Breast Cancer

, Volume 15, Issue 4, pp 303–308

Is triple negative a prognostic factor in breast cancer?

Authors

    • Department of Breast and Endocrine SurgeryKumamoto City Hospital
  • Nobuyuki Arima
    • Department of Clinical PathologyKumamoto City Hospital
Original Article

DOI: 10.1007/s12282-008-0042-3

Cite this article as:
Nishimura, R. & Arima, N. Breast Cancer (2008) 15: 303. doi:10.1007/s12282-008-0042-3

Abstract

Background

Breast cancer is characterized by hormone dependency, and endocrine therapy is a key treatment in breast cancer. Recently, targeted therapies such as Trastuzumab treatment for HER2-positive breast cancer has been important. Triple-negative (TN) breast cancer is characterized by lack of expression of estrogen receptor (ER) and progesterone receptor (PgR), and the absence of HER2 protein overexpression, and so there is no targeted therapy for this subtype. In this study, we examined the biological and prognostic characteristics in TN breast cancer.

Patients and methods

Between January 1998 and September 2006, 1,552 patients with primary breast cancer were investigated retrospectively in this study and ER, PgR and HER2 status were evaluated in all cases. Furthermore, p53 overexpression and Ki67 values were examined immunohistochemically.

Results

Patient distribution according to ER, PgR or HER2 status was as follows: ER and PgR positive: 57.9%, and ER and PgR negative: 25.1%. With regards to the HER2 status, HER2 positive was 23.3%, and triple negative (TN) was 14.0%. TN breast cancer has a high proliferation rate, high nuclear grade and frequent p53 overexpression. Patients with TN tumors had a significantly poorer disease-free survival (DFS) than those with non-TN tumors. After recurrence the overall survival (OS) rate in TN cases was significantly lower than that of the non-TN cases. Multivariate analysis revealed that TN was a significant factor for DFS and OS after recurrence.

Conclusion

TN breast cancer is a rare subtype with a high proliferation rate and a high nuclear grade, p53 overexpression, and lower DFS/OS. To improve the prognosis of TN breast cancer, a new effective strategy needs to be developed.

Keywords

Breast cancerTriple-negative subtypePrognosisBiology

Introduction

Breast cancer is characterized by hormone dependency, and endocrine therapy is a key treatment in breast cancer. Recently, targeted therapies such as Trastuzumab treatment for HER2-positive breast cancer have been important. Progress of these treatments has been brought on by the developments in biological research for breast cancer. The recent understanding of the molecular basis of breast cancer growth and progression has led to the identification of tumor subtypes with potentially different biologic behavior [13]. Triple-negative (TN) breast cancer is defined by a lack of expression of estrogen and progesterone receptor (ER, PgR), and HER2. To identify an appropriate treatment for TN breast cancer is an important problem. Because of the absence of specific treatment guidelines for this subgroup, TN breast cancer is managed with standard treatment. In this study, we examined the biological and prognostic characteristics in TN breast cancer.

Patients and methods

Patients

Between January 1998 and September 2006, 1,552 patients with primary breast cancer who were operated on in Kumamoto City Hospital were studied, and the ER, PgR receptors and HER 2 status were evaluated in all cases. The mean age of the patients was 55.4 (SD: 13.0) years (range: 26~95). Eight hundred and ninety-eight patients were postmenopausal. The average tumor size was 2.3 cm (range: 0.5~17 cm), and 575 cases were lymph node positive. Out of the 1,552 patients, breast-conserving surgery was performed on 883 patients (Table 1).
Table 1

Patient characteristics—1,552 cases (1998–2006.9)

Age (mean ± SD)

55.4 ± 13.0 years old (range: 26~95 years old)

Menopausal status

Pre-: 654, post-: 898

Tumor size (mean ± SD)

2.3 ± 1.7 cm (range: 0.5~17 cm)

Operation method

 

  Breast-conserving surgery

883 cases

  Mastectomy

669 cases

Nodal status

n0: 960 cases (including SLN(−))

n+: 575 cases, unknown: 17

Estrogen receptor (ER)

Positive: 1,120, negative: 432

Progesterone receptor (PgR)

Positive: 942, negative: 610

HER2 status by IHC

−: 751, 1+: 440, 2+: 123, 3+: 238

IHC immunohistochemistry

Histopathologic and immunohistologic examination

The items investigated were tumor size, presence or absence of lymph node metastasis, hormone receptor (HR) status, proliferation, and p53 overexpression. Immunostaining of ER, PgR, p53, and HER2 was done as previously described [4]. In brief, the slides were incubated at a dilution of 1:75 with anti-ER primary antibody (DAKO Glostrup, Denmark), at 1:700 with anti-PgR primary antibody (DAKO), and at 1:50 anti-p53 primary antibody (Japan Tanner, Osaka, Japan) and HER/neu oncoprotein antibody (DAKO; Herceptest). The DAKO EnVision system (DAKO EnVision labeled polymer, peroxidase) was used as the detection system for ER, PgR, and p53. ER and PgR expression was recorded as the percentage of stained cells and classified into two groups, one being positive (≥10% stained cells) and the other being negative (<10%) as previously reported [5, 6]. However in some cases, ER and PgR were evaluated by using the enzyme immunoassay procedure (EIA: positive if ≥10 fmol/mg protein). MIB-1 (Ki67), a marker of tumor proliferation, was evaluated by immunostaining and expressed as the percentage of positive cells. The proliferative activity of each tumor was classified semiquantitatively into three groups according to the percentage of positive nuclei (median value: 20%): <20%, 20 ≤ <50%, and 50%≤ [4, 7]. As shown in Table 1, the number of ER-positive cases was 1,120, and the number of PgR-positive cases was 942. The HER2 status was as follows: negative in 751 cases, 1+ in 440 cases, 2+ in 123 cases, and 3+ in 238 cases. Among the patients in the 2+ range for HER2, 65.2% (15/23) were positive for HER2 expression as examined by the fluorescence in situ hybridization (FISH) procedure. In this study, HER2 3+ and 2+ were judged as HER2 positive.

Adjuvant therapy

Postoperative adjuvant therapy has been performed based on the recommendation of the St. Gallen International Conference and centers on treatment with anthracycline-based regimens [CEF or EC therapy: cyclophosphamide (CPA) 600 mg/m2, epirubicin 60–75 mg/m2, (5-FU 500 mg/m2)] as chemotherapy. CMF therapy (CPA, methotrexate, and 5-FU) or oral anticancer drugs such as FU derivatives and CPA were given to some patients. Furthermore, taxanes (paclitaxel, docetaxel) were added to CE(F) therapy in some node-positive patients. With regard to endocrine therapy, tamoxifen (TAM) was given in most HR-positive patients, and some of the patients were treated with aromatase inhibitors after 2–3 years from the initiation of TAM. There were no cases with Trastuzumab in adjuvant setting. Trastuzumab was administered to about half of the patients with recurrence lesions.

Table 2 shows the distribution of adjuvant therapy according to TN or non TN. More than 80% of the patients with TN tumors received chemotherapy, and the regimen of chemotherapy was anthracycline-based in about half of them. On the other hand, in patients with non TN tumors, endocrine therapies were often administered. Out of the patients with non TN tumors, 45% of them had no chemotherapy; this was a significant difference in comparison with TN tumors.
Table 2

Adjuvant chemo-endocrine therapy in relation to ER/PgR and HER2 status

Chemotherapy regimen

Triple negative

Non-triple negative

None

39 (17.9)

746 (55.9)

CMF or FU + CPA

40 (18.3)

141 (10.6)

EC or CEF

105 (48.2)

337 (25.3)

Anthracycline + Taxane

34 (15.6)

112 (8.4)

P < 0.0001

P < 0.0001

Endocrine therapy

11 (5.0)

1,085 (81.3)

Total

218

1,334

Statistical methods

For statistical processing, the χ2 test and Fisher’s exact test were used for inter-group comparison with the results shown in Tables 2, 4, and 5. Wilcoxon’s (nonparametric) test was used to compare the mean values for tumor sizes or Ki67 values between the subgroups. The Kaplan-Meier test was used to calculate prognosis [cumulative disease-free and overall survival (OS) rate] and tested by using the log rank procedure. Cox’s proportional hazard model was used to perform univariate and multivariate analyses of the factors related to disease-free survival (DFS) and OS after recurrence. The median observation period was 50 months. During this follow-up period, there were 250 cases with recurrent tumors and 175 deaths due to recurrent tumors. A two-sided P value of <0.05 was considered statistically significant.

Results

Patient distribution according to ER/PgR or HER2 status

As shown in Table 3, patient distribution according to ER/PgR or HER2 status was as follows: ER and PgR positive: 57.9%, ER positive/PgR negative: 14.2%, and ER and PgR negative: 390 cases (25.1%). With regards to the HER2 status, HER2 positivity was 23.3% and TN was 14.0%.
Table 3

Patient distribution according to ER/PgR and HER2 status

 

HER2 (−)

HER2 (+)

Total

ER− PgR−

218 (14.0)

172 (11.1)

390 (25.1)

ER− PgR+

24 (1.5)

19 (1.2)

43 (2.8)

ER+  PgR−

151 (9.7)

69 (4.4)

220 (14.2)

ER+  PgR+

797 (51.4)

102 (6.5)

899 (57.9)

Total

1,190 (76.7)

362 (23.3)

1,552

( ): %/All cases (1,552)

Triple-negative status and clinico-pathological factors

Table 4 shows the clinical or biological features in relation to TN status. There was a significant difference between TN status and tumor size, p53 overexpression, Ki67 and the nuclear grade. In addition, TN tumors were significantly larger, had frequent p53 overexpression, higher Ki67 values, and a higher nuclear grade in comparison with non-TN tumors.
Table 4

Clinical and biological features in relation of triple negative status

 

Triple negative

Non-triple negative

P value

Tumor size (cm)

2.7 ± 1.9

2.3 ± 1.6

<0.0001

Involved nodes

 (−)

136 (63.3)

824 (62.4)

0.89

 (+)

79

496

 

p53 overexpression

 –

67

686

 

 1+

29

405

<0.0001

 2+

121 (55.8)

242 (18.2)

 

MIB-1(Ki67)

 –19%

36

598

 

 –49%

60

601

<0.0001

 50%–

121 (55.7)

135 (10.1)

 

Nuclear grade

 1

22

384

 

 2

112

800

<0.0001

 3

73 (35.3)

121 (9.3)

 
Out of 218 cases of TN breast cancer, ER and PgR status in 204 cases was evaluated by using the immunohistochemistry (IHC) procedure. Table 5 shows the relationship between the ER/PgR-positive cell rate (<1 vs. 1–9%) and the biological features in TN breast cancer. Cases with ER-positive cell rates less than 1% showed higher nuclear grade and higher Ki67 values compared to cases with 1% or more. With regards to PgR, cases with a PgR-positive cell rate less than 1% revealed significantly higher Ki67 values. Furthermore, there was a significant difference in the biological features in terms of hormone dependency even in TN tumors.
Table 5

ER and PgR positive cell rate and biological markers in triple-negative breast cancer

 

ER positve cell rate

P value

PgR positve cell rate

P value

–(<1%)

1–9%

–(<1%)

1–9%

p53

 −

39

14

 

49

4

 

 1+

22

9

0.17

28

3

0.92

 2+

96 (61.1)

19 (45.0)

 

109 (58.6)

11 (61.1)

 

Nuclear grade

 1

12

7

 

15

4

 

 2

40

21

0.0003

54

7

0.065

 3

108 (67.5)

14 (33.3)

 

115 (62.5)

7 (38.9)

 

Ki67 (%)

 Mean ± SD

55.5 ± 27.2

42.9 ± 27.3

0.008

54.5 ± 27.3

36.4 ± 26.7

0.008

Total

162

42

 

186

18

 

Disease-free and overall survival according to TN status

As shown in Fig. 1, patients with non TN tumors had a significantly better DFS than those with TN tumors. The OS rate of patients with TN tumors was significantly lower than those with non TN tumors.
https://static-content.springer.com/image/art%3A10.1007%2Fs12282-008-0042-3/MediaObjects/12282_2008_42_Fig1_HTML.gif
Fig. 1

Triple-negative status and prognosis. a Disease-free survival as a function of triple-negative (TN) status. Patients with non-TN tumors had a significantly better disease-free survival (DFS) than TN tumors. b Overall survival after recurrence. The overall survival rate of patients with TN tumors was significantly lower than those with non-TN tumors. TN triple negative, DFS disease-free survival

In order to clarify the prognostic factors, we evaluated several clinico-pathological factors for DFS and OS after recurrence by uni- and multivariate analysis. As shown in Table 6, menopausal status, tumor size, nuclear grade, nodal status, p53 overexpression, Ki67, and TN status were independently significant factors for DFS. Moreover, in terms of OS, multivariate analysis revealed that Ki67, nodal status, and TN status were significant factors (Table 7). In other words, TN was one of the major contributors for the increase in recurrence and the decrease in survival.
Table 6

Univariate and multivariate analysis of factors for disease-free survival after primary operation

Variables

Category

Univariate

Multivariate

Relative risk

P value

P value

Menopausal status

Post-/pre-

0.61

0.0004

0.001

Tumor size

2.1 cm– / –2.0 cm

3.94

<0.0001

<0.0001

Nuclear grade

Grade 2/1

4.50

<0.0001

0.007

Grade 3/1

6.47

<0.0001

0.04

Nodal status

Positive/negative

5.68

<0.0001

<0.0001

p53

−, +/++

2.53

<0.0001

0.007

Ki67

20%– / –19%

3.57

<0.0001

0.0004

50%– / –19%

4.84

<0.0001

0.01

Triple negative

Non TN/TN

0.49

<0.0001

0.01

Table 7

Univariate and multivariate analysis of factors for overall survival after recurrence

Variables

Category

Univariate

Multivariate

Relative risk

P value

P value

Menopausal status

Post-/pre-

1.34

0.09

0.10

Tumor size

2.1 cm– / –2.0 cm

1.58

0.03

0.36

Nuclear grade

Grade 2/1

0.74

0.42

 

Grade 3/1

1.04

0.93

 

Nodal status

Positive/negative

1.92

0.003

0.008

p53

−, +/++

2.22

<0.0001

0.11

Ki67

20%– / –19%

1.30

0.37

0.50

50%– / –19%

2.67

0.002

0.049

Triple negative

Non TN/TN

0.48

0.0002

0.027

DFI

2 years– / –2 years

0.55

0.006

0.14

5 years– / –2 years

0.31

0.10

0.66

Discussion

Analysis of the patterns of gene expression in breast cancer suggests that it is not a single entity, but is comprised of several biologically distinct subtypes with characteristic molecular profiles. Classification of breast cancers into basal type (TN), luminal, and HER2 has been proposed as a classification scheme based on gene expression profiles. The basal-like subtype is characterized by the low expression levels of the ER-related and the HER2-related group of genes, and therefore is often ‘triple negative’ on clinical assays for these proteins [8, 9]. Using the IHC analysis, tumors can be classified into the following categories: basal-like (TN), luminal (ER/PgR positive), and HER2. In this study, we evaluated prognostic and biological characteristics of TN breast cancer.

The prevalence of TN subtype was 14% in this study, and this was lower than, but similar to the number of 15–20% reported by the Carolina Breast Cancer Study (CBCS) [10]. The apparent characteristic of this subgroup was high proliferation potency. In addition, this subgroup had larger tumors with a higher grade of malignancy, including more frequent p53 overexpression and a higher nuclear grade than the non TN subgroup. In a cohort study [2], only 13% of luminal subtype-A tumors contained p53 mutations, whereas mutations were present in 82% of basal-like cancers (P < 0.001). These subtype lesions are usually high-grade tumors of ductal histology with a high proliferation rate. In addition, the majority of these patients express cytokeratin CK5/6, epidermal growth factor receptor (EGFR), and/or c-kit [1114]. It was reported [9] that TN breast cancer had many similarities to BRCA1-associated breast cancer.

According to the relationship between the ER/PgR-positive cell rate (<1% vs. 1–9%) and biological features in TN breast cancer, there was a significant difference in the biological features. This finding indicates that the ER/PgR-positive cell rates can distinguish the biological characteristics in TN breast cancer and may reflect the clinical course and response to endocrine therapy. The ER-, HER2-, CK5/6+ and/or EGFR+ profile seems to correlate better with basal-like gene expression profiles [12, 15]. Therefore, an analysis of the relationship between these profiles and ER/PgR-positive cell rates may be needed to define the cut-off points of HR in terms of TN breast cancer.

With regards to the prognosis, patients with TN tumors had poorer DFS than those with non-TN tumors. Several histological series of this subgroup have reported poorer clinical outcome data [12, 1618]. It was reported [19] that patients with HER2-positive tumors had higher recurrence rates in ER- and/or PgR-positive cases, but not in ER- and PgR-negative cases. In this study, patients with TN tumors had poorer survival after recurrence than those with non-TN tumors. Trastuzumab might contribute to a good result in HER2-positive tumors after recurrence. Multivariate analysis revealed that TN was a significant prognostic factor for DFS and OS after recurrence.

There is no specific systemic regimen recommended for the treatment of TN breast cancer. In general, chemotherapy is effective for patients with endocrine non-responsive or highly proliferative tumors [20]. TN tumors might benefit more from chemotherapy than endocrine-responsive tumors [21]. Neoadjuvant chemotherapy using adriamycin plus CPA (AC) followed by paclitaxel revealed that a pathological complete response (pCR) was seen in 45% of basal-like cancer, 45% of ERBB2-positive cancer, and only 6% of luminal cancer [22]. A subset analysis of CALGB 9344 [23] suggested that the impact of taxanes added to adjuvant AC was observed in the ER-negative subtypes (whether HER2-positive or -negative) and the ER-positive subtype if also HER2-positive. From these results, an appropriate treatment for TN cancer might be a standard regimen such as anthracyclines and taxanes. Siziopikou KP et al. [14] propose that the majority of the “triple-negative” patients have basal subtype tumors with high EGFR expression and that these tumors may be the subgroup of breast cancers that could potentially benefit the most from novel EGFR-targeted therapeutic strategies. However, there is no evidence for the efficacy of antibody treatment such as cetuximab to date. A randomized phase-III trial of bevacizumab combined with weekly paclitaxel suggested that TN breast cancer benefited from the addition of bevacizumab [24]. In fact, there is no role for endocrine- or HER2-targeted therapy for TN breast cancer. Therefore, a specific therapy for TN cancer, such as targeted or non-targeted therapy, is needed.

In conclusion, TN breast cancer is a rare subtype and has high proliferation rates and a high nuclear grade, p53 overexpression, and a lower DFS/OS. TN breast cancer does not have a so-called targeted treatment; however, standard chemotherapy might be effective in TN breast cancer. TN was one of the significant factors for DFS and OS after recurrence in breast cancer, although most TN patients were treated with standard chemotherapy. To improve the prognosis of TN breast cancer, a new effective strategy is necessary.

Copyright information

© The Japanese Breast Cancer Society 2008