Occult hematogenous micrometastases are the major cause for metastatic relapse and cancer-related death in patients with operable primary breast cancer. Although sensitive immunocytochemical and molecular methods allow detection of individual breast cancer cells in bone marrow (BM), a major site of metastatic relapse, current detection techniques cannot discriminate between nonviable shed tumor cells and seminal metastatic cells. To address this problem, we analyzed the relevance of erbB2 overexpression on disseminated cytokeratin-18-positive breast cancer cells in the BM of 52 patients with locoregionally restricted primary breast cancer using immunocytochemical double labeling with monoclonal antibody 9G6 to the p185erbB2 oncoprotein. Expression of p185erbB2 on BM micrometastases was detected in 31 of 52 (60%) patients independent of established risk factors such as lymph node involvement, primary tumor size, differentiation grade, or expression of p185erbB2 on primary tumor cells. After a median follow-up of 64 months, patients with p185erbB2-positive BM micrometastases had developed fatal metastatic relapses more frequently than patients with p185erbB2-negative micrometastases (21 versus 7 events; P = 0.032). In multivariate analysis, the presence of p185erbB2-positive micrometastases was an independent prognostic factor with a hazard ratio of 2.78 (95% confidence interval, 1.11–6.96) for overall survival (P = 0.029). We therefore conclude that erbB2 overexpression characterizes a clinically relevant subset of breast cancer micrometastases.

The human erbB2 proto-oncogene encodes for a Mr 185,000 transmembrane glycoprotein receptor (p185erbB2) that shares sequence homology with the epidermal growth factor receptor (1, 2). This protein has intracellular tyrosine kinase activity and an extracellular ligand-binding region (2). Although a specific ligand for p185erbB2 has not been identified, several glycoproteins have been identified that interact with the p185erbB2 receptor (3, 4, 5). Amplification of erbB2 and overexpression of p185erbB2 occur in approximately 15–30% of human primary breast cancers (6, 7, 8). Based on primary tumor tissue analysis, several studies suggest a direct correlation between p185erbB2 overexpression and metastatic relapse (6, 7, 8, 9, 10), whereas some studies have failed to do so (11, 12, 13). Technical issues concerning assay variability may be partly responsible for these discrepant results. However, analysis of primary tumors is only an indirect way to assess the extent of occult disease spread, which is the source of subsequent metastatic relapse.

In a recent study, we were able to demonstrate the prognostic relevance of occult metastatic cancer cells in BM3 using a standardized immunoassay for CK (14). In BM, CKs are expressed exclusively on invading epithelial tumor cells but not on autochthonous mesenchymal BM cells (15, 16). Convincing data exist that patients with BM micrometastases can still survive for extended periods of time and may not even experience tumor relapse at all (14, 17, 18, 19). Our present analysis further supports this hypothesis by demonstrating in patients with operable stage I–III breast cancer that p185erbB2 overexpression is an important characteristic of seminal metastatic tumor cells in the BM. This finding has important implications not only for the assessment of patient prognosis but also for the prediction of response to adjuvant anticancer therapy aimed at eliminating micrometastatic disease. p185erbB2 has become an important target for antibody-based biological therapy (20, 21, 22, 23), and the expression of this oncogene has been shown to predict response to chemotherapeutic treatment in breast cancer patients. Thus, the characterization of BM micrometastases will allow a better understanding of tumor progression and response to therapy in breast cancer.

Patients.

Patients admitted to the I. Frauenklinik, Ludwig-Maximilians University (München, Germany) and the II. Medizinische Klinik, Zentralklinikum (Augsburg, Germany) between February 1987 and November 1992 were included in this prospectively planned study. Patients with histopathologically confirmed breast cancer had either breast-conserving therapy (n = 32) or modified radical mastectomy (n = 20). Complete tumor resection with negative margins was ensured in all cases. Axillary dissection of levels I and II was performed in all patients, whereas level III was only excised in case of macroscopic metastatic involvement of the lower levels. Overt distant metastasis (stage IV disease) was an exclusion criterion, and adequate preoperative and postoperative imaging procedures (i.e., chest X-ray, liver ultrasound, and bone scan) ensured that only stage I–III breast cancer patients were included. In addition, detection of disseminated tumor cells in BM at the time of diagnosis was required for inclusion in this study. For this purpose, patients underwent BM aspiration from both upper iliac crests and the sternum, a procedure approved by the institutional ethical boards, after signed informed consent was received and before treatment.

Adjuvant treatment was applied according to international standards. In all patients treated by breast conservation, local telecobalt radiation therapy was administered. The median absorbed dose in the target area was either 50.0 Gy in 25 fractions or 50.4 Gy in 28 fractions in patients receiving concomitant chemotherapy. After locoregional therapy, patients (n = 7) with 1–3 involved axillary lymph nodes received six cycles of CMF every 21 days. In patients with ≥4 involved regional lymph nodes (n = 12), four courses of EC and three courses of CMF were given sequentially. Postmenopausal patients with lymph node involvement and positive estrogen receptor expression (n = 12) received 20–30 mg of tamoxifen daily for 2–5 years. At the time of patient recruitment, systemic treatment of node-negative patients was not generally recommended; however, five node-negative patients received either endocrine (n = 2) or cytotoxic chemotherapy (n = 3).

BM Preparation and Immunostaining.

The procedure for preparation of mononucleated cells from BM aspirates was performed exactly as described previously (16). In brief, after centrifugation using a Ficoll-Hypaque density gradient (density 1.077 g/mol) at 900 × g (30 min), mononucleated interphase cells were washed, and at least 8 × 105 cells were reproducibly centrifuged onto glass slides at 150 × g (5 min).

Micrometastatic cancer cells were then detected immunocytochemically using monoclonal antibody CK2 (Boehringer Mannheim, Mannheim, Germany) directed against CK18 at 2.5 μg/ml (24). For exact quantification, we analyzed a standardized sample size of 4 × 105 BM cells/patient. The specific antibody reaction was developed with the alkaline phosphatase anti-alkaline phosphatase (APAAP) technique and the Newfuchsin method to visualize specific antibody binding (25) as described previously (16). A standardized sample size of 4 × 105 BM cells/patient was analyzed, and CK18-positive cells were quantified. A BM sample was scored positive if one or more CK18-positive cells were detected.

In all positive cases, cancer cells detected in a second sample of 4 × 105 BM cells were phenotyped applying a previously described immunoenzymatic double labeling procedure as a reference method (26). At optimized concentrations, we used monoclonal antibody 9G6 (Dianova, Hamburg, Germany) directed against the extracellular domain of p185erbB2 oncoprotein (11). Because antigens localized in two separate compartments were detected, possible interference of two-color reactions as a source of false-negative results was minimized. To confirm the results of the our immunoenzymatic double labeling procedure, we also used an alternative double labeling protocol based on a combination of alkaline phosphatase and highly sensitive immunogold techniques (26), thus allowing clear discrimination between two signals colocalized within the same cell using epipolarized light illumination (27). A CK18-positive cell was considered erbB2 positive if a distinct extracellular staining signal was identified. Two experienced and independent observers (S. B. and K. P.), read all slides, with an interobserver concordance of more than 90%.

CK18- and p185erbB2-expressing cells of the human SKBr-3 mammary adenocarcinoma tumor cell line were seeded in mesenchymal cells of the human U-937 histiocytoma tumor cell line and then used as positive and negative immunostaining controls together with the patient BM specimens. An unrelated mouse myeloma antibody (MOPC21; Sigma, Deisenhofen, Germany) served as the IgG1 isotype staining control for the patient BM aspirates.

erbB2 overexpression on the routinely formalin-fixed, paraffin-embedded primary tumor tissue was determined by an automated standardized immunoassay using monoclonal antibody CB11 (Novocastra Laboratories, Newcastle, United Kingdom) directed against the internal domain of the p185erbB2 oncoprotein (28) and the avidin-biotin peroxidase staining technique (Ventana Medical Systems, Tucson, AZ). Staining intensity (0 to 3+) and the number of stained cells (≤10%, 11–50%, 51–80%, and >80%) were evaluated according to the United States Food and Drug Administration-approved score.

Statistical Analysis.

Data quality was optimized by controlling all reported histopathological and immunocytochemical results and by comparing reported events during follow-up with original patient files. The primary end point was overall survival, measured from the date of surgery to the time of the last follow-up or death. Follow-up findings were confirmed in all patients as of February 1, 1999. Kaplan-Meier curves were used for estimation of overall survival (29). Life table curves of the patients with and the patients without p185erbB2 expression on BM micrometastases were compared by the log-rank test. For survival analysis, only cancer-related deaths were considered; data on patients who were still alive at the end of our study were censored. We used a Cox multiple regression analysis to estimate the simultaneous prognostic effect of variables (30). To compare categorical variables, we used Fisher’s exact test. The differences of means of independent samples with continuous variables were calculated by the Mann-Whitney U test. All statistical analyses were performed in a two-tailed test at a significance level of P = 0.05 (SPSS 6.1.1 software package for Macintosh).

Distribution of p185erbB2 Expression on BM Micrometastases.

BM samples with CK18-positive cells from 52 breast cancer patients with stage I–III disease were available for double labeling. Because both double-staining protocols produced congruent results of p185erbB2 positivity on BM micrometastases (26), we were able to pool the data obtained by the immunoenzymatic and the immunogold-alkaline phosphatase technique. In 60% of our patients (31 of 52), micrometastases expressing p185erbB2 were detected (Table 1). The absence of p185erbB2-positive BM cells surrounding CK-positive/p185erbB2-positive tumor cells as well as the absence of false positive signals on consecutive cytospins stained with an isotype control antibody ensured the specificity of our findings. Representative examples of immunocytochemical double labeling of CK18 and p185erbB2 are shown in Fig. 1. The mean percentage of CK18/p185erbB2 double-positive cells among CK18-positive tumor cells was 72% (range, 25–100%). Expression of p185erbB2 on CK18-positive tumor cells was associated with an increased mean and median number of tumor cells present in BM (Fig. 2).

Lack of Correlation to Established Risk Factors.

Table 1 shows the clinical characteristics of the 52 patients who were analyzed for the expression of p185erbB2 on CK18-positive tumor cells in BM. Mean patient age at the time of diagnosis was 56 years. The size of the primary tumor was not associated with the presence of p185erbB2-positive micrometastases. We also did not find a significant correlation between the differentiation grade of the primary tumor and expression of p185erbB2 on CK18-positive tumor cells (P = 0.52). Presence or absence of regional lymph node metastasis was well balanced in our study population and was not associated statistically (P = 0.79) with detection of p185erbB2-positive micrometastases.

Comparison of Primary Tumor and Micrometastatic Cells.

Primary tumor specimens were available for immunohistochemical analysis in a representative subset of 24 patients (Table 2). The samples were analyzed using a different monoclonal antibody (CB11) than the BM samples because antibody 9G6 is not suitable for paraffin-embedded tumor sections. Only seven (29%) primary tumor specimens displayed an intense immunostaining with monoclonal antibody CB11 rated 3+ according to the United States Food and Drug Administration-approved grading system (Table 2). This 3+ score is known to correlate well with erbB2 overexpression and gene amplification. The fraction of p185erbB2-positive cancer cells varied on the primary tumor section between 11% and >80% and on BM cytospin preparations between 0% and 100% (Table 2). No correlation was found between the erbB2 staining score of the primary tumor and the presence of CK-positive/p185erbB2-positive cells in BM: 11 of 17 (65%) patients with none/weak (0/1+) or intermediate (2+) immunostaining had p185erbB2-positive BM micrometastases compared to 4 of 7 (57%) patients with intense (3+) immunostaining of their primary tumors (Table 2).

Presence of p185erbB2-positive Micrometastatic Cells and Unfavorable Prognosis.

After a median observation time of 64 months (range, 8–143 months), follow-up information was available for all 52 patients. A tumor relapse occurred in 29 of the 52 (56%) patients. Twenty-eight (97%) of these patients had disease recurrence with distant metastases, whereas one patient had a local tumor relapse only. Of the 21 patients with p185erbB2 overexpression on CK18-positive tumor cells, 18 patients had bone metastases in contrast to only 3 of the 7 patients with p185erbB2-negative metastatic tumor cells (P = 0.043).

At the time of follow-up, 20 (65%) cancer-related deaths had occurred in the 31 patients with p185erbB2-positive BM micrometastases compared to only 7 (33%) deaths among the 21 patients with p185erbB2-negative tumor cells (P = 0.047; Table 3). Of the 25 patients who were alive at the end of our study, all 14 patients with p185erbB2-negative tumor cells had no evidence of disease, whereas 2 (18%) of the 11 patients with p185erbB2-positive micrometastases had relapsed. As shown in Fig. 3, patients with p185erbB2-positive micrometastases had a worse survival than patients with p185erbB2-negative tumor cells; the adjusted survival rate at 5 years was 13% compared to 67%, respectively (P = 0.032 by the log-rank test).

A Cox multiple regression analysis was done to evaluate whether the detection of p185erbB2-positive BM tumor cells predicts overall survival independently of age, tumor size, differentiation grade, systemic treatment, and metastatic lymph node involvement. Only data stratified for type of systemic treatment were entered into the multivariate regression model to control for a possible bias by administration of adjuvant hormone or cytotoxic therapy to 12 and 19 patients of our study population, respectively, whereas 21 patients received no adjuvant systemic therapy. As shown in Table 4, overexpression of p185erbB2 on micrometastases was a stronger independent predictor of decreased survival (P = 0.029) than the presence of lymph node metastases (P = 0.039). None of the other variables tested in the model reached significance. Because the mean number of both CK18-positive tumor cells and CK18-positive clusters was significantly higher in patients with such cells than in patients without p185erbB2-positive BM micrometastases (Fig. 2), we investigated the influence of this variable in a second model (data not shown). In this model, overexpression of p185erbB2 on micrometastases remained the only independent predictor for decreased survival (P = 0.034; relative risk, 2.69).

This study suggests that overexpression of the erbB2 oncogene on occult metastatic cells in BM might determine their fate. Phenotyping of occult metastatic cancer cells may therefore add clinically relevant information beyond the mere identification of these cells by histogenetic markers such as CK18. This conclusion is also supported by the recent report of Untch et al.(31) demonstrating that the mere presence of CK18-positive cells in BM has no prognostic significance in breast cancer.

erbB2 expression in normal BM cells has recently been shown by detection of erbB2 mRNA in BM of noncarcinoma patients (32). Thus, it is obvious that erbB2 expression alone is not tumor specific and that a double labeling approach is therefore mandatory to ascertain the epithelial origin of p185erbB2-positive cells in BM. The double labeling approach also revealed important biological information: p185erbB2 expression on CK18-positive cells was associated with a significantly reduced survival rate at 5 years, with 67% overall survival in patients with p185erbB2-negative metastatic cells compared to 13% in patients with p185erbB2-positive micrometastases. The underlying cause of this poor prognosis might be a synchronous hematogenous spread of p185erbB2-positive tumor cells to other organs, such as liver and lungs, which signals a more aggressive tumor phenotype with greater potential for metastases to multiple sites. The recurrence in bone, in particular, was correlated with the presence of p185erbB2-positive metastatic cells in BM. This observation supports the supposition that among the occult metastatic cells in BM, precursor cells of later bone metastases carry erbB2 gene amplifications, mediating a survival advantage for these tumor cells within the BM microenvironment. This conclusion is also supported by our previous report demonstrating that p185erbB2 overexpression on BM tumor cells was present in 100% of patients with metastatic breast cancer (stage M1; Ref. 26). Further evidence for an involvement of erbB2 in the outgrowth of occult metastatic breast cancer cells derives from the observation that this oncogene is expressed on several metastatic breast cancer cell lines established from the BM of breast cancer patients (33).

Micrometastatic breast cancer cells displayed a distinctively higher frequency of p185erbB2 overexpression (60% in this study) than that of the primary tumor tissue in our series (29%) and that reported in the recent literature (6, 7, 8, 9, 10, 11). Although careful interpretation of our data is necessary in view of the relatively small number of cases and the relatively low number of micrometastases available for immunophenotyping, the results suggest a preferred selection of p185erbB2-positive micrometastases during tumor progression. Such a selection may be due to an influence of erbB2 on the integrity of intercellular connections. Interactions of homotypic epithelial adhesion molecules and extracellular matrix proteins are modulated by p185erbB2(34, 35, 36), which may enhance the detachment of single tumor cells from the primary tumor. The fact that this detachment is an early event in tumor progression may also explain why p185erbB2 overexpression on metastatic tumor cells did not correlate to clinical signs of tumor progression, i.e., increased size or higher grade of the primary tumor and metastatic lymph node involvement.

In view of the small size of our study population and the ongoing discussion on technical issues concerning detection of erbB2 amplification, methodological variables need to be considered as potentially confounding factors in our study. Our rate of 29% of primary tumors with intense (3+) p185erbB2 expression (Table 2) is consistent with previous reports on larger cohorts of unselected patients (6, 8). Utilization of a second antibody (CB11) for the studies on primary tumor tissue was necessary for technical reasons but may imply a potential source of incongruent results. However, Press et al.(28) demonstrated that both antibodies (9G6 and CB11) provide a 100% specificity and a comparable sensitivity (47% and 51%, respectively) with regard to detection of erbB2 overexpression. Thus, both antibodies used in our study identify erbB2 overexpression with comparable accuracy.

Previous studies have shown that micrometastatic tumor cells in BM of patients with various types of epithelial cancers are characterized by the expression of urokinase-type plasminogen-activator receptor (37), absence of p53 mutations (38), loss of MHC class I antigens (39), or absence of active proliferation at the time of primary diagnosis (26). Thus far, it remains unclear how these biological properties influence the fate of micrometastatic breast cancer cells. To our best knowledge, there are no other groups that have investigated HER2 expression of bone marrow micrometastases. In a small series of breast cancer patients (n = 19), Müller et al.(40) described a high frequency of erbB2 gene amplifications in BM micrometastases. Because no clinical follow-up data were provided for the study population, the clinical significance of their finding remained unclear. Our data on the biological relevance of p185erbB2 expression on disseminated breast cancer cells is supported by the recent observation of Brandt et al.(41). In this study, 19 of 29 patients had CK/p185erbB2 double-positive clustered cells in their peripheral blood. Results of in vitro motility experiments using single and clustered cells from primary breast cancer tissue strongly supported the assumption that CK/p185erbB2 double-positive clustered cells have a high potential for locomotion. From these data, they concluded that circulating p185erbB2-positive tumor cell clusters might be the potential precursors of distant metastases. Our present data now show for the first time that p185erbB2 overexpression characterizes a clinically relevant subset of established metastatic breast cancer cells in a secondary organ relevant for subsequent overt organ metastases.

The observed correlation between p185erbB2 expression on micrometastatic tumor cells in BM and poor prognosis may be biased by the fact that p185erbB2-expressing tumor cells seem to respond differently to chemotherapeutic agents. Several reports have suggested increased response to anthracyclines (42, 43) and resistance to cytotoxic alkalating agents (44) in breast cancer patients with p185erbB2 overexpressing tumors. We are well aware of the fact that the number of patients in our series is rather small, and that the variety of stage, grade, primary surgical treatment, and adjuvant systemic treatment may confound the interpretation of our data. However, the significant prognostic impact of p185erbB2-positive micrometastatic cells was confirmed as being independent of these potentially confounding variables by multivariate analysis.

In view of the data presented here, an anti-p185erbB2 anticancer treatment is desirable. One promising approach is therapy with monoclonal antibodies directed against the extracellular domain of p185erbB2 that have recently been introduced into the treatment of metastatic breast cancer (20, 21, 22, 23). Isolated micrometastatic cells may even represent better targets for antibody therapy than overt metastases (45), especially in an environment with abundant host-immune effector cells (such as BM) that enhance the antibody-mediated chemosensitivity by additional antibody/complement-dependent cytotoxicity. Moreover, the relative resistance of micrometastatic tumor cells in BM to current chemotherapeutic agents (46) might be successfully overcome by therapies that are independent of the proliferative status of the targeted cells. In this context, the BM assay described here could be used to stratify patients for adjuvant anti-p185erbB2 antibody therapy and to monitor eradication of micrometastatic cells during therapy. Based on the results of our small pilot study, this proposed treatment stratification and monitoring strategy needs to be confirmed by randomized prospective clinical trials. The first trial of this kind has recently been initiated at four German Breast Cancer Centers in Hamburg, Hannover, Munich, and Tuebingen.

Fig. 1.

Immunocytochemical double labeling of disseminated tumor cells in BM reveals coexpression of cytoplasmatic CK18 and extracellular p185erbB2 in single tumor cells. A, bright field illumination shows black extracellular staining of strong p185erbB2 expression, and (B) epipolarized light illumination (inset) visualizes extracellular staining of weaker p185erbB2 expression on CK18-positive (red cytoplasmatic stain) tumor cells. Surrounding BM cells are negative for both immunostainings (no counterstaining was performed; original magnification, ×1000).

Fig. 1.

Immunocytochemical double labeling of disseminated tumor cells in BM reveals coexpression of cytoplasmatic CK18 and extracellular p185erbB2 in single tumor cells. A, bright field illumination shows black extracellular staining of strong p185erbB2 expression, and (B) epipolarized light illumination (inset) visualizes extracellular staining of weaker p185erbB2 expression on CK18-positive (red cytoplasmatic stain) tumor cells. Surrounding BM cells are negative for both immunostainings (no counterstaining was performed; original magnification, ×1000).

Close modal
Fig. 2.

Box plot diagram showing the number of CK18-positive micrometastases according to coexpression of p185erbB2. Each box shows quantitative CK18-positive findings in patients with (n = 31) and without (n = 21) expression of p185erbB2 on BM micrometastases. Differences of mean (P = 0.0003, Mann-Whitney U test) and median (P = 0.0043, Wilcoxon test for independent variables) cell counts were statistically significant.

Fig. 2.

Box plot diagram showing the number of CK18-positive micrometastases according to coexpression of p185erbB2. Each box shows quantitative CK18-positive findings in patients with (n = 31) and without (n = 21) expression of p185erbB2 on BM micrometastases. Differences of mean (P = 0.0003, Mann-Whitney U test) and median (P = 0.0043, Wilcoxon test for independent variables) cell counts were statistically significant.

Close modal
Fig. 3.

Kaplan-Meier analysis for overall survival of 52 patients according to the expression of p185erbB2 on BM micrometastases. Top curve, patients with p185erbB2-negative micrometastases; 21 patients, 7 events (i.e., cancer-related deaths). Bottom curve, patients with p185erbB2-positive micrometastases; 31 patients, 21 events (i.e., cancer-related deaths). P = 0.032 by the log-rank test; values in parentheses represent probability of survival at the end of follow-up.

Fig. 3.

Kaplan-Meier analysis for overall survival of 52 patients according to the expression of p185erbB2 on BM micrometastases. Top curve, patients with p185erbB2-negative micrometastases; 21 patients, 7 events (i.e., cancer-related deaths). Bottom curve, patients with p185erbB2-positive micrometastases; 31 patients, 21 events (i.e., cancer-related deaths). P = 0.032 by the log-rank test; values in parentheses represent probability of survival at the end of follow-up.

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

Supported by the Deutsche Krebshilfe and the Dr. Mildred Scheel Stiftung (Bonn, Germany). Part of the material contained in this study was analyzed by I. H. as part of her M.D. thesis at Ludwig-Maximilians University, Medical Faculty (Munich, Germany).

3

The abbreviations used are: BM, bone marrow; CK, cytokeratin; CMF, 600 mg/m2 cyclophosphamide, 40 mg/m2 methotrexate, and 600 mg/m2 5-fluorouracil; EC, 90 mg/m2 epirubicin and 600 mg/m2 cyclophosphamide.

Table 1

Clinical characteristics of 52 patients with BM micrometastases analyzed for p185erbB2 overexpression

Expression of p185erbB2
Clinical characteristicsPatients (n = 52)Negative (n = 21)Positive (n = 31)P              a
Mean age (yr) 56 59 54 0.31 
Tumor size    0.50 
 ≤5 cm (pT1 and pT242 16 26  
 >5 cm (pT3 and pT410  
Grading    0.52 
 G1 and G2 39 17 22  
 G3 13  
Lymph node metastasis    0.79 
 Absent (pN026 10 16  
 Present (pN1 and pN226 11 15  
Systemic treatment    0.99 
 None 21 13  
 Endocrine/cytotoxicb 31 13 18  
Expression of p185erbB2
Clinical characteristicsPatients (n = 52)Negative (n = 21)Positive (n = 31)P              a
Mean age (yr) 56 59 54 0.31 
Tumor size    0.50 
 ≤5 cm (pT1 and pT242 16 26  
 >5 cm (pT3 and pT410  
Grading    0.52 
 G1 and G2 39 17 22  
 G3 13  
Lymph node metastasis    0.79 
 Absent (pN026 10 16  
 Present (pN1 and pN226 11 15  
Systemic treatment    0.99 
 None 21 13  
 Endocrine/cytotoxicb 31 13 18  
a

Ps were calculated by Fisher’s exact test for comparison of numbers of patients and by the Mann-Whitney U test for comparison of means.

b

Patients received 20–30 mg tamoxifen daily for 2–5 years (n = 12) or CMF (n = 7) and EC/CMF chemotherapy (n == 12), respectively.

Table 2

Expression of p185erbB2 on primary and micrometastatic tumor cells

Primary tumoraBMb
Patient no.Tumor stageIntensity of p185erbB2 staining% p185erbB2-positive cells% p185erbB2-positive cells
T1G2N1M0 51–80 
T3G2N1M0 11–50 
T1G1N1M0 51–80 
T4G2N1M0 >80 50 
T2G2N1M0 51–80 50 
T1G1N0M0 51–80 50 
T2G2N1M0 51–80 81 
T1G1N0M0 ++ 51–80 
T2G2N1M0 ++ 51–80 
10 T2G3N0M0 ++ 51–80 
11 T2G2N0M0 ++ 11–50 38 
12 T1G2N0M0 ++ 51–80 60 
13 T1G1N0M0 ++ 51–80 71 
14 T1G3N1M0 ++ 51–80 91 
15 T1G3N2M0 ++ 51–80 100 
16 T1G2N1M0 ++ >80 100 
17 T2G2N0M0 ++ 11–50 100 
18 T2G2N0M0 +++ >80 
19 T1G3N0M0 +++ >80 
20 T2G2N1M0 +++ 51–80 
21 T1G2N0M0 +++ >80 50 
22 T3G3N2M0 +++ 11–50 64 
23 T2G2N0M0 +++ >80 83 
24 T2G3N1M0 +++ 51–80 100 
Primary tumoraBMb
Patient no.Tumor stageIntensity of p185erbB2 staining% p185erbB2-positive cells% p185erbB2-positive cells
T1G2N1M0 51–80 
T3G2N1M0 11–50 
T1G1N1M0 51–80 
T4G2N1M0 >80 50 
T2G2N1M0 51–80 50 
T1G1N0M0 51–80 50 
T2G2N1M0 51–80 81 
T1G1N0M0 ++ 51–80 
T2G2N1M0 ++ 51–80 
10 T2G3N0M0 ++ 51–80 
11 T2G2N0M0 ++ 11–50 38 
12 T1G2N0M0 ++ 51–80 60 
13 T1G1N0M0 ++ 51–80 71 
14 T1G3N1M0 ++ 51–80 91 
15 T1G3N2M0 ++ 51–80 100 
16 T1G2N1M0 ++ >80 100 
17 T2G2N0M0 ++ 11–50 100 
18 T2G2N0M0 +++ >80 
19 T1G3N0M0 +++ >80 
20 T2G2N1M0 +++ 51–80 
21 T1G2N0M0 +++ >80 50 
22 T3G3N2M0 +++ 11–50 64 
23 T2G2N0M0 +++ >80 83 
24 T2G3N1M0 +++ 51–80 100 
a

Staining intensity: weak, +; intermediate, ++; strong, +++. Number of stained cells: H10%; 11–50%; 51–80%; >80%.

b

A mean number of 44 (range, 1–848) CK18-positive tumor cells per 4 × 105 BM cells was evaluated per patient.

Table 3

Expression of p185erbB2 on micrometastatic tumor cells in BM and patient outcome in stage I–III breast cancer

Expression of p185erbB2
Outcome of patientsPatients (n = 52)Negative (n = 21)Positive (n = 31)P              a
Deadb 27 20 0.047 
Alive 25 14 11  
 Recurrence  
 No evidence of disease 23 14  
Expression of p185erbB2
Outcome of patientsPatients (n = 52)Negative (n = 21)Positive (n = 31)P              a
Deadb 27 20 0.047 
Alive 25 14 11  
 Recurrence  
 No evidence of disease 23 14  
a

Ps were calculated by Fisher’s exact test for comparison of patient numbers.

b

Only cancer-related deaths were incorporated in this analysis.

Table 4

Multivariate analysis of variables influencing overall survival of 52 patients with stage I–III breast cancera

VariablesPRR (95% CI)
Age 0.52 b 
 <55 yrs (19/33); ≥55 yrs (8/19)   
Tumor size 0.81  
 H5 cm (22/42); >5 cm (5/10)   
Lymph node metastasis 0.039 2.48 (1.05–5.86) 
 Absent (11/26); present (16/26)   
Expression of p185erbB2 on micrometastases 0.029 2.78 (1.11–6.96) 
 Negative (7/21); positive (20/31)   
Grading 0.08  
 G1 and G2 (17/39); G3 (10/13)   
VariablesPRR (95% CI)
Age 0.52 b 
 <55 yrs (19/33); ≥55 yrs (8/19)   
Tumor size 0.81  
 H5 cm (22/42); >5 cm (5/10)   
Lymph node metastasis 0.039 2.48 (1.05–5.86) 
 Absent (11/26); present (16/26)   
Expression of p185erbB2 on micrometastases 0.029 2.78 (1.11–6.96) 
 Negative (7/21); positive (20/31)   
Grading 0.08  
 G1 and G2 (17/39); G3 (10/13)   
a

Cox proportional-hazard model for stepwise multivariate analysis. RR, relative risk; CI, confidence interval. The number of patients (events/total) is given in parentheses.

b

No relative risk was available because the variable was not significant with respect to the multivariate model; data were entered into the model stepwise after stratification for adjuvant systemic therapy (tamoxifen or CMF and EC/CMF chemotherapy).

We gratefully acknowledge excellent technical assistance of Tanja Siart and Simone Baier. We thank our colleagues at the Departments of Gynecology and Obstetrics in Munich (head: Prof. Günter Kindermann) and Augsburg (head: Prof. Artur Wischnik) and at the Departments of General Surgery in Augsburg (head: Prof. Jens Witte) for help in patient recruitment and follow-up. We also thank Dr. Nadia Harbeck for assistance in preparation of the manuscript. We gratefully acknowledge the provision of biotinylated monoclonal antibody CK2 by Dr. H. Bodenmüller (Boehringer Mannheim, Tutzing, Germany).

1
Schechter A. L., Stern D. F., Vaidyanathan L., Decker S. J., Drebin J. A., Greene M. I., Weinberg R. A. The neu oncogene: an erb-B-related gene encoding a 185,000-Mr tumour antigen.
Nature (Lond.)
,
312
:
513
-516,  
1984
.
2
Coussens L., Yang Feng T., Liao Y. C., Chen E., Gray A., McGrath J., Seeburg P. H., Libermann T. A., Schlessinger J., Francke U., et al Tyrosine kinase receptor with extensive homology to EGF receptor shares chromosomal location with neu oncogene.
Science (Washington DC)
,
230
:
1132
-1139,  
1985
.
3
Peles E., Bacus S. S., Koski R. A., Lu H. S., Wen D., Ogden S. G., Levy R. B., Yarden Y. Isolation of the neu/HER-2 stimulatory ligand: a 44 kD glycoprotein that induces differentiation of mammary tumor cells.
Cell
,
69
:
205
-216,  
1992
.
4
Holmes W. E., Sliwkowski M. X., Akita R. W., Henzel W. J., Lee J., Park J. W., Yansura D., Abadi N., Raab H., Lewis G. D., Sephard H. M., Kuang W-J., Wood W. I., Goeddel D. V., Vandlen R. L. Identification of heregulin, a specific activator of p185erbB2.
Science (Washington DC)
,
256
:
1205
-1210,  
1992
.
5
Lupu R., Colomer R., Kannan B., Lippman M. E. Characterization of a growth factor that binds exclusively to the erbB2 receptor and induces cellular responses.
Proc. Natl. Acad. Sci. USA
,
89
:
2287
-2291,  
1992
.
6
Pauletti G., Dandekar S., Rong H. M., Ramos L., Peng H. J., Seshadri R., Slamon D. J. Assessment of methods for tissue-based detection of the HER-2/neu alteration in human breast cancer: a direct comparison of fluorescence in situ hybridization and immunohistochemistry.
J. Clin. Oncol.
,
18
:
3651
-3664,  
2000
.
7
Slamon D. J., Clark G. M., Wong S. G., Levin W. J., Ullrich A., McGuire W. L. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene.
Science (Washington DC)
,
235
:
177
-182,  
1987
.
8
Slamon J. D., Godolphin W., Jones L. A., Holt J. A., Wong S. G., Keith D. E., Levin W. J., Stuart S. G., Udove J., Ullrich A., Press M. F. Studies of the HER-2/neu proto-oncogene in human breast cancer and ovarian cancer.
Science (Washington DC)
,
244
:
707
-712,  
1989
.
9
Toikkanen S., Helin H., Isola J., Joensuu H. Prognostic significance of HER-2 oncoprotein expression in breast cancer: a 30-year follow-up.
J. Clin. Oncol.
,
10
:
1044
-1048,  
1992
.
10
Gusterson B. A., Gelber R. D., Goldhirsch A., Price K. N., Säve-Söderborgh J., Anbazhagan R., Styles J., Rudenstam C. M., Golouh R., Reed R., Martinez-Tello F., Tiltman A., Torhorst J., Grigolato P., Bettelheim R., Neville A. M., Bürki K., Castiglione M., Collins J., Lindtner J., Senn H. J. Prognostic importance of c-erbB-2 expression in breast cancer.
J. Clin. Oncol.
,
10
:
1049
-1056,  
1992
.
11
Van de Vijver M., Peterse J. L., Mooi W. J., Wisman P., Lomans J., Dalesio O., Nusse R. Neu-protein over-expression in breast cancer. Association with comedo-type ductal carcinoma in situ and limited prognostic value in stage II breast cancer.
N. Engl. J. Med.
,
319
:
1239
-1245,  
1988
.
12
Barnes D. M., Lammie G. A., Millis R. R., Gullick W. L., Allen D. S., Altman D. G. An immunohistochemical evaluation of c-erbB-2 expression in human breast carcinoma.
Br. J. Cancer
,
58
:
448
-452,  
1988
.
13
Gasparini G., Gullick W. J., Bevilacqua P., Sainsbury J. R. C., Meli S., Boracchi P., Testolin A., La Malfa G., Pozza F. Human breast cancer: prognostic significance of the c-erbB-2 oncoprotein compared with epidermal growth factor receptor, DNA ploidy, and conventional pathologic features.
J. Clin. Oncol.
,
10
:
686
-695,  
1992
.
14
Braun S., Pantel K., Müller P., Janni W., Hepp F., Kentenich C. R. M., Gastroph S., Wischnik A., Dimpfl T., Kindermann G., Riethmüller G., Schlimok G. Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II or III breast cancer.
N. Engl. J. Med.
,
342
:
525
-533,  
2000
.
15
Braun S., Müller M., Hepp F., Schlimok G., Riethmüller G., Pantel K. Re: micrometastatic breast cancer cells in bone marrow at primary surgery: prognostic value in comparison with nodal status.
J. Natl. Cancer Inst. (Bethesda)
,
90
:
1099
-1100,  
1998
.
16
Pantel K., Schlimok G., Angstwurm M., Weckermann C., Schmaus W., Gath H., Passlick B., Izbicki J., Riethmüller G. Methodological analysis of immunocytochemical screening for disseminated epithelial tumor cells in bone marrow.
J. Hematother.
,
3
:
165
-173,  
1994
.
17
Mansi J. L., Gogas H., Bliss J. M., Gazet J. C., Berger U., Coombes R. C. Outcome of primary-breast-cancer patients with micrometastases: a long-term follow-up.
Lancet
,
354
:
197
-202,  
1999
.
18
Landys K., Persson S., Kovarik J., Hultborn R., Holmberg E. Prognostic value of bone marrow biopsy in operable breast cancer patients at the time of initial diagnosis: results of a 20-year median follow-up.
Breast Cancer Res. Treat.
,
49
:
27
-33,  
1998
.
19
Cote R. J., Rosen P. P., Lesser M. L., Old L. J., Osborne M. P. Prediction of early relapse in patients with operable breast cancer by detection of occult bone marrow micrometastases.
J. Clin. Oncol.
,
9
:
1749
-1756,  
1991
.
20
Cobleigh M. A., Vogel C. L., Tripathy D., Robert N. J., Scholl S., Fehrenbacher L., Walter J. M., Paton V., Shak S., Lieberman G., Slamon D. J. Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2-over-expressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease.
J. Clin. Oncol.
,
17
:
2639
-2648,  
1999
.
21
Pegram M. D., Lipton A., Hayes D. F., Weber B. L., Baselga J. M., Tripathy D., Baly D., Baughman S. A., Twaddell T., Glaspy J. A., Slamon D. J. Phase II study of receptor-enhanced chemosensitivity using recombinant humanized anti-p185HER2/neu monoclonal antibody plus cisplatin in patients with HER2/neu-overexpressing metastatic breast cancer refractory to chemotherapy treatment.
J. Clin. Oncol.
,
16
:
2659
-2671,  
1998
.
22
Slamon J. D., Leyland-Jones B., Shak S., Paton V., Bajamonde A., Fleming T., Eiermann W., Wolter J., Baselga J., Norton L. Addition of Herceptin™ (humanized anti-HER2 antibody) to first-line chemotherapy for HER2 over-expressing metastatic breast cancer markedly increases anti-cancer activity: a randomized, multinational controlled Phase III trial.
Proc. Am. Soc. Clin. Oncol.
,
17
:
98
1998
.
23
Baselga J., Tripathy D., Mendelsohn J., Baughman S., Benz C. C., Dantis L., Sklarin N. T., Seidman A. D., Hudis C. A., Moore J., Rosen P. P., Twaddell T., Henderson I. C., Norton N. Phase II study of weekly intravenous recombinant humanized anti-p185Her2 monoclonal antibody in patients with Her2/neu-over- expressing metastatic breast cancer.
J. Clin. Oncol.
,
14
:
737
-744,  
1996
.
24
Debus E., Moll R., Franke W. W. Immunohistochemical distinction of human carcinomas by cytokeratin typing with monoclonal antibodies.
Am. J. Pathol.
,
114
:
121
-130,  
1984
.
25
Cordell J. L., Falini B., Erber W. N., Ghosh A. K., Abdulaziz Z., MacDonald S., Pulford K. A. F., Stein H., Mason D. Y. Immunoenzymatic labeling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP-complexes).
J. Histochem. Cytochem.
,
32
:
219
-229,  
1984
.
26
Pantel K., Schlimok G., Braun S., Kutter D., Schaller G., Funke I., Izbicki J., Riethmüller G. Differential expression of proliferation-associated molecules in individual micrometastatic carcinoma cells.
J. Natl. Cancer Inst. (Bethesda)
,
85
:
1419
-1424,  
1993
.
27
Braun S., Hepp F., Sommer H. L., Pantel K. Tumor antigen heterogeneity of disseminated breast cancer cells: implications for immunotherapy of minimal residual disease.
Int. J. Cancer Predict. Oncol.
,
84
:
1
-5,  
1999
.
28
Press M. F., Hung G., Godolphin W., Slamon D. J. Sensitivity of HER-2/neu antibodies in archival tissue samples: potential source of error in immunohistochemical studies of oncogene expression.
Cancer Res.
,
54
:
2771
-2777,  
1994
.
29
Kaplan E. L., Meier P. Non-parametric estimation from incomplete observations.
J. Am. Stat. Assoc.
,
53
:
457
-481,  
1958
.
30
Cox D. R. Regression models and life-tables.
J. R. Stat. Soc.
,
34
:
187
-220,  
1972
.
31
Untch M., Kahlert S., Funke I., Boettcher B., Konecny G., Nestle-Kraemling C., Bauerfeind I. Detection of cytokeratin 18-positive cells in bone marrow of breast cancer patients: no prediction of bad outcome.
Proc. Am. Soc. Clin. Oncol.
,
18
:
639a
1999
.
32
Zippelius A., Kufer P., Honold G., Köllermann M. W., Oberneder R., Schlimok G., Riethmüller G., Pantel K. Limitations of reverse transcriptase-polymerase chain reaction for detection of micrometastatic epithelial cancer cells in bone marrow.
J. Clin. Oncol.
,
15
:
2701
-2708,  
1997
.
33
Putz E., Witter K., Offner S., Stosiek P., Zippelius A., Johnson J. P., Zahn R., Riethmüller G., Pantel K. Phenotypic characteristics of cell lines derived from disseminated cancer cells in bone marrow of patients with solid epithelial tumors: establishment of working models for human micrometastases.
Cancer Res.
,
59
:
241
-248,  
1999
.
34
Park J. M., Petitclerc E., Barrón M., Park J. J., Brooks P. C., Press M. F. HER-2/neu (c-erbB2) receptor tyrosine kinase overexpression alters integrin- mediated cell-adhesion.
Proc. AACR.
,
40
:
564
1999
.
35
Ye J., Xu R. H., Taylor-Papadimitriou J., Pitha P. M. Sp1 binding plays a critical role in erbB2- and v-ras-mediated downregulation of alpha2-integrin expression in human mammary epithelial cells.
Mol. Cell Biol.
,
16
:
6178
-6189,  
1996
.
36
D’Souza B., Taylor-Papadimitriou J. Overexpression of erbB2 in human mammary epithelial cells signals inhibition of transcription of the E-cadherin gene.
Proc. Natl. Acad. Sci. USA
,
91
:
7202
-7206,  
1994
.
37
Heiss M. M., Allgayer H., Gruetzner K. U., Funke I., Babic R., Jauch K. W., Schildberg F. W. Individual development and uPA-receptor expression of disseminated tumour cells in bone marrow: a reference to early systemic disease in solid cancer.
Nat. Med.
,
1
:
1035
-1039,  
1995
.
38
Offner S., Schmaus W., Witter K., Baretton G. B., Schlimok G., Passlick B., Riethmüller G., Pantel K. p53 mutations are not required for early dissemination of cancer cells.
Proc. Natl. Acad. Sci. USA
,
96
:
6942
-6946,  
1999
.
39
Pantel K., Schlimok G., Kutter D., Schaller G., Genz T., Wiebecke B., Backmann R., Funke I., Riethmüller G. Frequent down-regulation of major histocompatibility class I antigen expression on individual micrometastatic carcinoma cells.
Cancer Res.
,
51
:
4712
-4715,  
1991
.
40
Müller P., Weckermann D., Riethmüller G., Schlimok G. Detection of genetic alterations in micrometastatic cells in bone marrow of cancer patients by fluorescence in situ hybridization.
Cancer Genet. Cytogenet.
,
88
:
8
-16,  
1996
.
41
Brandt B., Roetger A., Heidl S., Jackisch C., Lelle R. J., Assmann G., Zanker K. S. Isolation of blood-borne epithelium-derived c-erb-B2 oncoprotein-positive clustered cells from peripheral blood of breast cancer patients.
Int. J. Cancer
,
76
:
824
-828,  
1998
.
42
Paik S., Bryant J., Park C., Fisher B., Tan-Chiu E., Hyams D., Fisher R. B., Sass R. E., Lippman M. E., Wickerham D. L., Wolmark N. Erb-B2 and response to doxorubicin in patients with axillary lymph node-positive, hormone receptor-negative breast cancer.
J. Natl. Cancer Inst. (Bethesda)
,
90
:
1361
-1370,  
1998
.
43
Muss H. B., Thor A. D., Berry D. A., Kute T., Liu E. T., Koerner F., Cirrincione C. T., Budman D. R., Wood W. C., Barcos M., Henderson I. C. c-erbB-2 expression and response to adjuvant therapy in women with node-positive early breast cancer.
N. Engl. J. Med.
,
330
:
1260
-1266,  
1994
.
44
Hancock M. C., Langton B. C., Chan T., Toy P., Monaham J. J., Mischak R. P., Shawver L. K. A monoclonal antibody against c-erbB-2 protein enhances the cytotoxicity of cis-diamine dichloro-platinum against human breast and ovarian cancer cell lines.
Cancer Res.
,
51
:
4575
-4580,  
1991
.
45
Jain R. K. Physiological barriers to delivery of monoclonal antibodies and other macromolecules in tumors.
Cancer Res.
,
50
:
2741
-2451,  
1990
.
46
Braun S., Kentenich C. R. M., Janni W., Hepp F., de Waal J., Willgeroth F., Sommer H. L. Lack of effect of adjuvant chemotherapy on the elimination of single dormant tumor cells in bone marrow of high-risk breast cancer patients.
J. Clin. Oncol.
,
18
:
80
-86,  
2000
.