- Email: [email protected]
Antigenic profiles of disseminated breast tumour cells and microenvironment in bone marrow
EJSO 2003; 29: 121±126 doi:10.1053/ejso.2002.1334
Antigenic profiles of disseminated breast tumour cells and microenvironment in bone marrow J. Zhang*, K.-W. Shen*, G. Liu*, J. Zhou², Q. Shen², Z.-Z. Shen* and Z.-M. Shao* *Department of Breast Surgery and ²Central Laboratory, Cancer Hospital/Cancer Institute, Fudan University, Shanghai, P. R. China
Aims: Thirty per cent of breast cancer patients with axillary lymph node negative at primary surgery will relapse within 10 years. This may be caused by disseminated tumour cells from the primary tumour. This study report the phenotypic profiles of disseminated tumour cells and microenvironmental characteristics in bone marrow of breast cancer. Methods: We detected the biologic markers on the disseminated tumour cells with immunocytochemical staining, analysed the immunological changes through flow-cytometry, and investigated the u-PA activity in the plasma of bone marrow. Results: With the immunocytochemical staining of EMA and CK19, we detected micrometastasis in thirty out of 72 (41.67%) breast cancer patients. Compared with the primary tumours, disseminated tumour cells expressed low protein cyclin D1, P53, Ki-67, EGFR, and high protein P21. The percentage of memory CD4 T cells was significantly higher in the micrometastasis-positive group than in the micrometastasis-negative group. Tumour size and axillary lymph node status were found to be significantly correlated with the u-PA activity level. Conclusions: Immunophenotypic profiles of disseminated tumour cells could be measured by immunocytochemical # 2002 Elsevier Science Ltd. All rights reserved. staining and microenvironment can be analysed by flow cytometry. Key words: breast cancer; bone marrow; disseminated tumour cells; phenotype.
INTRODUCTION Thirty per cent of breast cancer patients with axillary lymph node negative at primary surgery will relapse within 10 years. This is caused by disseminated tumour cells derived from the primary tumour.1 Bone marrow is an important site for micrometastases. Bone marrow might act as a reservoir for disseminated epithelial tumour cells.2 After the initial neoplastic transformation, tumour cells undergo further genetic changes resulting in selection of dominant clones with growth advantage and various metastatic potentials.3 The biologic characteristics of disseminated tumour cells are poorly understood.4±10 Once tumour cells and tumour cell emboli have reached the vascular or lymphatic compartments, they must survive a variety of immunological challenges. Factors which may impact on the metastatic process11 include a group of proteases which degrade
Correspondence to: Kun-Wei Shen, MD, Department of Breast Surgery, Cancer Hospital/Cancer Institute, Fudan University, 399 Ling-Ling Road, Shanghai, 200032, P. R. China. Fax: 86-21-64174774; E-mail: [email protected] 0748±7983/03/$30.00
extracellular matrix proteins and basement membranes. u-PA may have metastatic relevance and prognostic significance breast cancer.12,13 We here report the phenotypic profiles of disseminated tumour cells and microenviromental characteristics in the bone marrow in breast cancer patients, and the u-PA activity in the plasma of bone marrow.
PATIENTS AND METHODS Patients From December 2001 to March 2002, 72 primary breast cancer patients treated in Department of Breast Surgery, Shanghai Cancer Hospital, Fudan University, were studied. The age of the patients ranged from 32 to 73 years with a mean age of 50.9 years. Of these, 44.4% were pre-menopausal and 55.6% were postmenopausal (Table 1). Investigations to exclude the distant metastasis included chest radiography, mammography, ultrasound detection of the liver, and bone scans. No patients showed evidence of distant metastasis. All the patients underwent the surgery within two weeks after diagnosis. We also performed bone marrow aspiration #
2002 Elsevier Science Ltd. All rights reserved.
122
J. ZHANG ET AL.
Table 1 The correlation between the clinicopathologic factors and tumour cell detection Number
Micrometastases () patients (%)
P value
Menopausal status Pre-menopause Post-menopause
32 40
10 20
0.320
Side Left Right
32 40
16 14
0.500
Tumour size (cm) ,2 2±5 >5
26 38 8
8 14 8
0.400
ALNM* (ÿ) 1±3 4
40 14 18
18 2 10
0.227
Histological type Ductal Lobular Medullary DCIS Others
56 4 4 4 4
28 0 0 2 0
LVI² (ÿ) ()
50 12
20 4
1.000
ER status (ÿ) ()
32 36
16 10
0.291
Her-2 (ÿ) ()
6 62
2 24
1.000
Cyclin D1 (ÿ) ()
26 30
12 8
0.433
P53 (ÿ) ()
12 54
10 16
0.025
0.399
* ALNM: axillary lymph node metastasis. ² LVI: lymph/vascular invasion.
on six benign breast disease patients as the normal controls. Bone marrow aspiration was performed with written patient consent form and following an approved institutional reviewed protocol.
Preparation of bone marrow and immunocytochemical staining Ten-ml bone marrow samples were obtained from the sternum intraoperatively, and stored in two tubes with heparin. One five-ml sample was centrifuged through a Ficoll±Hypaque density gradient (density, 1.077 0.002 g/ml, Shanghai) at 2000 g for 20 min.
Mononucleated interface cells were collected, erythrocytes were destroyed by 10 mmol/l EDTA, and 106 cell were centrifuged onto seven glass slides for immunocytochemical staining. Antigens studied include EMA (dilution 1:50, Cell Marque), CK19 (dilution 1:50, Dako), P53 (dilution 1:50, Cell Marque), P21 (dilution 1:50, Cell Marque), Cyclin D1 (dilution 1:300, Dako), Ki-67 (dilution 1:100, Dako), EGFR (dilution 1:300, Dako). EMA is directed against membrane-bound surface antigens and CK19 against the cytoskeleton of epithelial cells. Micrometastasis was defined as the presence of one or more anaplastic cells with membrane staining of EMA or/and plasma staining of CK19 was confirmed. Morphologic features, including cell size and nuclear texture, were also used to increase the specificity. Smears were considered positive for micrometastases if P53, P21, Cyclin D1, positivity; Ki-67 with distinct nuclear staining and EGFR with membrane staining were identified. All slides were examined independently by two pathologists.
Flow cytometric analysis of lymphocyte makers The following monoclonal antibodies were used for flow cytometric analysis. CD3-PE, CD4-PE and CD45RO-FITC, CD8-PE and CD45RO-FITC, CD56-PE and CD45RO-FITC. IgG1-FITC and IgG1-PE were also used as the isotype controls; 10 000 gated viable lymphocytes were analysed using a FACS Caliber flowcytometer and CELLQuest software program (Becton Dickinson, San Jose, CA, USA). Windows for analysis of granulocytes and lymphocytes were set using sidescatter and forward-scatter dot plots.
u-PA activity assay In conclusion, micrometastases can be detected in bone marrow and their expression of a variety of surface CD antigens can be quantified by flow-cytometry. The u-PA ELISA kit detects both free active and inactive u-PA in the circulation. u-PA can catalyse the conversion of plasminogen into the active enzyme plasmin, and plasmin can activate type IV collagenase, which then degrades collagen and proteins of the surrounding tumour matrix and basement membranes. Five-ml bone marrow was centrifuged at 2000 g for 10 min to separate plasma, which was stored at ÿ80 C pending further assay. The u-PA activity of the plasma in the bone marrow was assayed using the CHEMICON u-PA assay kit (ECM 600, Chemicon International, CA), according to the manufacturer's instructions. In brief, positive control and samples were added to a clean 96-well plate, and the chromogenic substrate was added to each well. After incubation at 37 C for 4 h, absorbances were measured at 405 nm on an automated microplate reader (ELX 800, USA). All samples were run
ANTIGENIC PROFILES IN TUMOUR CELLS AND BONE MARROW
123
in duplicate, and inter-assay variations were less than 10%. Five ml blood was obtained simultaneously to time with bone marrow aspiration, and its u-PA activity was also analysed for each patient.
Statistical analysis We use statistic software SPSS10.0. The differences of the immunological parameters and u-PA activity between the groups were analysed by the independent-samples t-test. Frequencies were compared using chi-squared test for qualitative variable. The paired comparison of primary and disseminated tumour cells was performed using the paired w2 test.
RESULTS The correlation between primary tumour characteristics and micrometastases We detected micrometastases in 30 of 72 breast cancer patients. Large primary tumours were associated with a higher incidence of micrometastasis than were small tumours (P 0.040). Axillary lymph node involvement was not found to be correlated with the presence of occult tumour cells in the bone marrow. Eighteen patients had micrometastases without axillary lymph node involvement, and 20 patients had axillary lymph node involvement without micrometastasis. Of the biologic factors evaluated, only P53 protein expression was significantly associated with the incidence of micrometastases (P 0.025) (Table 1). Our method seemed to be specific because no disseminated tumour cells were found in six patients with benign breast conditions.
Phenotype of disseminated tumour cells in bone marrow Of 30 breast cancer patients with micrometastasis, two and ten patients also overexpressed P53 and cyclin D1 on the disseminated tumour cells. P21 was overexpressed for six cases, EGFR for six cases, and Ki67 for three cases, in 16 patients with disseminated tumour cells (Fig. 1). We also compared primary and disseminated tumour cells. Disseminated and primary tumour cells overexpressed cyclin D1 in eight, and six out of 18 patients, respectively and overexpressed P53 in four and fourteen, respectively, out of 22 patients where both samples were evaluated. There was no significant difference between primary and disseminated tumours for cyclin D1 and P53 protein expression (paired w2 test, w2 0.5 and 1.14, respectively, P . 0.05).
Figure 1 Phenotype of primary tumours and disseminated tumour cells in bone marrow. In our study, P53, Cyclin D1, P21, EGFR and Ki-67 proteins were over-expressed in 72.7%, 50.0%, 31.9%, 45.1% and 34.5%, respectively. On the disseminated tumour cells, P53, Cyclin D1, P21, EGFR and Ki-67 protein overexpression accounted for 6.67%, 33.3%, 37.5% and 18.75% of the patients who had micrometastases in bone marrow.
Analysis of immunological parameters in the bone marrow We investigated the immunological variation in the bone marrow micrometastases between different groups (Fig. 2). Compared with the normal controls, breast cancer patients had significantly higher CD4, CD4/CD8, percentage of the memory CD8 T cell and the ratio of the memory helper T cell and memory suppressor T cell (P 0.050, 0.007, 0.018 and 0.003, respectively). There was no correlation between the immunological parameters and axillary node status. The percentage of memory CD4 T cell and the ratio of the memory helper T cell and memory suppressor T cell were significantly higher in micrometastases () group than those in the micrometastases (ÿ) group (P 0.033 and 0.049, respectively) (Table 2).
Analysis of u-PA activity in the plasma of bone marrow Forty-two breast cancer patients were evaluated for u-PA activity in bone marrow. Compared with patients with benign breast disorder, breast cancer patients were significantly correlated with high u-PA activity in bone marrow (P 0.032). We also studied the u-PA activity in the blood of the breast cancer patients, and found a marginal elevation for u-PA activity compared with the bone marrow (P 0.043). Among the clinicopathologic factors, tumour size and axillary lymph node status were significantly correlated with the u-PA activity level in the bone marrow (0.01 and 0.000, respectively). However,
124
J. ZHANG ET AL. we did not find significant difference for u-PA activity between micrometastasis-positive group and -negative group (Table 3).
DISCUSSION
Figure 2 Analysis of immunological parameters in the bone marrow. We used the following monoclonal antibodies to explore the immunological parameters of a breast cancer patient: CD-3PE, CD4-PE and CD45RO-FITC, CD8-PE and CD45RO-FITC, CD56-PE and CD45RO-FITC. A, CD4: 24.8% (4.3% 20.5%); CD4CD45RO: 20.5%; CD4/ CD4 CD45RO : 82.7% (20.5%/24.8%). B, C: with a similar method, we can get other parameters from representative flow-cytometry dot plots in B and C.
Occult dissemination of tumour cells in patients with operable breast cancer can cause metastases. It is usually missed by conventional tumour staging.14,15 Immunocytochemical detection assays now enable the detection of one disseminated cancer cell in a background of 105±106 non-malignant bone marrow cells.16 Occult tumour cells may be present in bone marrow in 20% to 45% of operable breast cancer patients and 20±70% of advanced breast cancer patients.4 In our study, with the monoclonal antibody of EMA and CK19, we detected disseminated tumour cells in 30 out of 72 breast cancer patients. In 1980, Sloane et al. reported the detection of occult tumour cells in the bone marrow of breast cancer patients.17 Micrometastasis may be as an independent prognostic factor.6,18,19 Braun et al. reported that the presence of occult bone marrow micrometastasis was independent of the axillary lymph node status (P 0.13).20 In our study, primary tumour size is significantly associated with the incidence of micrometastasis. The specific biologic characteristics of the disseminated tumour cells remain poorly understood.4±10 Jiang et al. reported an inverse relationship between p21 and mutant-type p53 expression in human breast cancer.21 Pantel found 11 of 69 cancer patients expressed Ki-67 or P120 on disseminated tumour cells.8 We also found low Ki67 and EGFR expression on the disseminated tumour cells which may be non-cycling and nonproliferation. Cell clones with altered protein expression profiles which confer more aggressive growth properties may have evolutionary advantages in clone selection processes. Detection of micrometastases in bone marrow may help to identify the earliest metastatic population of tumour cells, and thus elucidate the events involved in the early metastatic process. Ree et al. compared the immunomagnetically sorted target cells from bone marrow, positive axillary lymph nodes and primary tumours by differential display analysis. They identified two candidate genes, differentially expressed in bone marrow, which were involved in stimulatory signal transduction, cell cycle regulation, and angiogenesis.3 CD45 is a transmembrane protein tyrosine phosphatase. When activated, it changes from CD45RA to a lower-molecular weight form CD45RO.22 Memory T cell with the marker CD45RO is characterized by the ability to generate an earlier, more intense and long-lasting immune response, Feuerer et al. reported the enrichment of memory T cells, in the bone marrow
ANTIGENIC PROFILES IN TUMOUR CELLS AND BONE MARROW
125
Table 2 Analysis of the immunological variation in the bone marrow between different groups Group A (n 54) Breast cancer
Group B (n 6) Normal control
Group C (n 28) Micrometastases ( ÿ )
Group D (n 26) Micrometastases ()
58.8 8.0 *29.1 8.0 29.5 7.0 *1.1 0.4 20.6 7.5 54.7 13.1 *42.1 12.6 *1.4 0.6
58.1 5.5 39.3 11.0 24.3 13.0 2.0 1.2 21.9 2.7 41.7 14.6 23.5 4.4 2.8 1.3
57.9 9.4 26.8 6.9 28.8 5.2 0.9 0.2 19.8 8.1 ²48.7 11.7 41.4 14.5 ²1.2 0.4
59.8 6.5 31.5 8.6 30.2 8.6 1.3 0.5 21.5 7.1 61.2 11.6 43.2 10.6 1.7 0.8
CD3 CD4 CD8 CD4/CD8 CD56 CD4CD45RO/CD4 CD8CD45RO/CD8 CD4CD45RO/CD8CD45RO
CD3: total T cell; CD4: help T cell; CD8: suppressor T cell; CD56: including NK cell; CD45RO: memory T cell; CD4CD45RO/CD4: the percentage of the memory help T cell; CD8CD45RO/CD8: the percentage of the memory suppressor T cell; CD56CD45RO/CD56: the percentage of the memory CD56 cell; CD4CD45RO/ CD8CD45RO: the ratio of the memory helper T cell and memory suppressor T cell. * Group A was compared with Group B, P , 0.05. ² Group C was compared with Group D, P , 0.05.
Table 3 Analysis of u-PA activity in the plasma of bone marrow Number u-PA activity (unit) P value Age (y) 50 . 50
26 16
2.5 1.8 2.6 1.3
0.868
Tumour size (cm) ,2 2±5 .5
12 24 6
2.0 0.3 2.3 1.1 4.9 2.7
0.01
ALNM* (ÿ) ()
22 20
2.0 0.5 3.2 2.1
0.000
Micrometastases (ÿ) ()
30 12
2.6 1.6 2.4 1.7
0.769
Histological type Ductal Lobular Others
32 4 6
2.5 1.6 2.1 0.3 3.3 1.9
ER status (ÿ) ()
16 26
1.9 0.3 2.9 1.9
0.674
0.166
* ALNM: axillary lymph node metastasis.
of breast cancer patients.11 We found elevation of memory lymphocytes, and particularly CD8 T cells; in the breast cancer patients compared with normal controls. We also found the percentage of memory CD4 T cell was significantly higher in the micrometastasis-positive group than that in micrometastasis-negative group. Further follow-up may identify the correlation between our indices and the development of overt metastasis.
The u-PA content of cytosols or detergent extracts of breast cancer tissues may be significantly correlated with prolonged disease-free survival.23±25 Our study showed that the patients with large tumour size and axillary lymph node metastasis were likely to have high u-PA activity, as in other studies.23,25 Through long-time follow-up, we hope to identify whether high levels of uPA activity correlate with the development of overt metastasis.
ACKNOWLEDGEMENTS This research was supported in part by the Outstanding Young Investigator Award of National Natural Science Foundation of China (30025015) and Award of Shanghai Scientific and Technologic Committee (014119032).
REFERENCES 1. 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 2001; 19: 3669±74. 2. Pantel K, Schlimok G, Angstwurm M et al. Methodological analysis of immunocytochemical screening for disseminated epithelial tumour cells in bone marrow. J Hematother 1994; 3: 165±73. 3. Ree AH, Engebraaten O, Hovig E et al. Differential display analysis of breast carcinoma cells enriched by immunomagnetic target cell selection: gene expression profiles in bone marrow target cells. Int J Cancer 2002; 97: 28±33. 4. Diel IJ, Cote RJ. Bone marrow and lymph node assessment for minimal residual disease in patients with breast cancer. Cancer Treat Rev 2000; 26: 53±65. 5. Solomayer EF, Diel IJ, Meyberg GC et al. Prognostic relevance of cathepsin D detection in micrometastatic cells in the bone marrow of patients with primary breast cancer. Breast Cancer Res Treat 1998; 49: 145±54.
126 6. Solomayer EF, Diel IJ, Wallwiener D et al. Prognostic relevance of urokinase plasminogen activator detection in micrometastatic cells in the bone marrow of patients with primary breast cancer. Br J Cancer 1997; 76: 812±8. 7. Braun S, Hepp F, Sommer HL et al. Tumour-antigen heterogeneity of disseminated breast cancer cells: implications for immunotherapy of minimal residual disease. Int J Cancer 1999; 84: 1±5. 8. Pantel K, Schlimok G, Braun S et al. Differential expression of proliferation-associated molecules in individual micrometastatic carcinoma cells. J Natl Cancer Inst 1993; 85(17): 1419±24. 9. Passlick B, Sienel W, Seen-Hibler R et al. The 17-1A antigen is expressed on primary, metastatic and disseminated non-small cell lung carcinoma cells. Int J Cancer 2000; 87: 548±52. 10. Putz E, Witter K, Offner S et al. Phenotypic characteristics of cell lines derived from disseminated cancer cells in bone marrow of patients with solid epithelial tumours: establishment of working models for human micrometastasis. Cancer Res 1999; 59: 241±8. 11. Feuerer M, Rocha M, Bai L et al. Enrichment of memory T cells and other profound immunological changes in the bone marrow from untreated breast cancer patients. Int J Cancer 2001; 92: 96±105. 12. Pedersen AN, Christensen IJ, Stephens RW et al. The complex between urokinase and its type-1 inhibitor in primary breast cancer: relation to survival. Cancer Res 2000; 60: 6927±34. 13. Fisher JL, Field CL, Zhou H et al. Urokinase plasminogen activator system gene expression is increased in human breast carcinoma and its bone metastases ± a comparison of normal breast tissue, non-invasive and invasive carcinoma and osseous metastases. Breast Cancer Res Treat 2000; 61: 1±12. 14. Hosch SB, Braun S, Pantel K. Characterization of disseminated tumour cells. Semin Surg Oncol 2001; 20: 265±71. 15. Braun S, Rosenberg R, Thorban S et al. Implications of occult metastatic cells for systemic cancer treatment in patients
J. ZHANG ET AL.
16. 17. 18. 19. 20. 21. 22.
23.
24. 25.
with breast or gastrointestinal cancer. Semin Surg Oncol 2001; 20: 334±46. Lacroix J, Doeberitz MK. Technical aspects of minimal residual disease detection in carcinoma patients. Semin Surg Oncol 2001; 20: 252±64. Sloane JP, Ormerod MG, Neville AM. Potential pathological application of immunocytochemical methods to the detection of micrometastasis. Cancer Res 1980; 40: 3079±82. Mansi JL, Gogas H, Bliss JM et al. Outcome of primarybreast-cancer patients with micrometastases: a long-term follow-up study. Lancet 1999; 17: 197±202. Diel IJ. Bone marrow staging for breast cancer: is it better than axillary node dissection? Semin Oncol 2001; 28: 236±44. 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 2000; 342: 525±33. Jiang M, Shao ZM, Wu J et al. p21/waf1/cip1 and mdm-2 expression in breast carcinoma patients as related to prognosis. Int J Cancer 1997; 74: 529±34. Muraro PA, Pette M, Bielekova B et al. Human autoreactive CD4 T cells from naive CD45RA and memory CD45RO subsets differ with respect to epitope specificity and functional antigen avidity. J Immunol 2000; 164: 5474±81. Janicke F, Pache L, Schmitt M et al. Both the cytosols and detergent extracts of breast cancer tissues are suited to evaluate the prognostic impact of the urokinase-type plasminogen activator and its inhibitor, plasminogen activator inhibitor type 1. Cancer Res 1994; 54: 2527±30. Duffy MJ, Reilly D, McDermott E et al. Urokinase plasminogen activator as a prognostic marker in different subgroups of patients with breast cancer. Cancer 1994; 74: 2276±80. Xiao J, Zhang G, Xia W. Expression and significance of urokinasetype plasminogen activator in breast cancer. Zhonghua Zhong Liu Za Zhi 1999; 21: 196±8.
Antigenic profiles of disseminated breast tumour cells and microenvironment in bone marrow J. Zhang*, K.-W. Shen*, G. Liu*, J. Zhou², Q. Shen², Z.-Z. Shen* and Z.-M. Shao* *Department of Breast Surgery and ²Central Laboratory, Cancer Hospital/Cancer Institute, Fudan University, Shanghai, P. R. China
Aims: Thirty per cent of breast cancer patients with axillary lymph node negative at primary surgery will relapse within 10 years. This may be caused by disseminated tumour cells from the primary tumour. This study report the phenotypic profiles of disseminated tumour cells and microenvironmental characteristics in bone marrow of breast cancer. Methods: We detected the biologic markers on the disseminated tumour cells with immunocytochemical staining, analysed the immunological changes through flow-cytometry, and investigated the u-PA activity in the plasma of bone marrow. Results: With the immunocytochemical staining of EMA and CK19, we detected micrometastasis in thirty out of 72 (41.67%) breast cancer patients. Compared with the primary tumours, disseminated tumour cells expressed low protein cyclin D1, P53, Ki-67, EGFR, and high protein P21. The percentage of memory CD4 T cells was significantly higher in the micrometastasis-positive group than in the micrometastasis-negative group. Tumour size and axillary lymph node status were found to be significantly correlated with the u-PA activity level. Conclusions: Immunophenotypic profiles of disseminated tumour cells could be measured by immunocytochemical # 2002 Elsevier Science Ltd. All rights reserved. staining and microenvironment can be analysed by flow cytometry. Key words: breast cancer; bone marrow; disseminated tumour cells; phenotype.
INTRODUCTION Thirty per cent of breast cancer patients with axillary lymph node negative at primary surgery will relapse within 10 years. This is caused by disseminated tumour cells derived from the primary tumour.1 Bone marrow is an important site for micrometastases. Bone marrow might act as a reservoir for disseminated epithelial tumour cells.2 After the initial neoplastic transformation, tumour cells undergo further genetic changes resulting in selection of dominant clones with growth advantage and various metastatic potentials.3 The biologic characteristics of disseminated tumour cells are poorly understood.4±10 Once tumour cells and tumour cell emboli have reached the vascular or lymphatic compartments, they must survive a variety of immunological challenges. Factors which may impact on the metastatic process11 include a group of proteases which degrade
Correspondence to: Kun-Wei Shen, MD, Department of Breast Surgery, Cancer Hospital/Cancer Institute, Fudan University, 399 Ling-Ling Road, Shanghai, 200032, P. R. China. Fax: 86-21-64174774; E-mail: [email protected] 0748±7983/03/$30.00
extracellular matrix proteins and basement membranes. u-PA may have metastatic relevance and prognostic significance breast cancer.12,13 We here report the phenotypic profiles of disseminated tumour cells and microenviromental characteristics in the bone marrow in breast cancer patients, and the u-PA activity in the plasma of bone marrow.
PATIENTS AND METHODS Patients From December 2001 to March 2002, 72 primary breast cancer patients treated in Department of Breast Surgery, Shanghai Cancer Hospital, Fudan University, were studied. The age of the patients ranged from 32 to 73 years with a mean age of 50.9 years. Of these, 44.4% were pre-menopausal and 55.6% were postmenopausal (Table 1). Investigations to exclude the distant metastasis included chest radiography, mammography, ultrasound detection of the liver, and bone scans. No patients showed evidence of distant metastasis. All the patients underwent the surgery within two weeks after diagnosis. We also performed bone marrow aspiration #
2002 Elsevier Science Ltd. All rights reserved.
122
J. ZHANG ET AL.
Table 1 The correlation between the clinicopathologic factors and tumour cell detection Number
Micrometastases () patients (%)
P value
Menopausal status Pre-menopause Post-menopause
32 40
10 20
0.320
Side Left Right
32 40
16 14
0.500
Tumour size (cm) ,2 2±5 >5
26 38 8
8 14 8
0.400
ALNM* (ÿ) 1±3 4
40 14 18
18 2 10
0.227
Histological type Ductal Lobular Medullary DCIS Others
56 4 4 4 4
28 0 0 2 0
LVI² (ÿ) ()
50 12
20 4
1.000
ER status (ÿ) ()
32 36
16 10
0.291
Her-2 (ÿ) ()
6 62
2 24
1.000
Cyclin D1 (ÿ) ()
26 30
12 8
0.433
P53 (ÿ) ()
12 54
10 16
0.025
0.399
* ALNM: axillary lymph node metastasis. ² LVI: lymph/vascular invasion.
on six benign breast disease patients as the normal controls. Bone marrow aspiration was performed with written patient consent form and following an approved institutional reviewed protocol.
Preparation of bone marrow and immunocytochemical staining Ten-ml bone marrow samples were obtained from the sternum intraoperatively, and stored in two tubes with heparin. One five-ml sample was centrifuged through a Ficoll±Hypaque density gradient (density, 1.077 0.002 g/ml, Shanghai) at 2000 g for 20 min.
Mononucleated interface cells were collected, erythrocytes were destroyed by 10 mmol/l EDTA, and 106 cell were centrifuged onto seven glass slides for immunocytochemical staining. Antigens studied include EMA (dilution 1:50, Cell Marque), CK19 (dilution 1:50, Dako), P53 (dilution 1:50, Cell Marque), P21 (dilution 1:50, Cell Marque), Cyclin D1 (dilution 1:300, Dako), Ki-67 (dilution 1:100, Dako), EGFR (dilution 1:300, Dako). EMA is directed against membrane-bound surface antigens and CK19 against the cytoskeleton of epithelial cells. Micrometastasis was defined as the presence of one or more anaplastic cells with membrane staining of EMA or/and plasma staining of CK19 was confirmed. Morphologic features, including cell size and nuclear texture, were also used to increase the specificity. Smears were considered positive for micrometastases if P53, P21, Cyclin D1, positivity; Ki-67 with distinct nuclear staining and EGFR with membrane staining were identified. All slides were examined independently by two pathologists.
Flow cytometric analysis of lymphocyte makers The following monoclonal antibodies were used for flow cytometric analysis. CD3-PE, CD4-PE and CD45RO-FITC, CD8-PE and CD45RO-FITC, CD56-PE and CD45RO-FITC. IgG1-FITC and IgG1-PE were also used as the isotype controls; 10 000 gated viable lymphocytes were analysed using a FACS Caliber flowcytometer and CELLQuest software program (Becton Dickinson, San Jose, CA, USA). Windows for analysis of granulocytes and lymphocytes were set using sidescatter and forward-scatter dot plots.
u-PA activity assay In conclusion, micrometastases can be detected in bone marrow and their expression of a variety of surface CD antigens can be quantified by flow-cytometry. The u-PA ELISA kit detects both free active and inactive u-PA in the circulation. u-PA can catalyse the conversion of plasminogen into the active enzyme plasmin, and plasmin can activate type IV collagenase, which then degrades collagen and proteins of the surrounding tumour matrix and basement membranes. Five-ml bone marrow was centrifuged at 2000 g for 10 min to separate plasma, which was stored at ÿ80 C pending further assay. The u-PA activity of the plasma in the bone marrow was assayed using the CHEMICON u-PA assay kit (ECM 600, Chemicon International, CA), according to the manufacturer's instructions. In brief, positive control and samples were added to a clean 96-well plate, and the chromogenic substrate was added to each well. After incubation at 37 C for 4 h, absorbances were measured at 405 nm on an automated microplate reader (ELX 800, USA). All samples were run
ANTIGENIC PROFILES IN TUMOUR CELLS AND BONE MARROW
123
in duplicate, and inter-assay variations were less than 10%. Five ml blood was obtained simultaneously to time with bone marrow aspiration, and its u-PA activity was also analysed for each patient.
Statistical analysis We use statistic software SPSS10.0. The differences of the immunological parameters and u-PA activity between the groups were analysed by the independent-samples t-test. Frequencies were compared using chi-squared test for qualitative variable. The paired comparison of primary and disseminated tumour cells was performed using the paired w2 test.
RESULTS The correlation between primary tumour characteristics and micrometastases We detected micrometastases in 30 of 72 breast cancer patients. Large primary tumours were associated with a higher incidence of micrometastasis than were small tumours (P 0.040). Axillary lymph node involvement was not found to be correlated with the presence of occult tumour cells in the bone marrow. Eighteen patients had micrometastases without axillary lymph node involvement, and 20 patients had axillary lymph node involvement without micrometastasis. Of the biologic factors evaluated, only P53 protein expression was significantly associated with the incidence of micrometastases (P 0.025) (Table 1). Our method seemed to be specific because no disseminated tumour cells were found in six patients with benign breast conditions.
Phenotype of disseminated tumour cells in bone marrow Of 30 breast cancer patients with micrometastasis, two and ten patients also overexpressed P53 and cyclin D1 on the disseminated tumour cells. P21 was overexpressed for six cases, EGFR for six cases, and Ki67 for three cases, in 16 patients with disseminated tumour cells (Fig. 1). We also compared primary and disseminated tumour cells. Disseminated and primary tumour cells overexpressed cyclin D1 in eight, and six out of 18 patients, respectively and overexpressed P53 in four and fourteen, respectively, out of 22 patients where both samples were evaluated. There was no significant difference between primary and disseminated tumours for cyclin D1 and P53 protein expression (paired w2 test, w2 0.5 and 1.14, respectively, P . 0.05).
Figure 1 Phenotype of primary tumours and disseminated tumour cells in bone marrow. In our study, P53, Cyclin D1, P21, EGFR and Ki-67 proteins were over-expressed in 72.7%, 50.0%, 31.9%, 45.1% and 34.5%, respectively. On the disseminated tumour cells, P53, Cyclin D1, P21, EGFR and Ki-67 protein overexpression accounted for 6.67%, 33.3%, 37.5% and 18.75% of the patients who had micrometastases in bone marrow.
Analysis of immunological parameters in the bone marrow We investigated the immunological variation in the bone marrow micrometastases between different groups (Fig. 2). Compared with the normal controls, breast cancer patients had significantly higher CD4, CD4/CD8, percentage of the memory CD8 T cell and the ratio of the memory helper T cell and memory suppressor T cell (P 0.050, 0.007, 0.018 and 0.003, respectively). There was no correlation between the immunological parameters and axillary node status. The percentage of memory CD4 T cell and the ratio of the memory helper T cell and memory suppressor T cell were significantly higher in micrometastases () group than those in the micrometastases (ÿ) group (P 0.033 and 0.049, respectively) (Table 2).
Analysis of u-PA activity in the plasma of bone marrow Forty-two breast cancer patients were evaluated for u-PA activity in bone marrow. Compared with patients with benign breast disorder, breast cancer patients were significantly correlated with high u-PA activity in bone marrow (P 0.032). We also studied the u-PA activity in the blood of the breast cancer patients, and found a marginal elevation for u-PA activity compared with the bone marrow (P 0.043). Among the clinicopathologic factors, tumour size and axillary lymph node status were significantly correlated with the u-PA activity level in the bone marrow (0.01 and 0.000, respectively). However,
124
J. ZHANG ET AL. we did not find significant difference for u-PA activity between micrometastasis-positive group and -negative group (Table 3).
DISCUSSION
Figure 2 Analysis of immunological parameters in the bone marrow. We used the following monoclonal antibodies to explore the immunological parameters of a breast cancer patient: CD-3PE, CD4-PE and CD45RO-FITC, CD8-PE and CD45RO-FITC, CD56-PE and CD45RO-FITC. A, CD4: 24.8% (4.3% 20.5%); CD4CD45RO: 20.5%; CD4/ CD4 CD45RO : 82.7% (20.5%/24.8%). B, C: with a similar method, we can get other parameters from representative flow-cytometry dot plots in B and C.
Occult dissemination of tumour cells in patients with operable breast cancer can cause metastases. It is usually missed by conventional tumour staging.14,15 Immunocytochemical detection assays now enable the detection of one disseminated cancer cell in a background of 105±106 non-malignant bone marrow cells.16 Occult tumour cells may be present in bone marrow in 20% to 45% of operable breast cancer patients and 20±70% of advanced breast cancer patients.4 In our study, with the monoclonal antibody of EMA and CK19, we detected disseminated tumour cells in 30 out of 72 breast cancer patients. In 1980, Sloane et al. reported the detection of occult tumour cells in the bone marrow of breast cancer patients.17 Micrometastasis may be as an independent prognostic factor.6,18,19 Braun et al. reported that the presence of occult bone marrow micrometastasis was independent of the axillary lymph node status (P 0.13).20 In our study, primary tumour size is significantly associated with the incidence of micrometastasis. The specific biologic characteristics of the disseminated tumour cells remain poorly understood.4±10 Jiang et al. reported an inverse relationship between p21 and mutant-type p53 expression in human breast cancer.21 Pantel found 11 of 69 cancer patients expressed Ki-67 or P120 on disseminated tumour cells.8 We also found low Ki67 and EGFR expression on the disseminated tumour cells which may be non-cycling and nonproliferation. Cell clones with altered protein expression profiles which confer more aggressive growth properties may have evolutionary advantages in clone selection processes. Detection of micrometastases in bone marrow may help to identify the earliest metastatic population of tumour cells, and thus elucidate the events involved in the early metastatic process. Ree et al. compared the immunomagnetically sorted target cells from bone marrow, positive axillary lymph nodes and primary tumours by differential display analysis. They identified two candidate genes, differentially expressed in bone marrow, which were involved in stimulatory signal transduction, cell cycle regulation, and angiogenesis.3 CD45 is a transmembrane protein tyrosine phosphatase. When activated, it changes from CD45RA to a lower-molecular weight form CD45RO.22 Memory T cell with the marker CD45RO is characterized by the ability to generate an earlier, more intense and long-lasting immune response, Feuerer et al. reported the enrichment of memory T cells, in the bone marrow
ANTIGENIC PROFILES IN TUMOUR CELLS AND BONE MARROW
125
Table 2 Analysis of the immunological variation in the bone marrow between different groups Group A (n 54) Breast cancer
Group B (n 6) Normal control
Group C (n 28) Micrometastases ( ÿ )
Group D (n 26) Micrometastases ()
58.8 8.0 *29.1 8.0 29.5 7.0 *1.1 0.4 20.6 7.5 54.7 13.1 *42.1 12.6 *1.4 0.6
58.1 5.5 39.3 11.0 24.3 13.0 2.0 1.2 21.9 2.7 41.7 14.6 23.5 4.4 2.8 1.3
57.9 9.4 26.8 6.9 28.8 5.2 0.9 0.2 19.8 8.1 ²48.7 11.7 41.4 14.5 ²1.2 0.4
59.8 6.5 31.5 8.6 30.2 8.6 1.3 0.5 21.5 7.1 61.2 11.6 43.2 10.6 1.7 0.8
CD3 CD4 CD8 CD4/CD8 CD56 CD4CD45RO/CD4 CD8CD45RO/CD8 CD4CD45RO/CD8CD45RO
CD3: total T cell; CD4: help T cell; CD8: suppressor T cell; CD56: including NK cell; CD45RO: memory T cell; CD4CD45RO/CD4: the percentage of the memory help T cell; CD8CD45RO/CD8: the percentage of the memory suppressor T cell; CD56CD45RO/CD56: the percentage of the memory CD56 cell; CD4CD45RO/ CD8CD45RO: the ratio of the memory helper T cell and memory suppressor T cell. * Group A was compared with Group B, P , 0.05. ² Group C was compared with Group D, P , 0.05.
Table 3 Analysis of u-PA activity in the plasma of bone marrow Number u-PA activity (unit) P value Age (y) 50 . 50
26 16
2.5 1.8 2.6 1.3
0.868
Tumour size (cm) ,2 2±5 .5
12 24 6
2.0 0.3 2.3 1.1 4.9 2.7
0.01
ALNM* (ÿ) ()
22 20
2.0 0.5 3.2 2.1
0.000
Micrometastases (ÿ) ()
30 12
2.6 1.6 2.4 1.7
0.769
Histological type Ductal Lobular Others
32 4 6
2.5 1.6 2.1 0.3 3.3 1.9
ER status (ÿ) ()
16 26
1.9 0.3 2.9 1.9
0.674
0.166
* ALNM: axillary lymph node metastasis.
of breast cancer patients.11 We found elevation of memory lymphocytes, and particularly CD8 T cells; in the breast cancer patients compared with normal controls. We also found the percentage of memory CD4 T cell was significantly higher in the micrometastasis-positive group than that in micrometastasis-negative group. Further follow-up may identify the correlation between our indices and the development of overt metastasis.
The u-PA content of cytosols or detergent extracts of breast cancer tissues may be significantly correlated with prolonged disease-free survival.23±25 Our study showed that the patients with large tumour size and axillary lymph node metastasis were likely to have high u-PA activity, as in other studies.23,25 Through long-time follow-up, we hope to identify whether high levels of uPA activity correlate with the development of overt metastasis.
ACKNOWLEDGEMENTS This research was supported in part by the Outstanding Young Investigator Award of National Natural Science Foundation of China (30025015) and Award of Shanghai Scientific and Technologic Committee (014119032).
REFERENCES 1. 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 2001; 19: 3669±74. 2. Pantel K, Schlimok G, Angstwurm M et al. Methodological analysis of immunocytochemical screening for disseminated epithelial tumour cells in bone marrow. J Hematother 1994; 3: 165±73. 3. Ree AH, Engebraaten O, Hovig E et al. Differential display analysis of breast carcinoma cells enriched by immunomagnetic target cell selection: gene expression profiles in bone marrow target cells. Int J Cancer 2002; 97: 28±33. 4. Diel IJ, Cote RJ. Bone marrow and lymph node assessment for minimal residual disease in patients with breast cancer. Cancer Treat Rev 2000; 26: 53±65. 5. Solomayer EF, Diel IJ, Meyberg GC et al. Prognostic relevance of cathepsin D detection in micrometastatic cells in the bone marrow of patients with primary breast cancer. Breast Cancer Res Treat 1998; 49: 145±54.
126 6. Solomayer EF, Diel IJ, Wallwiener D et al. Prognostic relevance of urokinase plasminogen activator detection in micrometastatic cells in the bone marrow of patients with primary breast cancer. Br J Cancer 1997; 76: 812±8. 7. Braun S, Hepp F, Sommer HL et al. Tumour-antigen heterogeneity of disseminated breast cancer cells: implications for immunotherapy of minimal residual disease. Int J Cancer 1999; 84: 1±5. 8. Pantel K, Schlimok G, Braun S et al. Differential expression of proliferation-associated molecules in individual micrometastatic carcinoma cells. J Natl Cancer Inst 1993; 85(17): 1419±24. 9. Passlick B, Sienel W, Seen-Hibler R et al. The 17-1A antigen is expressed on primary, metastatic and disseminated non-small cell lung carcinoma cells. Int J Cancer 2000; 87: 548±52. 10. Putz E, Witter K, Offner S et al. Phenotypic characteristics of cell lines derived from disseminated cancer cells in bone marrow of patients with solid epithelial tumours: establishment of working models for human micrometastasis. Cancer Res 1999; 59: 241±8. 11. Feuerer M, Rocha M, Bai L et al. Enrichment of memory T cells and other profound immunological changes in the bone marrow from untreated breast cancer patients. Int J Cancer 2001; 92: 96±105. 12. Pedersen AN, Christensen IJ, Stephens RW et al. The complex between urokinase and its type-1 inhibitor in primary breast cancer: relation to survival. Cancer Res 2000; 60: 6927±34. 13. Fisher JL, Field CL, Zhou H et al. Urokinase plasminogen activator system gene expression is increased in human breast carcinoma and its bone metastases ± a comparison of normal breast tissue, non-invasive and invasive carcinoma and osseous metastases. Breast Cancer Res Treat 2000; 61: 1±12. 14. Hosch SB, Braun S, Pantel K. Characterization of disseminated tumour cells. Semin Surg Oncol 2001; 20: 265±71. 15. Braun S, Rosenberg R, Thorban S et al. Implications of occult metastatic cells for systemic cancer treatment in patients
J. ZHANG ET AL.
16. 17. 18. 19. 20. 21. 22.
23.
24. 25.
with breast or gastrointestinal cancer. Semin Surg Oncol 2001; 20: 334±46. Lacroix J, Doeberitz MK. Technical aspects of minimal residual disease detection in carcinoma patients. Semin Surg Oncol 2001; 20: 252±64. Sloane JP, Ormerod MG, Neville AM. Potential pathological application of immunocytochemical methods to the detection of micrometastasis. Cancer Res 1980; 40: 3079±82. Mansi JL, Gogas H, Bliss JM et al. Outcome of primarybreast-cancer patients with micrometastases: a long-term follow-up study. Lancet 1999; 17: 197±202. Diel IJ. Bone marrow staging for breast cancer: is it better than axillary node dissection? Semin Oncol 2001; 28: 236±44. 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 2000; 342: 525±33. Jiang M, Shao ZM, Wu J et al. p21/waf1/cip1 and mdm-2 expression in breast carcinoma patients as related to prognosis. Int J Cancer 1997; 74: 529±34. Muraro PA, Pette M, Bielekova B et al. Human autoreactive CD4 T cells from naive CD45RA and memory CD45RO subsets differ with respect to epitope specificity and functional antigen avidity. J Immunol 2000; 164: 5474±81. Janicke F, Pache L, Schmitt M et al. Both the cytosols and detergent extracts of breast cancer tissues are suited to evaluate the prognostic impact of the urokinase-type plasminogen activator and its inhibitor, plasminogen activator inhibitor type 1. Cancer Res 1994; 54: 2527±30. Duffy MJ, Reilly D, McDermott E et al. Urokinase plasminogen activator as a prognostic marker in different subgroups of patients with breast cancer. Cancer 1994; 74: 2276±80. Xiao J, Zhang G, Xia W. Expression and significance of urokinasetype plasminogen activator in breast cancer. Zhonghua Zhong Liu Za Zhi 1999; 21: 196±8.