|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Detection of Isolated Tumor Cells in Bone Marrow Is an Independent Prognostic Factor in Breast Cancer
From the Departments of Surgery and Pathology, Ullevål University Clinic; Departments of Pathology, Oncology, and Surgery, The Norwegian Radium Hospital; Department of Surgery, Baerum Hospital; and Department of Surgery, Aker University Hospital, Oslo, Norway. Address reprint requests to Gro Wiedswang, MD, Surgical Department, Ullevål University Hospital, Kirkeveien 165, N-0407 Oslo, Norway; email: gro.wiedswang{at}ulleval.no.
Purpose: This study was performed to disclose the clinical impact of isolated tumor cell (ITC) detection in bone marrow (BM) in breast cancer. Patients and Methods: BM aspirates were collected from 817 patients at primary surgery. Tumor cells in BM were detected by immunocytochemistry using anticytokeratin antibodies (AE1/AE3). Analyses of the primary tumor included histologic grading, vascular invasion, and immunohistochemical detection of c-erbB-2, cathepsin D, p53, and estrogen receptor (ER)/progesterone receptor (PgR) expression. These analyses were compared with clinical outcome. The median follow-up was 49 months. Results: ITC were detected in 13.2% of the patients. The detection rate rose with increasing tumor size (P = .011) and lymph node involvement (P < .001). Systemic relapse and death from breast cancer occurred in 31.7% and 26.9% of the BM-positive patients versus 13.7% and 10.9% of BM-negative patients, respectively (P < .001). Analyzing node-positive and node-negative patients separately, ITC positivity was associated with poor prognosis in the node-positive group and in node-negative patients not receiving adjuvant therapy (T1N0). In multivariate analysis, ITC in BM was an independent prognostic factor together with node, tumor, and ER/PgR status, histologic grade, and vascular invasion. In separate analysis of the T1N0 patients, histologic grade was independently associated with both distant disease-free survival (DDFS) and breast cancer–specific survival (BCSS), ITC detection was associated with BCSS, and vascular invasion was associated with DDFS. Conclusion: ITC in BM is an independent predictor of DDFS and BCSS. An unfavorable prognosis was observed for node-positive patients and for node-negative patients not receiving systemic therapy. A combination of several independent prognostic factors can classify subgroups of patients into excellent and high-risk prognosis groups.
A LARGE proportion of breast cancer patients experience systemic relapse, approximately 25% to 30% of node-negative patients and 60% of node-positive patients.1,2 At present, tumor, node, metastasis (TNM) classification, histopathologic tumor grade, hormone receptor status, and age are used for prognostication and treatment decisions. However, these factors are not sufficient for an accurate selection of patients into low-risk groups, with no need for adjuvant therapy, and high-risk groups for systemic relapse. Extensive work has been performed to develop methods for the assessment of risk. The presence of isolated tumor cells (ITC) in the bone marrow (BM) at the time of operation has been shown to be a predictor of early systemic relapse in node-positive patients in several independent studies.3–8 However, the reported results from ITC analysis of node-negative patients have been contradictory.4,5,8–10 Most studies have used immunocytochemical (ICC) techniques for the detection of ITC. Methodologic differences between studies exist9,11 that may cause differences in both sensitivity and specificity. Therefore, guidelines for analysis and evaluation of ITC by ICC in BM, proposed by The European Working Group for Standardization of Tumor Cell Detection, have been used in this study.12 The aim of this study was to examine early-stage breast cancer patients for the presence of ITC in BM, clarify their prognostic value, and analyze and relate candidate prognostic factors in the primary tumors to the presence of ITC and clinical outcome in 817 unselected patients.
Patients From May 1995 to December 1998, 920 breast cancer patients were included from the following five hospitals in Oslo, Norway: Ullevål University Hospital, Norwegian Radium Hospital, Baerum Hospital, Aker University Hospital, and Buskerud Hospital. After informed, written consent, BM aspiration was performed under general anesthesia, just before primary surgery for suspected breast cancer. The routine diagnostic work-up included mammography, chest x-ray, blood sampling, and clinical examination. The treatment was breast-conserving surgery or breast ablation and axillary clearance. The routine selection of patients to adjuvant treatment was as described in Table 1  . The results of the BM examination were not known to the clinicians responsible for the treatment. The standard adjuvant chemotherapy regimen consisted of six cycles of intravenous cyclophosphamide 600 mg/m2, methotrexate 40 mg/m2, and fluorouracil 600 mg/m2. In 17 patients, preoperative chemotherapy was administered, and 13 patients received high-dose chemotherapy. Tamoxifen was used as endocrine therapy in receptor-positive patients. Radiotherapy was offered after breast-conserving surgery for all node-positive premenopausal patients and postmenopausal patients with four or more positive lymph nodes. The follow-up consisted of clinical examination at 6- to 12- month intervals at the hospital outpatient departments or by the patients’ primary doctors, with mammography annually. Further diagnostic work-up was performed only if the patients had symptoms or signs of progression.
Preparation of the BM A total of 40 mL of BM was aspirated from anterior and posterior iliac crests bilaterally, 10 mL per site, and processed as described previously.13 After separation by density centrifugation, mononuclear cells (MNC) were collected, and cytospins were prepared (5 x 105 MNC/slide).
ICC Staining
Detection of ITC
Negative Immunomagnetic Separation (IMS) Technique
Analysis of Primary Tumor and Axillary Lymph Nodes
Statistical Analysis
Of the 920 patients initially enrolled onto the study, histopathologic tumor analysis revealed seven patients with benign lesions, 38 in situ carcinomas, and two nonepithelial cancers. Eight patients had distant metastasis at diagnosis. For two of the patients, the cytospin quality was inappropriate for conclusions about the presence of ITC, and 46 patient samples were noninterpretable because of a positive result in both the specific test and the negative control. Excluding these patients, 817 assessable patients remained for further exploration.
Patient Characteristics and Detection of ITC
Detection of ITC and Disease Recurrence The median follow-up time was 49 months (range, 0.5 to 85 months). Three of the patients had less than 3 months of follow-up; one patient died because of a perioperative complication, and two patients emigrated. One hundred seventy-five patients (21.4%) experienced recurrence of the disease. Locoregional relapse alone was diagnosed in 48 patients (5.9%), and systemic relapse was diagnosed in 127 patients (15.5%). Locoregional relapse alone was not associated with the presence of ITC in the BM. Systemic relapse was significantly more frequent in the BM-positive group (30.6%) than in the BM-negative group (13.3%, P < .001, Table 3 ). Among the node-negative patients, systemic relapse occurred with the same frequency whether or not ITC was detected. However, a markedly increased frequency of systemic relapse was recorded in node-positive patients compared with node-negative patients (51.8% v 25.5%, respectively; P = .002). BM positivity was highly associated with skeletal metastasis (Table 3). In the node-positive group, skeletal involvement was detected in 42.9% of the BM-positive patients compared with 14.1% of the BM-negative patients (P < .001). Analyzing only the patients experiencing systemic relapse, the skeleton was involved in 81.8% of the BM-positive patients compared with 57.4% of the BM-negative patients (data not shown). The rate of metastasis to the liver was also significantly higher in BM-positive patients. The presence of pulmonary metastases did not correlate to BM status.
Negative IMS In an attempt to increase the sensitivity of ITC detection, a negative IMS step was accomplished in addition to direct ICC in 582 patients. Negative IMS was successful in 464 patients (80%). In these patients, negative IMS or the combination of negative IMS and direct ICC did not improve the prognostic value of ITC detection in BM (Table 4 ). Accordingly, the results from direct ICC have been used in the following analysis.
Survival Analysis During the observation period, 88 (10.8%) of the 817 patients died of metastatic disease. In the BM-positive group, 25.9% (28 of 108 patients) died of breast cancer, compared with 8.5% (60 of 709 patients) of the BM-negative patients (P < .001). Kaplan-Meier survival analyses demonstrate a markedly reduced distant disease-free survival (DDFS) and BCSS among the BM-positive patients (Fig 1 ). Separate analyses of the node-negative and node-positive patients revealed a marked difference in DDFS and BCSS according to BM status in node-positive but not in node-negative patients (Fig 2A to 2D). The node-negative patients with the highest risk of relapse (T2, grade 2 to 3) received systemic adjuvant treatment. The majority of the node-negative patients were not offered such treatment and belonged to the T1N0 subgroup (n = 321). In this group, neither hormone receptor status nor histologic grade initiated the use of systemic therapy. These nontreated node-negative patients were chosen for further subgroup analysis. The T1N0 patients with ITC in BM experienced reduced BCSS (P = .014) but no difference in DDFS (Fig 2E and 2F) compared with ITC-negative patients.
Univariate and Multivariate Analysis BM status, nodal status, tumor status, histologic grade, hormone receptor status, vascular invasion, and p53, c-erbB-2, and cathepsin D expression were all tested in univariate analysis. Except for cathepsin D, all variables were significantly associated with DDFS and BCSS (P < .001). Subsequently, multivariate analyses (Cox regression) of the subgroup with a complete data set of prognostic factors were performed (n = 593). The data were stratified according to use of adjuvant therapy. BM status, node status, tumor status, histologic grade, and ER/PgR expression were all found to be independent prognostic factors for both DDFS and BCSS, and vascular invasion was significant only for DDFS (Table 5 ). Multivariate analysis was then performed separately for T1N0 and node-positive patients (Table 6). In the T1N0 group, histologic grade was the only independent prognostic factor for both DDFS and BCSS, whereas ITC detection was independently associated with BCSS only, and vascular invasion was associated with DDFS. Among the node-positive patients, tumor status, ER/PgR expression, and BM status were independent factors for both survival analyses, whereas histologic grade was significant only for BCSS.
Combination of Independent Prognostic Factors in Survival Analyses To disclose patients with especially low risk of metastasis with questionable need of systemic adjuvant therapy and patients with high risk of adjuvant systemic treatment failure, the value of combining the independent prognostic factors in risk assessment was explored. In the T1N0 group, the independent prognostic factors were stepwise added for exploration of the most optimal combination of factors for prediction of DDFS and BCSS. This analysis revealed that only histologic grade was needed for optimal categorization of patients into low- and high-risk groups. The addition of ITC detection and analysis of vascular invasion did not result in further prognostic information (data not shown). The Kaplan-Meier survival plot for histologic grade 1 and 2 versus 3 in the T1N0 group is shown in Figure 3A and 3B. Of the patients, 83.2% had grade 1 or 2 tumors. Among these patients, only 3.7% (10 of 267 patients) experienced systemic relapse, and 1.1% (three of 267 patients) died of breast cancer, compared with 18.5% (10 of 54 patients) and 9.3% (five of 54 patients), respectively, of patients with histologic grade 3 tumors. Similar stepwise analysis of factors was performed for the node-positive patients. As shown in Fig 3C and 3D, hormone receptor–positive patients with T1 tumors and BM-negative status (30.1% of the node positive group) had an excellent prognosis. Only 3.0% (two of 66 patients) of patients experienced systemic relapse, and 1.5% (one of 66 patients) died of breast cancer. In node-positive patients with T2 tumors and/or hormone receptor–negative status and/or BM-positive status, 37.3% (57 of 153 patients) experienced systemic relapse, and 27.5% (42 of 153 patients) died of breast cancer.
This study is, to date, one of the largest prospective analyses of clinical relevance of detection of ITC in breast cancer patients. The study population is unselected because the majority of all patients referred to the inclusion hospitals were asked to participate. Overall, the patient population represents earlier breast cancer stage than in previous studies,3–8 with a median tumor size of 1.7 cm and approximately two thirds of the patients being node-negative. Approximately one third of the patients were participants in an official mammography screening program.13 This study confirms that the occurrence of ITC in BM predicts future systemic relapse and death from breast cancer, and the detection of these cells is an independent prognostic factor for systemic relapse and breast cancer–specific death. Among the node-negative patients, detection of ITC in BM is a prognostic factor for patients not receiving adjuvant therapy. However, the analysis of all independent prognostic factors reveals that analysis of histologic grade alone can identify a large group of node-negative patients (approximately 80%) with an especially good prognosis that may not need adjuvant systemic therapy. The combination of tumor, hormone receptor, and BM status can categorize node-positive patients into a good prognosis group and a high-risk group for early disease progression.
To date, the reported clinical significance of ITC detection in BM in node-negative patients has differed.4,5,8 In two previous studies, BM analysis was found to identify node-negative patients with an excellent/good prognosis.4,5 Using a different technique for cell preparation and ICC detection, Gebauer et al8 did not find this association. Therefore, it can be argued that methodologic factors as well as the size of the study population determine the impact of BM analysis when the micrometastatic load is low. Indeed, if the proportion of BMs harboring tumor cells is low, the BM results will be influenced by a higher relative frequency of unspecific reactions (false-positive cells). Supporting this is our previous estimation of a 3% to 4% rate of false-positive reactions not detected by negative control analysis.13 We observed a significant reduction in BCSS, but not in DDFS, in the node-negative BM-positive patients not treated with adjuvant therapy. Because the frequency of BM-positive patients in this group is low, the false-positive background probably reduces the true difference in survival between the BM-positive and BM-negative group. With a median follow-up of 49 months, our observation time may be too short for final conclusions about the prognostic relevance of the BM analysis for late recurrences. However, a reanalysis of the original study by Mansi et al10 with long-term follow-up revealed that the significance of BM analysis is reduced over time. As compared with our study, Braun et al5 observed prognostic impact of ITC detection in node negative patients with shorter follow-up time. This argues against too short of follow-up time as being an important limiting factor for exploring the clinical value of ITC detection in node-negative patients. Irrespective of the presence of unspecific reactions, a method with higher sensitivity should detect, if present, more patients with clinically relevant positive BM findings. Our results may indicate a lower sensitivity of the ITC detection than Braun et al, although the BM cells have been processed similarly, and the same number of cells was examined in both studies. However, our efforts to increase the sensitivity by analyzing a higher number of cells by negative IMS did not alter the predictive value of the BM analysis in the node-negative group (Table 4 The proportion of positive BMs in this report is lower than in most other studies.3–8 This can partly be explained by a different stage distribution in the patient population. Also, traditional markers for aggressiveness, such as ER/PgR negativity and histologic grade 3, were less frequently registered in our study. Methodologic aspects may also have affected the results. Some studies have not reported an appropriate use of controls.4,6,8 A large proportion of false-positive results was eliminated in our study by the use of negative controls containing the same number of cells as the specific test.13 This reduced the positivity in our study from 18.0% to 13.2%. Differences in the use of negative controls might explain some of the variations in the reported rate of BM positivity. The choice of mAbs for the ICC detection and the morphologic criteria used for the scoring of tumor cells may also influence the results.4–6,8,12 The predilection of distant metastases to the skeleton in the BM-positive patients has also been shown by others.5,8 Interestingly, the rate of metastasis to the liver, but not to the lungs, was also significantly higher in BM-positive than in BM-negative patients. This indicates that disseminating tumor cells have different properties resulting in metastasis to distinct organs. Characterization of the tumor cells in BM may turn out to be important both for understanding the aspects of homing to particular organs and for unveiling the targets for therapy.17–23
By analysis of histologic grade in the primary tumor alone, approximately 80% of the T1N0 patients could be placed into a low-risk group (Fig 3 The question remains whether BM analysis should be performed routinely at the time of primary operation or, instead, be used in the monitoring of patients after receiving adjuvant systemic treatment. The mere detection of these cells in stage II breast cancer cannot be used for individualized adjuvant treatment decisions. Therefore, the routine use of this technique at the time of diagnosis must await studies in which further characterization of these cells can reveal predictive factors of value in the treatment decision. Several reports have shown the possibility for analysis of selected markers on disseminated cells, such as uPA/uPAR complex, c-erbB2, and p53.17,18,21 The clinical significance of such characterization of ITC has to be compared with the results of detailed array-based primary tumor analyses in the future.24–26 Presently, there is a lack of established surrogate markers for the evaluation of the efficacy of a selected systemic adjuvant therapy. With an increasing number of effective drugs available for use in breast cancer, methods for detection of nonresponders to any given treatment are needed. Recently, the presence of ITC in BM after adjuvant treatment has been shown to be associated with poor outcome.27,28 BM monitoring after the completion of systemic treatment opens the possibility for early treatment intervention for patients with persisting ITC in BM. In conclusion, this study confirms the independent prognostic value of detection of ITC in BM in breast cancer. However, detection of ITC cannot replace primary tumor analyses for optimal prognostication, but it is a supplement. In the future, extensive analysis of both primary tumor and ITC in BM are likely to constitute important tools for the management of breast cancer patients in general.
The authors indicated no potential conflicts of interest.
We thank the Micrometastasis laboratory, under the leadership of Anne Renolen, for the skilled processing and sample analyses.
Supported by the Norwegian Cancer Society and the Norwegian Foundation of Health and Rehabilitation, Oslo, Norway. Presented in part at the San Antonio Breast Cancer Conference, December 2002, and the International Conference on Tumor Cell Dissemination in Breast Cancer, Tuebingen, Germany, January 2003.
1. Early Breast Cancer Trialists’ Collaborative Group: Polychemotherapy for early breast cancer: An overview of the randomised trials. Lancet 352:930–942, 1998[CrossRef][Medline] 2. Early Breast Cancer Trialists’ Collaborative Group: Tamoxifen for early breast cancer: An overview of the randomised trials. Lancet 351:1451–1467, 1998[CrossRef][Medline] 3. Cote RJ, Rosen PP, Lesser ML, et al: Prediction of early relapse in patients with operable breast cancer by detection of occult bone marrow micrometastases. J Clin Oncol 9:1749–1756, 1991[Abstract]
4. Diel IJ, Kaufmann M, Costa SD, et al: Micrometastatic breast cancer cells in bone marrow at primary surgery: Prognostic value in comparison with nodal status. J Natl Cancer Inst 88:1652–1658, 1996
5. Braun S, Pantel K, Muller P, et al: 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 6. Mansi JL, Berger U, Easton D, et al: Micrometastases in bone marrow in patients with primary breast cancer: Evaluation as an early predictor of bone metastases. BMJ 295:1093–1096, 1987[Medline] 7. Harbeck N, Untch M, Pache L, et al: Tumour cell detection in the bone marrow of breast cancer patients at primary therapy: Results of a 3-year median follow-up. Br J Cancer 69:566–571, 1994[Medline]
8. Gebauer G, Fehm T, Merkle E, et al: Epithelial cells in bone marrow of breast cancer patients at time of primary surgery: Clinical outcome during long-term follow-up. J Clin Oncol 19:3669–3674, 2001 9. Funke I, Schraut W: Meta-analyses of studies on bone marrow micrometastases: An independent prognostic impact remains to be substantiated. J Clin Oncol 16:557–566, 1998[Abstract] 10. Mansi JL, Gogas H, Bliss JM, et al: Outcome of primary-breast-cancer patients with micrometastases: A long-term follow-up study. Lancet 354:197–202, 1999[CrossRef][Medline] 11. Braun S, Pantel K: Prognostic significance of micrometastatic bone marrow involvement. Breast Cancer Res Treat 52:201–216, 1998[CrossRef][Medline] 12. Borgen E, Naume B, Nesland JM, et al: Standardization of the immunocytochemical detection of cancer cells in BM and blood: I. Establishment of objective criteria for the evaluation of immunostained cells. Cytotherapy 1:377–388, 1998[CrossRef]
13. Naume B, Borgen E, Kvalheim G, et al: Detection of isolated tumor cells in bone marrow in early-stage breast carcinoma patients: Comparison with preoperative clinical parameters and primary tumor characteristics. Clin Cancer Res 7:4122–4129, 2001 14. Naume B, Borgen E, Nesland JM, et al: Increased sensitivity for detection of micrometastases in bone- marrow/peripheral-blood stem-cell products from breast-cancer patients by negative immunomagnetic separation. Int J Cancer 78:556–560, 1998[CrossRef][Medline] 15. 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 19:403–410, 1991[Medline] 16. Wiedswang G, Borgen E, Kåresen R, et al: Detection of isolated tumour cells in bone marrow from breast cancer patients: The significance of anterior and posterior iliac crest aspirations and the number of mononuclear cells analysed. Cytotherapy 5:40–45, 2003[Medline]
17. Braun S, Schlimok G, Heumos I, et al: ErbB2 overexpression on occult metastatic cells in bone marrow predicts poor clinical outcome of stage I–III breast cancer patients. Cancer Res 61:1890–1895, 2001 18. Heiss MM, Allgayer H, Gruetzner KU, et al: 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[CrossRef][Medline]
19. Pantel K, Cote RJ, Fodstad O: Detection and clinical importance of micrometastatic disease. J Natl Cancer Inst 91:1113–1124, 1999 20. Klein CA, Blankenstein TJF, Schmidt-Kittler O, et al: Genetic heterogeneity of single disseminated tumour cells in minimal residual cancer. Lancet 360:683–689, 2002[CrossRef][Medline]
21. Offner S, Schmaus W, Witter K, et al: p53 gene mutations are not required for early dissemination of cancer cells. Proc Natl Acad Sci USA 96:6942–6946, 1999
22. Mehes G, Witt A, Kubista E, et al: Circulating breast cancer cells are frequently apoptotic. Am J Pathol 159:17–20, 2001
23. Pantel K, Schlimok G, Braun S, et al: Differential expression of proliferation-associated molecules in individual micrometastatic carcinoma cells. J Natl Cancer Inst 85:1419–1424, 1993
24. Sorlie T, Perou CM, Tibshirani R, et al: Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci USA 98:10869–10874, 2001 25. Perou CM, Sorlie T, Eisen MB, et al: Molecular portraits of human breast tumours. Nature 406:747–752, 2000[CrossRef][Medline] 26. van’t-Veer LJ, Dai H, van de Vijver MJ, et al: Gene expression profiling predicts clinical outcome of breast cancer. Nature 415:530–536, 2002[CrossRef][Medline]
27. Braun S, Kentenich C, Janni W, et al: 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 28. Janni W, Hepp F, Rjosk D, et al: The fate and prognostic value of occult metastatic cells in the bone marrow of patients with breast carcinoma between primary treatment and recurrence. Cancer 92:46–53, 2001[CrossRef][Medline] Submitted February 3, 2003; accepted June 18, 2003. This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
10% (ER, PgR, cathepsin D, c-erbB-2) or 
2 test and linear by linear test were used to compare categorical variables. For statistical analysis, the SPSS (Version 10.1; SPSS, Inc, Chicago, IL) software was used. 
2 cm in 60.7% of the patients, and 62.8% of tumors were node-negative. ICC analysis of 2 x 106 cells disclosed one or more ITC in 13.2% of patients. The median age was 58 years in both the BM-positive and BM-negative group. Table 2
and Both Methods (n = 464)
