast cancer death rates is assumed to be a result of earlier detection, improved treatment, and perhaps decreased incidence (1). Codification of tumor staging into the TNM system by the American Joint Committee on Cancer (AJCC) celebrated its 50th anniversary in 2009, and relevance of TNM staging for breast cancer in the 21st century was questioned by the Breast Cancer Task Force (2). While T, N, and M still provide value in establishing the future outcome of breast cancer patients, determination of estrogen receptor (ER), progesterone receptor (PR), and epidermal growth factor receptor 2 (HER2) status has both prognostic and predictive values in breast cancer management (3–6), and testing for them is now considered standard of care (7). The triple negative breast carcinoma subtype (ER /PR /HER2 ) has been associated with worse overall prognosis in comparison with other subtypes in study populations consisting of ethnic minorities and young women (8–16). Only a few studies tried to address causes for possible differences in overall prognosis and survival between African-American and Caucasian women with breast cancer, and relationship of ER/PR/HER2 status/subtype of breast cancer with socioeconomic influences such as poverty, diet, body mass index, history of pregnancy, oral contraceptive use, and postmenopausal hormone replacement was discussed with not entirely conclusive results (10,12,13,17,18). A single study of an allwhite patient population from Scotland suggested that a higher risk of ER-negative breast cancer is correlated with a lifelong history of poverty (19). Similar results were reported in the United States in a report by Gordon (20). The studies mentioned previously clearly identified the need for more research to better understand the relationship between race and breast cancer. Owing to the geographic location of our institution in East Tennessee, greater than 90% of the total patient population is white. More than 95% of breast cancer patients diagnosed in our institution are also treated and followed in our institution. We evaluated the prognostic value of breast cancer subtypes, Ki-67 proliferation index, and pathologic tumor characteristics on breast cancer survival in a large cohort of Caucasian women diagnosed and treated for breast cancer in our institution.

MATERIALS AND METHODS Selection and Description of Participants Our study population is a retrospective cohort group taken from a prospectively monitored data base of 593 Caucasian women diagnosed with primary breast cancer between January 1, 2000 and September 8, 2004, with an age range of 23–98. This time frame was chosen to assure a minimum of 5 years of followup data. The patients were treated and followed at our academic institution located in Knoxville, Tennessee, where greater than 90% of the total patient population is Caucasian. After obtaining Institutional Review Board approval, we searched our institution’s cancer center data base for newly diagnosed invasive breast cancer cases in women, spanning from the year 2000 to 2004. A total of 628 new invasive breast cancer cases were identified in women. Among those, 593 (94%) patients were Caucasian, 27 (4.2%) were African American, and eight (1.2%) were of unknown ethnicity. We chose only the 593 Caucasian women as our study cohort and analyzed the following clinicopathologic characteristics: age, tumor type, tumor size, tumor grade, Nottingham score, tumor-node-metastasis status, anatomic stage/prognostic groups, ER/PR/ HER2 status, Ki-67 proliferation index, and overall survival status.

Determination of Clinicopathologic Tumor Characteristics, Breast Cancer Subtype, and KI-67 Proliferation Index Tumor type was determined based on World Health Organization classification (21). Tumor size, Nottingham score, and histologic tumor grade (the Nottingham combined histologic grade) were evaluated at the time of the pathologic assessment of the tumors based on established criteria (21–23) and were extrapolated from the pathology reports. Tumor-node-metastasis status was assigned based on the AJCC staging system that was in place at the time of assessment (24,25) and was restaged in our study to the newest AJCC staging system (edn 7) (2). Stages IA and IB were combined in our study. Estrogen receptor, PR, and HER2 status and Ki-67 proliferation index were determined at the time of the patient’s cancer assessment by immunohistochemical

24 • lynn ferguson et al.

methods on a selected tumor block of the needle core biopsy tissue and/or a selected tumor block after final surgical excision. ER and PR were considered positive with >1% of the nuclear staining, based on the newest ASCO/CAP guidelines for immunohistochemical testing for estrogen and progesterone receptor in breast cancer (26). HER2 was considered overexpressed with 3+ score positivity by immunocytochemical methods and negative with 0+ or 1+ immunoreactivity. Tumor blocks with 2+ immunohistochemical scores were sent for fluorescence in situ hybridization to an independent reference laboratory, and a ratio of >2.2 of HER2-neu gene to chromosome 17 was considered positive for HER2 gene overexpression. Ki-67 proliferation index was determined in the area with highest Ki-67 nuclear labeling. A total of 300 proliferating and nonproliferating cells were counted, and the percentage of proliferating cells was calculated and reported as percent proliferating cells. Patient’s Inclusion into the Study, Treatment and Survival Patients with synchronous invasive cancers in both breasts with different ER/PR/HER2 status, multifocal cancers in the same breast with different ER/PR/HER2 status, and patients with newly diagnosed breast cancer with synchronous or metachronous malignancies of other organs were excluded from our study. All of the patients were diagnosed with stages I to IV primary breast cancer and underwent surgical treatment, adjuvant chemotherapy, and radiation therapy according to standard treatment protocols at the time of the diagnosis, based on NCCN guidelines on similar evidence-based data at the time of diagnosis (27). ER-positive patients underwent antihormonal treatment and HER2-positive patients received anti-HER2 therapy only when diagnosed with systemic metastatic disease. Follow-up status and outcome information were obtained from our cancer registry data base, and in the case of patients who were still living, the last date of follow-up was March 11, 2011. Statistical Analysis Breast cancer subtypes (ER/PR/HER2 status) and other clinicopathologic characteristics of invasive carcinomas were examined using frequency statistics, after dividing patients into five groups based on their cancer subtype per the St. Gallen International Expert

Consensus recommendation from 2011 (28) as follows: group 1—luminal A-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index less than 14% (ER+/PR+/HER2 ; ER+/PR / HER2 ; and ER /PR+/HER2 ); group 2—luminal B-/HER2-negative-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index more than 14% (ER+/PR+/HER2 ; ER+/PR / HER2 ; and ER /PR+/HER2 ); group 3—luminal B-/HER2-positive-like subtype = ER and/or PR positive, HER2 positive, and any Ki-67 proliferation index (ER+/PR+/HER2+; ER+/PR /HER2+; and ER /PR+/ HER2+); group 4—HER2-positive-/nonluminal-like subtype with ER and PR negative, HER2 positive (ER /PR /HER2+), and any Ki-67 proliferation index; and group 5—triple negative subtype = ER and PR negative and HER2 negative (ER /PR /HER2 ), and any Ki-67 proliferation index. Totally, 554 patients had complete data on ER/PR/HER2 status and the following clinicopathologic characteristics: age, tumor morphology, Nottingham score and overall grade, tumor status, anatomic stage/prognostic groups, and survival status. Five hundred and twenty-nine patients with known ER/PR/HER2 status had known nodal status, 549 patients had known metastasis status, 541 patients had known tumor size, and 464 had known Ki-67 proliferation index. However, with ER/ PR/HER2 grouping based on the St. Gallen International Consensus recommendations that included Ki-67 proliferation index, 493 patients were eligible to be analyzed in this study. The Kaplan–Meier curve was used as a univariate analysis looking at how the five ER/PR/HER2 groups differed in terms of survival. The Cox regression curve was used as a multivariate analysis looking at how the five ER/PR/HER2 groups differed in terms of survival when controlled for age, grade, tumor status, nodal status, and metastasis status (extraneous variables), or when controlled for age, grade, and anatomic stage/ prognostic groups. The Kruskal–Wallis test was used to test for any effects in Ki-67 proliferation index between the five ER/PR/HER2 groups. The Mann–Whitney U-test was utilized in a post hoc fashion to explain any significant main effect found. Then, 2 9 5 factorial ANOVA was employed to test for significant main effects and interactions between the five ER/PR/HER2 groups and mortality in regard to Ki-67. All analyses were conducted using SPSS Version 19 (SPSS, Chicago, IL), and statistical significance was assumed at a p < .05 level.

Breast Cancer Prognostic Markers & Survival • 25

Table 1. Clinicopathologic Characteristics of Invasive Carcinomas

ER/PR/HER2 frequency Age* Type** Grade** Nottingham score* Size (mm)* Ki-67 %* T stage** N stage** M stage** Anatomic stage** Survival months

Group 1

Group 2

Group 3

Group 4

Group 5

149 = 30.2% 63.07 IDC 1–2 5 18.16 7.62 1 0 0 IA-B 109.6

163 = 33% 59.24 IDC 2 7 23.41 30.08 1 0 0 IA-B 111.3

55 = 11.1% 53.18 IDC 3 7 24.56 31.37 1 0 0 IA-B 117.4

38 = 7.7% 53.68 IDC 3 8 24.2 43.13 1 0 0 IA-B-IIA 107.5

88 = 17.8% 52.82 IDC 3 8 23.08 51.55 1 0 0 IA-B 100.9

*mean value. **most frequent; IDC = invasive ductal carcinoma. group 1—luminal A-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index less than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER /PR+/HER2 ). group 2—luminal B-/HER2-negative-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index more than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER / PR+/HER2 ). group 3—luminal B-/HER2-positive-like subtype = ER and/or PR positive, HER2 positive, and any Ki-67 proliferation index (ER+/PR+/HER2+; ER+/PR /HER2+; and ER /PR+/HER2+). group 4—HER2-positive-/nonluminal-like subtype = ER and PR negative, HER2 positive (ER /PR /HER2+), and any Ki-67 proliferation index. group 5—triple negative subtype = ER and PR negative and HER2 negative (ER /PR /HER2 ) and any Ki-67 proliferation index.

RESULTS Our patient population of 493 Caucasian women was divided into five groups based on their cancer subtype per the St. Gallen International Expert Consensus recommendations from 2011 (28): group 1— luminal A-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index less than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER /PR+/HER2 ); group 2—luminal B-/HER2-negative-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index more than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER /PR+/ HER2 ); group 3—luminal B-/HER2-positive-like subtype = ER and/or PR positive, HER2 positive, and any Ki-67 proliferation index (ER+/PR+/HER2+; ER+/ PR /HER2+; and ER /PR+/HER2+); group 4— HER2-positive-/nonluminal-like subtype with ER and PR negative, HER2 positive (ER /PR /HER2+), and any Ki-67 proliferation index; and group 5—triple negative subtype = ER and PR negative and HER2 negative (ER /PR /HER2 ), and any Ki-67 proliferation index. The clinicopathologic characteristics of invasive carcinomas for each group are summarized in Table 1, showing that the most frequent prognostic subtype was luminal B-/HER2-negative-like subtype (group 2), comprising 33% of the total, immediately followed by luminal A-like subtype. The youngest patients were in triple negative carcinoma subtype (mean age 52.82), and the oldest patients were in luminal A-like subtype (mean age 63.07). The most frequent tumor type for

all groups was invasive ductal carcinoma. Mean Nottingham score for luminal A-like subtype was 5, luminal B-/HER2-negative-like subtype and luminal B-/ HER2-positive-like subtype was 7, and HER2-positive-/nonluminal-like subtype and triple negative subtype was 8. The most frequent overall Nottingham grade 3 was seen in triple negative, HER2-positive-/nonluminal-like subtype, and luminal B-/HER2-positive-like subtype, and grades 2 and 1–2 were seen in luminal B-/HER2negative-like subtype and luminal A-like subtype. Of note, an overall Nottingham grade 1 was not seen in triple negative carcinoma subtype or in HER2-positive-/nonluminal-like subtype (Table 2). Table 3 shows frequencies of tumor stage, node stage, and metastasis stage for each of the five ER/PR/ HER2 groups. The most frequent tumor stage and node stage were 1 and 0, respectively, for all groups, while patients from HER2-positive-/nonluminal-like group had most frequent metastases. Table 4 shows distribution of anatomic stage/prognostic groups for each of ER/PR/HER2 groups. The most frequent cancer stage was stages IA-B, while HER2-positive-/nonluminal-like group had equal distribution of stages IA-B and stage IIA. Figure 1 shows the Kaplan–Meier curve for cumulative survival in months according to the five ER/PR/ HER2 groups. This curve showed no significant difference in survival, p = .358. It is interesting to note that 5-year survival rate was the best for luminal A-like group = 77%, followed by luminal B-/HER-negativelike group = 76%, HER2-positive-/nonluminal-like

26 • lynn ferguson et al.

Table 2. Distribution of Overall Nottingham Grade in Patients Divided into Five Groups Based on ER/PR/HER2 Carcinoma Subtypes Classified Based on the St. Gallen International Expert Consensus Recommendations from 2011 (28)

Group Group Group Group Group Total

1 2 3 4 5

Grade 1

Grade 2

Grade 3

Total

71 22 9 0 0 102

76 77 22 12 15 202

2 64 24 26 73 189

149 163 55 38 88 493

group 1—luminal A-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index less than 14% (ER+/PR+/HER2-; ER+/PR /HER2 ; and ER /PR+/ HER2 ). group 2—luminal B-/HER2-negative-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index more than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER /PR+/HER2 ). group 3—luminal B-/HER2-positive-like subtype = ER and/or PR positive, HER2 positive, and any Ki-67 proliferation index (ER+/PR+/HER2+; ER+/PR /HER2+; and ER /PR+/ HER2+). group 4—HER2-positive-/nonluminal-like subtype = ER and PR negative, HER2 positive (ER /PR /HER2+), and any Ki-67 proliferation index. group 5—triple negative subtype = ER and PR negative and HER2 negative (ER /PR / HER2 ) and any Ki-67 proliferation index.

Table 3. Frequencies of Tumor Stage, Nodal Stage and Metastasis Stage of Invasive Carcinomas in Five ER/PR/HER2 Carcinoma Subtypes Classified Based on the St. Gallen International Expert Consensus Recommendations from 2011 (28) T1 Group Group Group Group Group Total

1 2 3 4 5

112 = 92 = 32 = 18 = 46 = 300

T2

75% 56.4% 58.1% 47.3% 52.2

26 = 48 = 11 = 15 = 32 = 132

N0 Group Group Group Group Group Total

1 2 3 4 5

100 = 100 = 27 = 20 = 47 = 294

17% 29.4 20% 39.4% 36.3

N1

70.9% 64.5% 50% 57.1% 56.6%

30 = 35 = 20 = 12 = 24 = 121

21.2% 22.5% 37% 34.3% 28.9%

M0

group = 74%, and luminal B-/HER2-positive-like subtype = 67%. The lowest 5-year survival (69%) was seen in the triple negative group. When the Cox regression curve for cumulative survival in months according to the five ER/PR/HER2 groups was controlled for anatomic stage/prognostic groups, tumor grade, and age when predicting for survival, stages IIB through IV and age had significant effects on survival (Fig. 2 and Table 5), but hormone receptor status and grade did not within the model. The results found that the chances of dying increases 4.8% for every year increase in age. When the Cox regression curve for cumulative survival in months according to the five ER/PR/HER2 groups was controlled for tumor-node-metastasis status, grade, and age in terms of survival, only tumor stage, node stage, metastasis stage, and age were significant predictors of death, while hormone receptor status and grade were not significant. The patients with T2 tumors, N2, N3 tumors and M1 tumors were 1.72 times; 2.79 times, 2.63 times, and 4.0 times more likely to die than patients with T1, N0/N1 tumors, and M0 tumors, respectively (95% CI 1.093–2.729 for T2; 1.48–5.25 for N2, 1.17–5.92 for N3; and 1.66 –9.64 for M1; Fig. 3), when controlled for hormone receptor status. Figure 4 demonstrates the Ki-67 proliferation index for each of the five ER/PR/HER2 groups. There was a significant difference in Ki-67 proliferation index

Group Group Group Group Group Total

1 2 3 4 5

141 = 150 = 52 = 34 = 83 = 460

95.3% 9.4% 94.5% 91.9% 94.3%

T3 8= 14 = 6= 0= 8= 36

5.3% 8.5% 10.9% 0% 9%

3= 9= 6= 5= 2= 25

N2 8= 12 = 3= 2= 10 = 35

5.6% 7.7% 55.6% 5.7% 12.0%

M1 7= 10 = 3= 3= 5= 28

4.7% 6.3% 5.5% 8.1% 5.7%

T4

Total

2% 5.5% 10.9% 13.1% 2.2%

149 163 55 38 88 493

N3

Total

3= 8= 4= 1= 2= 18

2.1% 5.2% 7.4% 2.9% 2.4%

141 155 54 35 83 468 Total 148 160 55 37 88 488

group 1—luminal A-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index less than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER /PR+/ HER2 ). group 2—luminal B-/HER2-negative-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index more than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER /PR+/HER2 ). group 3—luminal B-/HER2-positive-like subtype = ER and/or PR positive, HER2 positive, and any Ki-67 proliferation index (ER+/PR+/HER2+; ER+/PR /HER2+; and ER /PR+/ HER2+). group 4—HER2-positive-/nonluminal-like subtype = ER and PR negative, HER2 positive (ER /PR /HER2+), and any Ki-67 proliferation index. group 5—triple negative subtype = ER and PR negative and HER2 negative (ER /PR / HER2 ) and any Ki-67 proliferation index.

values between the five groups with p < .001. Mann– Whitney U-tests found significant difference in the Ki67 proliferation index between groups 1 and 2, groups 1 and 3, groups 1 and 4, groups 1 and 5, groups 2 and 4, groups 2 and 5, and groups 3 and 5, with p < .001; groups 3 and 4, p = .003; and groups 4 and 5, p = .043, while no significant effect was seen between groups 2 and 3, p = .413. The means are presented in Table 1. Despite the significant main effect and subsequent post hoc findings, when mortality was accounted for in the 2 9 5 factorial ANOVA, a nonsignificant interaction was found, F(4, 454) = .90, p = .465, g2 = .01, power = .29, meaning there was evidence that Ki-67 proliferation index does not have an impact on survival of patients when divided into the five ER/PR/HER2 groups.

Breast Cancer Prognostic Markers & Survival • 27

Table 4. Distribution of Anatomic Stage/Prognostic Groups of Invasive Carcinomas in Five ER/PR/ HER2 Carcinoma Subtypes Classified Based on the St. Gallen International Expert Consensus Recommendations from 2011 (28)

Group Group Group Group Group Total

1 2 3 4 5

IA-B

IIA

IIB

IIIA

IIIB

IIIC

IV

Total

89 79 22 12 37 239

33 33 11 12 15 104

10 18 8 6 17 59

6 13 5 1 12 37

1 3 4 3 1 12

3 6 2 1 1 13

7 10 3 3 5 28

149 162 55 38 88 492

group 1—luminal A-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index less than 14% (ER+/PR+/HER2-; ER+/PR /HER2 ; and ER /PR+/ HER2 ). group 2—luminal B-/HER2-negative-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index more than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER /PR+/HER2 ). group 3—luminal B-/HER2-positive-like subtype = ER and/or PR positive, HER2 positive, and any Ki-67 proliferation index (ER+/PR+/HER2+; ER+/PR /HER2+; and ER /PR+/ HER2+). group 4—HER2-positive-/nonluminal-like subtype = ER and PR negative, HER2 positive (ER /PR /HER2+), and any Ki-67 proliferation index. group 5—triple negative subtype = ER and PR negative and HER2 negative (ER /PR / HER2 ) and any Ki-67 proliferation index.

Figure 2. Survival for the five ER/PR/HER2 groups classified based on the St. Gallen International Expert Consensus recommendations from 2011 (28), when controlled for anatomic stage/ prognostic groups, tumor grade, and age. Stage IIB and higher (IIB-IV), and age were significant predictors of death, but breast carcinoma subtype and grade were not significant.

Table 5. Cox Regression for Five ER/PR/HER2 Carcinoma Subtypes Classified Based on the St. Gallen International Expert Consensus Recommendations from 2011 (28) When Controlled for Grade, Stage, and Age Significance (p-value)

Figure 1. The Kaplan–Meier curve for cumulative survival in months according to the five ER/PR/HER2 groups classified based on the St. Gallen International Expert Consensus recommendations from 2011 (28). This curve showed no significant difference in survival, with p = .358.

DISCUSSION We evaluated the prognostic value of breast cancer subtypes, Ki-67 proliferation index and pathologic tumor characteristics on breast cancer survival in Caucasian women in our institution, where greater than 90% of the total patient population is white. The median follow-up length of time of our patient cohort was 9 years.

Group 1 Group 2 Group 3 Group 4 Group 5 Grade 1 Grade 2 Grade 3 Stages IA-B Stage IIA Stage IIB Stage IIIA Stage IIIB Stage IIIC Stage IV Age

.144 .192 .454 .864 .346 .563 .814 .380 .000 .912 .013 .000 .008 .000 .000 .000

Adjusted odds ratio

95% CI lower

95% CI upper

.720 .768 .931 1.367

.439 .384 .412 .714

1.180 1.533 2.106 2.620

1.068 1.353

.615 .688

1.858 2.660

.970 2.115 3.496 3.352 5.856 16.475 1.048

.561 1.172 1.973 1.368 2.651 9.680 1.033

1.674 3.814 6.195 8.212 12.937 28.038 1.064

group 1—luminal A-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index less than 14% (ER+/PR+/HER2-; ER+/PR /HER2 ; and ER /PR+/ HER2 ). group 2—luminal B-/HER2-negative-like subtype = ER and/or PR positive, HER2 negative, and Ki-67 proliferation index more than 14% (ER+/PR+/HER2 ; ER+/PR /HER2 ; and ER /PR+/HER2 ). group 3—luminal B-/HER2-positive-like subtype = ER and/or PR positive, HER2 positive, and any Ki-67 proliferation index (ER+/PR+/HER2+; ER+/PR /HER2+; and ER /PR+/ HER2+). group 4—HER2-positive-/nonluminal-like subtype = ER and PR negative, HER2 positive (ER /PR /HER2+), and any Ki-67 proliferation index. group 5—triple negative subtype = ER and PR negative and HER2 negative (ER /PR / HER2 ) and any Ki-67 proliferation index.

28 • lynn ferguson et al.

Figure 3. Survival for the five ER/PR/HER2 groups classified based on the St. Gallen International Expert Consensus recommendations from 2011 (28), when controlled for tumor-node-metastasis status, tumor grade, and age. T2 tumors, N2 and N3 tumors, M1 tumors, and age were significant predictors of death, but breast carcinoma subtype and grade were not significant.

60 50 40 30 Ki-67 PI %

20 10 0

Group Group Group Group Group 1 2 3 4 5

Figure 4. Ki-67 proliferation index (mean values) for the five ER/ PR/HER2 groups classified based on the St. Gallen International Expert Consensus recommendations from 2011 (28). Ki-67 proliferation index was significantly different between majority of the groups (p < .001 for groups 1&2, 1&3, 1&4, 1&5, 2&4, 2&5, 3&5; p = .003 for groups 3&4; p = .043 for groups 4&5).

From our patient population of 593 Caucasian women, 493 patients had complete data on ER/PR/ HER2 status and corresponding Ki-67 proliferation index and were divided into five groups according to their hormone receptor status/subtype based on the cancer subtype classification per the St. Gallen International Expert Consensus recommendation from 2011 (28). St. Gallen’s recommendations account for

clinicopathologic classification of breast carcinoma subtypes using immunohistochemical methods and Ki67 proliferation index rather than defining subtypes by genetic array. Luminal A-like subtype was seen in 30% of our patients, luminal B-/HER2-negative-like subtype in 33%, luminal B-/HER2-positive-like subtype in 11%, HER2-positive-/nonluminal-like subtype in 7.7% and triple negative carcinoma subtype in 17.8%. The percent of patients in each group in our study was similar to previously published data on breast cancer in Caucasian women when tumors are divided according to the hormone receptor subtype (11,12,17,29) with percents for triple negative carcinoma ranging from 15% in O’Brien’s study to 21.8% in Lund’s study (11,17). The triple negative carcinoma subtype occurred in younger patients (mean age 52.82) than the luminal A-like carcinoma subtype (mean age 62.07) in our study. This is similar to the data reported by Tawfik et al. (12) who showed in their study that triple negative patients were significantly younger than nontriple negative patients. The most frequent tumor type in our patient population was invasive ductal carcinoma (72%) followed by invasive lobular carcinoma (11%), invasive mammary carcinoma with ductal and lobular features (10%), and other types (7%). This is in an expected range for breast cancer (21), and is similar to the previously published study by Tawfik et al. (12) in a cohort of 700 patients from University of Kansas Medical Center, with 83% of patients in their study population being Caucasians. Interestingly, although, in Tawfik’s study, 6.1% of triple negative patients had overall Nottingham grade 1 (well-differentiated) tumors, in our study, none of our triple negative patients, nor patients with HER2-positive-/nonluminal-like subtype had well-differentiated tumors. The Kaplan–Meier curve for cumulative survival in months for patients divided into the five ER/PR/HER2 subtypes (luminal A-like, luminal B-like/HER2 , luminal B-like/HER2+, HER2+/nonluminal like, and triple negative) showed no significant difference in survival in our study, although some trend was seen in triple negative patients that had the shortest survival. Interestingly, though, luminal B-/HER2-positive-like patients in our study showed the longer survival than luminal A-like patients. These results are in part concordant with data published by O’Brien et al. (17), who evaluated influence of breast tumor subtypes, race and long-term survival in the Carolina Breast

Breast Cancer Prognostic Markers & Survival • 29

Cancer Study on 1149 invasive breast cancer patients, 631 of which were Caucasians and 518 of which were African American. In their study, triple negative patients and ER /PR /HER2+ patients had a higher mortality rate when compared with luminal A and B subtypes, and mortality of triple negative white patients was higher than triple negative African-American patients. This led them to conclude that the triple negative breast cancer subtype was associated with poor prognosis, but was not a more aggressive disease in African-American patients than Caucasians, which was similar to the results published by Tawfik et al. (12). Results from the molecular phenotypes of tumors collected between 1976–1997 in the Nurses Health Study (3) also showed that triple negative tumors (subdivided to basal like and unclassified), luminal B and HER2-type tumors had worse prognosis than luminal A tumor subtype, when tumor subtype was defined by immunocytochemistry. However, in our study, the longest surviving group was luminal B-/ HER2-positive like. The latter result could not be explained by difference in treatment, because luminal A-like and luminal B-/HER2-negative-like groups received same hormonal treatment as luminal B-/ HER2-positive patients, and the anti-HER2 therapy was given to three patients only with luminal B-/ HER2-positive subtype because of stage IV disease (Table 3). As the triple negative patients in our study had the shortest survival, followed by patients from HER2positive-/nonluminal-like group, it will be interesting to find out in our ongoing study whether ER/PR/HER2 subtype had any significant impact on survival in patients from our institution at a different time interval when antiHER2 treatment was starting to be used in combination with adjuvant chemotherapy in HER2-positive patients (year 2005 forward). The Cox regression curve for cumulative survival in months for patients divided into the five ER/PR/HER2 subtypes in our study, which controlled for anatomic stage, grade and age showed that stage IIB through stage IV, and age had significant effect on survival (Fig. 2 and Table 5), but breast carcinoma subtype and grade did not. Similar results were obtained when the Cox regression curve for cumulative survival in months according to the five ER/PR/HER2 groups in our study was controlled for tumor-node-metastasis status, grade, and age in terms of survival. Only tumor-node-metastasis stage and age were significant predictors of death, while hormone receptor status was not significant (Fig. 3). Our data support the notion that TNM staging for

breast cancer in the 21st century is still relevant (2) and that determination of estrogen receptor, progesterone receptor, and HER2 status as both prognostic and predictive markers in breast cancer management (3–6) is valid but is probably also influenced by the type of classification system used for determining the ER/PR/ HER2 subtypes, characteristics of populations studied, time periods of the study, as well as choice of variables used in statistical analysis. Ki-67 proliferation index was highest in the triple negative breast carcinoma group in our study, followed by HER2-positive-/nonluminal-like group, luminal B-/HER2-positive-like, luminal B-/HER2-negative-like, and luminal A-like subtypes. It has been shown that proliferation was the strongest prognosticator in node-negative breast cancer (30). Combined evaluation of the Ki-67 proliferation index and Oncotype DX recurrence score (the 10-year risk of recurrent distant disease predictive of clinical benefits from chemotherapy in lymph node-negative, ER-positive patients who are to receive tamoxifen therapy) (31) was recently proposed by Gwin et al. (32). The combined evaluation would identify tumors with high recurrence potential in a low-risk recurrence score group and consequently identify patients who would benefit from adjuvant chemotherapy. We evaluated overall survival in our patients without regard of node status and divided into five ER/PR/HER2 subtypes per recommendations from St. Gallen panelists (28), in which classification of two subtypes also depended on Ki-67 proliferation index. We showed that in our study population, Ki-67 did not have significant impact on survival. It is possible that Ki-67 proliferation index is strong prognosticator and has a complementary value only in certain breast cancer patients, which was addressed in a recent study by Aleskandarany et al. (33). They showed that Ki-67 proliferation index could separate subgroups of patients within luminal/hormone receptor-positive breast cancer with significantly different clinical outcome, but not in HER2-positive-/nonluminal-like or triple negative patients. It will be interesting to find out in our ongoing study whether Ki-67 proliferation index had impact on patients’ survival when breast carcinoma subtypes were again classified per St. Gallen panelists’ recommendations, but this time using tumor grade rather than Ki-67 proliferation index (28). This was recently used in the study by von Minckwitz et al. (34) because Ki-67 data were not available for their study cohort. Our ongoing study

30 • lynn ferguson et al.

will hopefully also validate St. Gallen panelists’ recommendations for the use of tumor grade in breast carcinoma subtype classification in the absence of Ki-67 proliferation index. In summary, we evaluated the prognostic value of breast cancer subtypes, Ki-67 proliferation index, and pathologic tumor characteristics on breast cancer survival in Caucasian women in our institution, where greater than 90% of the total patient population is white. We showed that age and tumor-node-metastasis status/anatomic stage of breast carcinomas are predictive of survival in our patient population, but Ki-67 proliferation index, tumor grade, Nottingham score, and breast carcinoma subtype are not. Our data suggest that prognostic value of breast cancer hormonal status on survival of Caucasian women warrants more investigations.

REFERENCES 1. American Cancer Society. 2012; Available at: http://www. cancer.org/acs/groups/content/@epidemiologysurveilance/documents/ document/acspc-031941.pdf. Accessed 04/21, 2012. 2. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, eds AJCC Cancer Staging Manual. 7th edn. New York: Springer, 2010. 3. Dawood S, Hu R, Homes MD, et al. Defining breast cancer prognosis based on molecular phenotypes: results from a large cohort study. Breast Cancer Res Treat 2011;126:185–92. 4. Cianfrocca M, Goldstein LJ. Prognostic and predictive factors in early-stage breast cancer. Oncologist 2004;9:606–16. 5. Hayes DF. Prognostic and predictive factors revisited. Breast 2005;14:493–9. 6. Payne SJ, Bowen RL, Jones JL, Wells CA. Predictive markers in breast cancer–the present. Histopathology 2008;52:82–90. 7. Carlson RW, Allred DC, Anderson BO, et al. Breast cancer. Clinical practice guidelines in oncology. J Natl Compr Canc Netw 2009;7:122–92. 8. Maskarinec G, Sen C, Koga K, Conroy SM. Ethnic differences in breast cancer survival: status and determinants. Womens Health (Lond Engl) 2011;7:677–87. 9. Sparano JA, Wang M, Zhao F, et al. Race and hormone receptor-positive breast cancer outcomes in a randomized chemotherapy trial. J Natl Cancer Inst 2012;104:406–14. 10. Al-Abbadi MA, Washington TA, Saleh HA, Tekyi-Mensah SE, Lucas DR, Briston CA. Differential expression of HER-2/NEU receptor of invasive mammary carcinoma between Caucasian and African American patients in the Detroit metropolitan area. Correlation with overall survival and other prognostic factors. Breast Cancer Res Treat 2006;97:3–8. 11. Lund MJ, Trivers KF, Porter PL, et al. Race and triple negative threats to breast cancer survival: a population-based study in Atlanta. GA. Breast Cancer Res Treat 2009;113:357–70. 12. Tawfik O, Davis K, Kimler BF, et al. Clinicopathological characteristics of triple-negative invasive mammary carcinomas in African-American versus Caucasian women. Ann Clin Lab Sci 2010 Fall;40:315–23. 13. Brawley OW. Toward a better understanding of race and cancer. Clin Cancer Res 2010;16:5920–2. 14. Carvalho FM, Bacchi LM, Santos PP, Bacchi CE. Triple-negative breast carcinomas are a heterogeneous entity that differs between young and old patients. Clinics (Sao Paulo) 2010;65:1033–6.

15. Carvalho LV, Pereira EM, Frappart L, et al. Molecular characterization of breast cancer in young Brazilian women. Rev Assoc Med Bras 2010;56:278–87. 16. Keegan TH, Derouen MC, Press DJ, Kurian AW, Clarke CA. Occurrence of breast cancer subtypes in adolescent and young adult women. Breast Cancer Res 2012;14:R55. 17. O’Brien KM, Cole SR, Tse CK, et al. Intrinsic breast tumor subtypes, race, and long-term survival in the Carolina Breast Cancer Study. Clin Cancer Res 2010;16:6100–10. 18. Yang XR, Chang-Claude J, Goode EL, et al. Associations of breast cancer risk factors with tumor subtypes: a pooled analysis from the Breast Cancer Association Consortium studies. J Natl Cancer Inst 2011;103:250–63. 19. Thomson CS, Hole DJ, Twelves CJ, Brewster DH, Black RJ. Scottish Cancer Therapy Network. Prognostic factors in women with breast cancer: distribution by socioeconomic status and effect on differences in survival. J Epidemiol Community Health 2001;55:308–15. 20. Gordon NH. Socioeconomic factors and breast cancer in black and white Americans. Cancer Metastasis Rev 2003;22:55–65. 21. Tavassoli FA, Devilee P, eds Pathology and genetics of tumours of the breast and female genital organs. Lyon: IARC, 2003. 22. Elston CW, Ellis IO. Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up. Histopathology 1991;19:403–10. 23. Fitzgibbons PL, Page DL, Weaver D, et al. Prognostic factors in breast cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000;124:966–78. 24. Fleming ID, Cooper JS, Hensen DE, eds AJCC cancer staging manual. 5th edn. Philadelphia: Lippincott-Raven, 1997. 25. Greene FL, Page DL, Fleming ID, eds AJCC cancer staging manual. 6th edn. New York: Springer, 2002. 26. Hammond ME, Hayes DF, Dowsett M, et al. American Society of Clinical Oncology/College Of American Pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Clin Oncol 2010;28:2784–95. 27. Carlson RW, Anderson BO, Bensinger W, et al. NCCN Practice Guidelines for Breast Cancer. Oncology (Williston Park) 2000;14:33–49. 28. Goldhirsch A, Wood WC, Coates AS, et al. Strategies for subtypes-dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011. Ann Oncol 2011;22:1736–47. 29. Dookeran KA, Dignam JJ, Holloway N, et al. Race and the Prognostic Influence of p53 in Women with Breast Cancer. Ann Surg Oncol 2012;19:2334–44. 30. Baak JP, Gudlaugsson E, Skaland I, et al. Proliferation is the strongest prognosticator in node-negative breast cancer: significance, error sources, alternatives and comparison with molecular prognostic markers. Breast Cancer Res Treat 2009;115:241–54. 31. Cronin M, Sangli C, Liu ML, et al. Analytical validation of the Oncotype DX genomic diagnostic test for recurrence prognosis and therapeutic response prediction in node-negative, estrogen receptor-positive breast cancer. Clin Chem 2007;53:1084–91. 32. Gwin K, Pinto M, Tavassoli FA. Complementary value of the Ki-67 proliferation index to the oncotype DX recurrence score. Int J Surg Pathol 2009;17:303–10. 33. Aleskandarany MA, Green AR, Benhasouna AA, et al. Prognostic value of proliferation assay in the luminal, HER2-positive, and triple-negative biologic classes of breast cancer. Breast Cancer Res 2012;14:R3. 34. von Minckwitz G, Untch M, Blohmer JU, et al. Definition and impact of pathologic complete response on prognosis after neoadjuvant chemotherapy in various intrinsic breast cancer subtypes. J Clin Oncol 2012;30:1796–804.