Adjuvant chemotherapy for patients with lymph node-negative breast carcinoma is being recommended currently based on the St. Gallen classification. The prognostic importance of HER-2 status in patients with lymph node-negative breast carcinoma has been investigated extensively, with contradictory results. The authors investigated the clinical relevance of HER-2 overexpression when combined with the St. Gallen classification in lymph node-negative breast carcinoma.
The medical records of patients with breast carcinoma negative for lymph node involvement who underwent surgery between January 1995 and December 2000 at the Seoul National University College of Medicine (Seoul, Korea) were reviewed retrospectively. Risk groups based on the St. Gallen classification were categorized as average or minimal risk. The prognostic values of HER-2 in combination with the St. Gallen classification were analyzed with respect to disease-free survival (DFS) rates.
A total of 906 patients were eligible for analysis. The overall 7-year DFS rate was 87.5%. The 7-year DFS rates for patients with HER-2–positive and HER-2–negative tumors were, respectively, 77.9% and 91.2% (P = 0.002). The 7-year DFS rates for patients with average and minimal risk group were 85.0% and 97.9%, respectively. The authors found that HER-2 overexpression significantly predicted the risk of disease recurrence (odds ratio = 3.03 [95% confidence interval, 1.63–5.63]). Furthermore, when HER-2 status was combined with the St. Gallen classification, the DFS rate of the HER-2–positive average risk group was 73.3% compared with 88.4% for the HER-2–negative average risk group (P = 0.007).
Women with axillary lymph node-negative breast carcinoma have a good prognosis. However, approximately 20% of individuals will experience disease recurrence and die of systemic disease. Although some patients with negative axillary lymph nodes may benefit from adjuvant chemotherapy, large numbers of women must be treated to benefit those destined to experience disease recurrence. Thus, the definition of the risk profiles of individual patients at the initial prognosis and the identification of appropriate optimal systemic treatment modalities remain a major challenge.1, 2
Currently, adjuvant chemotherapy for patients with lymph node-negative breast carcinoma is usually administered based on the St. Gallen classification,3, 4 which divides women with lymph node-negative breast carcinoma into two groups—an average-risk group and a minimal-risk group. The average-risk group comprises patients with endocrine-nonresponsive (estrogen receptor [ER] negative and progesterone receptor [PgR] negative) and endocrine-responsive disease (ER negative and/or PgR positive, and at least one of the following features: pT > 2 cm, histologic Grade 2–3, or age < 35 years). According to these guidelines, > 80% of all patients with lymph node-negative breast carcinoma should receive adjuvant systemic therapy.5 However, the average-risk group is so broad and heterogeneous that some patients survive long term without adjuvant therapy, whereas others experience disease recurrence. Thus, it has been suggested that the St. Gallen classification is not a satisfactory way of predicting prognosis for patients with lymph node-negative breast carcinoma, and many studies have been conducted to determine more convincing guidelines.5–7
Over the past 30 years, a number of potential prognostic markers related to the biologic processes of cancer cells have been investigated extensively.8–11 Among them, HER-2 has realized clinical importance over the last few years.12–14 HER-2, an oncogenic transmembrane growth factor receptor, has been shown to be amplified or overexpressed in 20–40% of patients with primary invasive breast carcinoma.15–17 An extensive body of evidence indicates that a positive HER-2 status is correlated with a poor outcome.18, 19 However, the prognostic value of HER-2 overexpression or amplification in patients with lymph node-negative breast carcinoma has been controversial.17, 19, 20
Most previous studies of HER-2 in lymphnode-negative breast carcinoma focused on the independent clinical relevance of HER-2 overexpression. Like any single prognostic marker, HER-2 overexpression dose not predict all disease recurrences, so the consideration of HER-2 overexpression with other prognostic factors is important. The current study was undertaken to identify patients with lymph node-negative breast carcinoma who were most at risk of developing disease recurrence and to determine prognostic factors using the combination of the St. Gallen classification and HER-2 overexpression.
MATERIALS AND METHODS
The study population consisted of 906 consecutive patients who underwent surgery for primary breast carcinoma between January 1995 and December 2000 at the Seoul National University Hospital. Patients with pathologically proven axillary lymph node-negative breast carcinoma were eligible. Patients with a T4 tumor or with pure ductal carcinoma in situ were excluded. Clinicopathologic data and follow-up information were retrieved from medical records. Primary treatment consisted of either radical or modified mastectomy or quadrantectomy followed by radiotherapy. All patients received a complete axillary lymph node dissection.
St. Gallen Classification
Patients were placed in the the average-risk group or the minimal-risk group based on the St. Gallen classification published in 2003.3 Histologic grade was assessed using the Elston–Ellis modification of the Bloom–Richardson grading system. ER and PgR status was defined by immunohistochemistry (IHC).
The average-risk group comprised patients with at least one of the following features: ER/PgR negative, pT > 2 cm, histologic Grade 3, or age < 35 years. Our classification differed from the St. Gallen classification, as patients with histologic Grade 2 disease without another risk factor were not included in the average-risk group. In our study, including such patients would have caused the allocation of 96% of classifiable patients to the average-risk group. Moreover, the classification of patients with a histologic Grade 2 tumor to the minimal-risk group did not have a significant effect on patient survival in previous studies.5, 21
HER-2 Immunohistochemical Staining
HER-2 protein expression was detected by IHC using NCL-CB-11 (mouse antihuman immunoglobulin G1 [IgG1] monoclonal antibody; Novocastra Laboratories Ltd., New Castle-Upon-Tyne, U.K.). HER-2 staining was observed by light microscopy. Membrane staining was evaluated for HER-2 protein expression, whereas cytoplasmic staining was ignored. The intensity of membrane staining was assessed using the Dako HerceptTest scoring system (Dako Corporation, Carpinteria, CA): 0, ≤ 10% of tumor cells showed membrane staining; 1+, partial membrane staining in ≥ 10% of tumor cells; 2+, weak to moderate complete staining in ≥ 10% of tumor cells; and 3+, strong complete membrane staining in ≥ 10% of tumor cells. HER-2 data were available for 693 patients, and 182 tumors had a staining intensity score of 3+, 81 tumors had a score of 2+, 115 tumors had a score of 1+, and 315 tumors had a score of 0. Because previous studies had shown a poor correlation between weakly positive IHC results 2+ and HER-2 gene amplification, but an excellent correlation between highly positive IHC results 3+ and HER-2 gene amplification, tumors with 3+ in IHC were classified as positive for HER-2 overexpression.22
Statistical analysis was performed using SPSS 11.0 for Windows (SPSS Inc., Chicago, IL). The role of HER-2 overexpression and the usefulness of other clinicopathologic markers as prognostic indicators were investigated in terms of disease-free survival (DFS). The relevance of HER-2 expression in combination with the St. Gallen categories was also evaluated. The DFS rate after surgery was computed using the Kaplan–Meier product limit method. DFS curves were compared using the log-rank test. The first occurrence of locoregional or distant disease recurrence or of contralateral breast carcinoma, was considered disease recurrence. Seven-year DFS rates and their 95% confidence intervals (95% CI) were calculated.
All lymph node-negative patients who enrolled in the current study were observed regardless of whether a tumor specimen was available for clinicopathologic or molecular analysis (N = 906). The median patient age was 46 years (range, 20–89 years). Fifty-three percent received some form of systemic therapy, 34% received chemotherapy only, 11% received hormonal therapy only, and 8% received both chemotherapy and hormonal therapy. We identified 301 patients who received 6 courses of CMF (cyclophosphamide, methotrexate, 5-fluorouracil [5-FU]), 14 patients who received 6 courses of CAF (cyclophosphamide, doxorubicin, 5-FU), 64 patients who received 4 courses of AC (doxorubicin, cyclophosphamide), 4 patients who received < 4 courses of CMF, and 1 patient who received only 1 course of CAF.
As of November 1, 2003, the median follow-up period was 56 months and 72 patients developed disease recurrence. The 7-year DFS rate for all patients was 87.5%. Tumor specimens with HER-2 overexpression represented 27% of tumor specimens with available HER-2 data. Table 1 displays the clinicopathologic characteristics of the study population.
Table 1. Characteristics of Patients and Univariate Analysis of Possible Prognostic Variables for 7-Year DFS
Of the prognostic variables entered into the univariate analysis, HER-2 overexpression, ER status, PgR status, and tumor size correlated with DFS. By multivariate analysis, after adjusting for tumor size, histologic grade, hormonal receptor status, and age, only HER-2 status showed independent prognostic relevance (odds ratio [OR] = 3.03; 95% CI, 1.63–5.63) (Table 2).
Table 2. Prognostic Factors for 7-Year DFS by Multivariate Analysis
By applying our less stringent St. Gallen classification, 781 patients (86%) were classifiable, and 624 patients (80% of the classifiable patients) were included in the average-risk group. The 7-year DFS rates of the average-risk and minimal-risk groups were 85.0% and 97.9%, respectively (P = 0.001; Table 3). These results confirmed the clinical relevance of the St. Gallen classification in patients with lymph node-negative breast carcinoma. Four hundred fifteen patients (67% of the average-risk group) received adjuvant chemotherapy and/or hormonal therapy, whereas 209 patients in the average-risk group did not receive chemotherapy or hormonal therapy. The 7-year DFS rate of this untreated subgroup was 87.7 %. No difference was observed between 7-year DFS rates of treated and untreated subgroups in the average-risk group (83.3% vs. 87.7%, P = 0.93).
Table 3. Seven-year DFS According to the St. Gallen Classification
St. Gallen classification
No. of patients (%)
7-yr DFS rate (%)
DFS: disease-free survival.
We evaluated the relation between HER-2 expression and the St. Gallen classification. Of the patients in the average-risk group with available HER-2 data, 141 (27.8%) had HER-2–positive tumors and 367 (72.2%) had HER-2–negative tumors. Of the patients in the minimal-risk group with available HER-2 data, 36 (23.8%) patients had HER-2–positive tumors and 115 (76.2%) had HER-2–negative tumors. HER-2 expression was not related to the St. Gallen classification (P = 0.92; Table 4).
Table 4. The Relation between HER-2 Expression and the St. Gallen Classification (P = 0.92)
St. Gallen classification
Next, we analyzed the DFS rates of subgroups with or without HER-2 overexpression in the St. Gallen risk groups. The 7-year DFS rate was 73.3% for patients with HER-2–positive tumors and 88.4% for patients with HER-2–negative tumors among average-risk patients (P = 0.007; Table 5). Remarkably, when we combined HER-2 expression and the St. Gallen classification, we found a twofold elevated risk of disease recurrence when tumor specimens showed HER-2 overexpression and average risk characteristics synchronously, compared with those that had only average-risk characteristics (26.7% vs. 11.6%, P = 0.007).
Table 5. Seven-Year DFS According to HER-2 Expression in the St. Gallen Risk Group
No. of patients (%)
7-yr DFS rate (%)
DFS: disease-free survival.
Based on our observations, we further evaluated the clinical relevance of HER-2 overexpression in association with the St. Gallen classification, by subgrouping patients with lymph node-negative breast carcinoma into average risk with HER-2 overexpression, average risk without HER-2 overexpression, and minimal risk. Six hundred sixty-eight patients were classifiable with available HER-2 data. Respectively, 21%, 54%, and 21% of patients belonged to the average-risk with HER-2 overexpression group, the average-risk without HER-2 overexpression group, and the minimal-risk group. The corresponding 7-year DFS rates were 73.3%, 88.4%, and 97.9% (P = 0.0001; Fig. 1).
The central purpose of the current study was to help guide therapeutic choice for patients with lymph node-negative breast carcinoma. Two key problems remain unresolved for patients with lymph node-negative breast carcinoma, namely, who should receive adjuvant chemotherapy and what adjuvant chemotherapy should be prescribed. Adjuvant treatment for patients with lymph node-negative breast carcinoma is currently being recommended by the St. Gallen classification. However, decisions regarding adjuvant chemotherapy according to the established St. Gallen classification may cause overtreatment or undertreatment in clinical practice.5, 23 For example, in our retrospective study, 182 patients (90% of untreated patients in the average-risk group) survived without disease recurrence. Thus, it can be deduced that > 80% patients in the average-risk group would have survived without adjuvant therapy, which provides substance for the need to consider modifying the St. Gallen classification.
Pathologic HER-2 overexpression means that thousands of HER-2 receptor molecules are present on the cell surface, and that growth signals are being transmitted continuously to the nucleus, even in the absence of ligands. This is a biologic description, but from a clinical point of view some shadowy areas remained. This molecular mechanism of HER-2 expression could, in fact, account for the prognostic role of this biomarker in breast carcinoma, but much literature data fail to unequivocally confirm this hypothesis, especially in patients with lymph node-negative breast carcinoma.9, 24–27 Mirza et al.9 reviewed published studies on HER-2 expression in patients with lymph node-negative breast carcinoma. They showed that HER-2 expression was a positive prognostic factor in 6 of 13 studies by univariate analysis, and in 2 of 6 studies by multivariate analysis.
For patients with lymph node-negative breast carcinoma, the current accepted prognostic factors include clinicopathologic (e.g., age, tumor size, and histologic grade) and biologic characteristics (e.g., ER/PgR status and HER-2 expression). With the molecular approaches, the importance of molecular prognostic markers in breast carcinoma has been increased. Furthermore, very promising results have been obtained by the simultaneous evaluation of thousands of genes which are significant in terms of prognosis, as described by van de Vijer et al.6 However, few of the many putative biomolecular prognosticators identified are currently used in clinical practice. The main reasons for the lack of a translation of laboratory findings to clinical practice are related to the way in which many experimental data for molecular prognostic factors have not been evaluated in combination with clinicopathologic data.
Among numerous molecular prognostic factors, HER-2 overexpression has attracted the greatest interest as a potential prognostic marker in breast carcinoma. However, the prognostic relevance of HER-2 expression, especially when considered alone, is open to debate. We found very few studies in which HER-2 overexpression had been evaluated in combination with clinicopathologic data in patients with lymph node-negative breast carcinoma.13
We tested the prognostic value of HER-2 expression, and found that HER-2 overexpression has an independent clinical relevance by univariate and multivariate analyses. Analysis using a combination of the St. Gallen classification and HER-2 status permitted us to attribute discriminating prognostic power to HER-2 expression within the average-risk group. Patients with HER-2–overexpressing tumors showed a risk of disease recurrence that was twice that of tumors not overexpressing HER-2 in the average-risk group. Our results give some insight into the heterogeneous nature of the average-risk group, and suggest that lymph node-negative breast carcinoma can be reclassified into an average-risk with HER-2 overexpression group, an average-risk without HER-2 overexpression group, and a minimal-risk group.
The 7-year DFS rate of the average-risk with HER-2 overexpression group was lowest among the patients with lymph node-negative breast carcinoma and may be near that of patients with lymph node-positive breast carcinoma with a low tumor burden. Thus, we suggest a new guideline for treatment according to our clinicopathologic and molecular-based classification of patients with lymph node-negative breast carcinoma (Fig. 2). Anthracycline-based chemotherapy in patients with lymph node-negative breast carcinoma is controversial, because it is toxic and these patients have more reasonable prognostic features. However, recent studies have shown the superiority of the doxorubicin-containing CAF regimen compared with CMF as adjuvant chemotherapy in patients with lymph node-negative breast carcinoma.28 The average-risk with HER-2 overexpression group should receive intensive adjuvant therapy with an anthracycline-based regimen and the average-risk without HER-2 overexpression group should receive the less toxic nonanthracycline-based chemotherapy.
HER-2 status and the St. Gallen classification seem to provide independent and complementary prognostic information for patients with lymph node-negative breast carcinoma. Moreover, intensive adjuvant chemotherapy may be of more benefit to average-risk patients with HER-2 overexpression. However, the clinical use of HER-2 status in association with the St. Gallen classification requires confirmation by a large prospective study.