Sepp Kaul

Detection of tumor cells in bone marrow of patients with primary breast cancer: a prognostic factor for distant metastasis.

PURPOSE At the time of primary surgery, approximately 90% of all patients with breast cancer are free of metastases, but in the next 5 years almost 50% of them will relapse. We evaluated the significance of the presence of tumor cells in bone marrow of patients with primary breast cancer to investigate their predictive value for relapse. PATIENTS AND METHODS Two hundred sixty patients with primary breast cancer were examined for tumor cells in bone marrow aspirates taken from six sites of the skeleton. After density centrifugation, cells in interphase were smeared and stained. For the immunocytologic reaction, we used a new monoclonal antibody (2E11) that was reactive with the core protein of the tumor-associated glycoprotein TAG12. TAG12 is secreted by nearly all human breast carcinomas. RESULTS A significant correlation was found between tumor-cell detection and tumor stage (P < .0001), nodal status (P < .0001), and tumor grading (P = .002). A good relation to progesterone receptor (PR; P = .008) was found, but there was no correlation to estrogen receptor (ER) and menopausal status. Follow-up examinations showed distant metastases in 26 of 211 patients (15%). Twenty-two relapses occurred among the 81 patients with 2E11-positive cells in bone marrow, but only four occurred among the 130 patients without tumor-cell detection. CONCLUSIONS This study suggests that tumor-cell detection in bone marrow of patients with primary breast carcinoma is a good predictor for all distant relapses (P < .0005, Cox multiple regression analysis) and provides additional information in regard to other prognostic factors. The highest predicting value for distant metastasis results from the combination of nodal status, negative PR, and tumor-cell presence in bone marrow.

Expression of MUC-1 Epitopes on Normal Bone Marrow: Implications for the Detection of Micrometastatic Tumor Cells

PURPOSE The expression of the carcinoma-associated mucin MUC-1 is thought to be restricted to epithelial cells and is used for micrometastatic tumor cell detection in patients with solid tumors, including those with breast cancer. Little is known, however, about the expression of MUC-1 epitopes in normal hematopoietic cells. MATERIALS AND METHODS MUC-1 expression was analyzed by flow cytometry and immunocytology on bone marrow (BM) mononuclear cells and purified CD34+ cells from healthy volunteers, using different anti-MUC-1-specific monoclonal antibodies. In addition, Western blotting of MUC-1 proteins was performed. RESULTS Surprisingly, 2% to 10% of normal human BM mononuclear cells expressed MUC-1, as defined by the anti-MUC-1 antibodies BM-2 (2E11), BM-7, 12H12, MAM-6, and HMFG-1. In contrast, two antibodies recognizing the BM-8 and the HMFG-2 epitopes of MUC-1 were not detected. MUC-1+ cells from normal BM consisted primarily of erythroblasts and normoblasts. In agreement with this, normal CD34+ cells cultured in vitro to differentiate into the erythroid lineage showed a strong MUC-1 expression on day 7 proerythroblasts. Western blotting of these cells confirmed that the reactive species is the known high molecular weight MUC-1 protein. CONCLUSION Our data demonstrate that some MUC-1 epitopes are expressed on normal BM cells and particularly on cells of the erythroid lineage. Hence the application of anti-MUC-1 antibodies for disseminated tumor cell detection in BM or peripheral blood progenitor cells may provide false-positive results, and only carefully evaluated anti-MUC-1 antibodies (eg, HMFG-2) might be selected. Furthermore, MUC-1-targeted immunotherapy in cancer patients might be hampered by the suppression of erythropoiesis.

Telomerase activity correlates with tumor aggressiveness and reflects therapy effect in breast cancer

Telomerase appears to be an important factor for the control of cellular proliferation capacity and for tumorigenesis. Enzyme activity is highly increased in almost all human tumors and distinguishes them from benign lesions. Besides its diagnostic value, telomerase activity appears to be associated with tumor progression. The purpose of our study was to evaluate the significance of telomerase activity as a clinical marker in breast cancer. Twenty‐five tumor samples from breast cancer patients were analyzed retrospectively for telomerase activity in chemotherapy‐treated and untreated tumors. For each patient an identical number of cells was measured quantitatively for telomerase activity using the Telomerase‐PCR‐ELISA based on the TRAP (telomerase repeat amplification protocol) method. The findings were compared to clinical course, therapy and staging parameters. Telomerase activity was detected in all breast cancers. A significant correlation was found between enzyme activity and tumor size, lymph node status and stage: with ongoing tumor progression, telomerase activity appeared to increase in primary carcinomas. No correlation was seen between enzyme activity and the clinical course of patients. Without exception, telomerase activity was strongly decreased in all chemotherapy‐treated tumors compared to untreated tumors. Our preliminary data indicate that telomerase activity is associated with aggressiveness of breast tumors and appears to mirror the anti‐proliferative effects of chemotherapy. Int. J. Cancer (Pred. Oncol.) 79:8–12, 1998.© 1998 Wiley‐Liss, Inc.

Brain metastasis model in athymic nude mice using a novel MUC1-secreting human breast-cancer cell line, MA11.

The human breast epithelial cell line HBL 100 synthesizes and secretes FGF2, is able to grow in soft agar but is not tumorigenic in nude mice. Transfection of this cell line with the FGF4 gene led to its tumorigenic conversion in a dose‐dependent manner as assessed by growth under serum‐free conditions, growth in soft agar and growth as xenografts in nude mice. Clones of FGF4‐transfected cells producing different amounts of FGF4 secreted similar quantities of FGF2. Exagenously added recombinant FGF4 stimulated growth of clones that do not express this growth factor, but did not affect growth of FGF4‐producers. Neutralizing IgGs directed against FGF2 strongly inhibited growth of clones that do not produce FGF4, but did not affect growth of FGF4‐producers, indicating that the latter are FGF2‐independent. Cross‐linking experiments done with 125I‐FGF2 showed a down‐regulation of FGF receptors from the cell surface of FGF4‐expressing clones. Taken together, these results indicate that the FGF signalling pathway may be involved during tumoral progression of human breast epithelial cells and that there is a dose‐dependent relationship between the quantity of FGF4 produced and tumor development.

Immunocytochemical Detection of HPV High-Risk Type L1 Capsid Proteins in LSIL and HSIL as Compared with Detection of HPV L1 DNA

OBJECTIVE To investigate the prevalence of HPV L1 capsid proteins in HPV-infected HSIL and LSIL. STUDY DESIGN Cervical smears from 74 women with cytologically and histologically confirmed LSIL (n = 32) and HSIL (n = 42) were collected prospectively to detect HPV high-risk (hr) types 16, 18, 33, 35, 39, 45, 56 and 58 L1-DNA by standardized L1-consensus primer PCR (MY 09/11) and L1 capsid proteins by immunocytochemistry using monoclonal antibodies T31 (HPV16) and T16 (HPV hr) in a standardized protocol. RESULTS In HSIL and LSIL, L1 DNA was found for HPV hr in 93% and 59% and for HPV16 in 69% and 37% of the specimens, respectively. HPV L1 capsid proteins were detected in HSIL and LSIL for HPV hr in 33% and 44% and for HPV16 in 29% and 31% of the specimens, respectively. Expression of L1 capsid proteins was significantly reduced, by 59.6% for HPV hr L1 DNA-positive HSIL (P < .01) and by 40.4% for HPV 16 L1 DNA-positive HSIL (P < .01). In HPV 16 DNA-positive and HPV hr DNA-positive LSIL, no significant reduction of corresponding L1 capsid protein expression could be demonstrated. CONCLUSION These data suggest a disturbed viral cellular interaction in HPV 16 and HPV hr-infected HSIL, with loss of viral L1 capsid antigen. In this context there is a possible role of T31 and T16 as prognostic markers to predict the prognosis of CIN.

Immunomagnetic selection of CD34+ peripheral blood stem cells for autografting in patients with breast cancer

Contamination of transplants with tumour cells may contribute to relapse after peripheral blood stem cell transplantation (PBSCT). We studied the feasibility of CD34+ cell selection from blood‐derived autografts obtained following G‐CSF‐supported cytotoxic chemotherapy in a group of 25 patients with breast cancer (10 with high‐risk stage II/III and 15 with stage IV without bone or bone marrow involvement).

Serum levels of soluble E-selectin are associated with the clinical course of metastatic disease in patients with liver metastases from breast cancer.

Increasing evidence indicates that the expression of the endothelial adhesion molecule E-selectin is associated with progression and metastasis of breast cancer. Patients with liver metastases also show increased serum levels of the soluble form of E-selectin. It was our aim to compare serum levels of soluble E-selectin (sES) in such patients with the biology of the primary tumor and the course of the metastatic disease under therapy. We examined 69 patients with liver metastases from breast cancer who were selected to receive systemic tumor therapy because of progressive disease (n = 44) or newly detected liver metastases (n = 25). Serum concentrations of sES were measured before each therapy cycle using a specific ELISA. Serum concentrations of sES before the start of therapy were compared to clinical parameters and histopathological findings referring to the primary tumor. Secondly, serum levels of sES were compared to serum concentrations of the corresponding tumor markers. We observed a possible trend for certain unfavorable prognostic parameters (e.g., young women, low-graded tumors, human epidermal growth factor receptor 2 overexpression) to be related to higher serum levels of sES. Serum levels of sES were correlated with tumor marker levels in a logarithmical relation (r = 0.44, P < 0.0005). In some cases it could be demonstrated that serum levels of sES changed similarly to the course of tumor marker levels. We conclude that serum levels of sES are associated with the clinical course of liver metastases from breast cancer. Further investigations are needed to clarify if serum levels of sES may serve as tumor marker in certain clinical situations. E-selectin should be evaluated as a possible target for antimetastatic therapy studies.

Tandem High-Dose Chemotherapy in High-Risk Primary Breast Cancer: A Multivariate Analysis and a Matched-Pair Comparison With Standard-Dose Chemotherapy

Stem cell-supported high-dose chemotherapy (HDCT) is currently being evaluated in patients with high-risk primary breast cancer (HRPBC), as defined by extensive axillary lymph node involvement. Conclusive results from randomized studies with sufficient patient numbers and follow-up are pending. We retrospectively analyzed 144 HRPBC patients enrolled in a single-arm trial of tandem HDCT at the University of Heidelberg to evaluate the prognostic value of nodal ratio, HER2/neu status, and cytokeratin-positive bone marrow cells and to compare the outcomes of these patients with those of a conventionally treated control group of 91 patients matched by nodal ratio, tumor size, combined hormone-receptor status, and HER2/neu status. The tandem HDCT regimen consisted of 2 cycles of induction chemotherapy followed by 2 cycles of blood stem cell-supported high-dose ifosfamide, 12 g/m2; carboplatin, 900 mg/M2; and epirubicin, 180 mg/m2. Conventionally treated patients received a regimen containing anthracycline without taxanes (52 patients) or CMF (cyclophosphamide, methotrexate, and 5-flurouracil; 39 patients). With a median follow-up of 3.8 years, disease-free, distant disease-free, and overall survival rates were 62%, 65%, and 84%, respectively. In univariate analysis, besides the hormone receptor status (P = .007), HER2/neu overexpression was the strongest predictor of earlier death (P = .017). In multivariate analysis, a nodal ratio of > or =0.8 was found to be the only independent predictor of relapse (relative risk [RR] = 2.09; 95% confidence interval [CI], 1.21-3.60; P = .008) and only the absence of hormone receptors was associated with earlier death (RR = 3.59; 95% CI, 1.45-8.86; P = .006). Despite a trend toward later distant relapse after HDCT compared with standard-dose chemotherapy with a median follow-up of 3 years (P = .059), thus far, matched-pair analysis has not demonstrated significantly better survival rates after HDCT in all matched patients (P = .786) or in the subgroups of anthracycline-treated patients and patients with and without overexpression of HER2/neu. So far, the follow-up time has been too short to draw definite conclusions; however, patients with a nodal ratio of > or =0.8, receptor-negative tumors, or HER2/neu overexpression are at high risk for relapse and death, irrespective of the kind of adjuvant chemotherapy.

Serum Levels of Hepatocyte Growth Factor/Scatter Factor in Patients with Liver Metastases from Breast Cancer

Objective: Recent studies have shown that the pleiotropic cytokine hepatocyte growth factor/scatter factor (HGF/SF) and its receptor c-Met play major roles in the malignant progression of numerous tumors. For patients with breast cancer liver metastases, increased serum levels of HGF/SF have been reported. We studied the relationship between the clinical course of the disease and the serum levels of HGF/SF in such patients. Methods: We examined 51 patients treated for breast cancer liver metastases. Serum concentrations of HGF/SF were measured before each therapy cycle and compared to the corresponding tumor marker levels. Results: Mean serum levels of HGF/SF in patients with liver metastases were increased above the reported reference levels of primary breast cancer patients. Serum levels of HGF/SF were correlated with tumor marker levels in a logarithmic relation (r = 0.47, p < 0.001). In some cases serum concentrations of HGF/SF changed similarly to the course of the corresponding tumor markers. Conclusions: Serum levels of HGF/SF are associated with the clinical course of metastatic breast cancer patients with liver metastases. Further studies are required to clarify the potential value of the HGF/SF serum concentration as a tumor marker. HGF/SF and its receptor c-Met should be further evaluated as therapeutic targets.

Detection of tumor cells in leukapheresis products from patients with breast cancer using immunocytochemical staining method.

Abstract We used a combination of 4 monoclonal antibodies (BM7, BM8 against MUC1, 5D3 aganist CK8,18,19 and HEA125 against human epithelial antigen) and a sensitive immunocytochemical staining using cytospin preparation to identify breast tumor cells in leukapheresis products (LP). This assay allowed detection of one tumor cell in 1×106 mononuclear cells (MC). In clinical specimens, tumor cells were detected in LP from 6 of 42 (14.3%) patients in the adjuvant treatment group, from 2 of 11 (18.2%) patients in the neoadjuvant treatment group and from 9 of 43 (20.1%) in the group of patients with metastatic disease. Tumor cell counts ranged from 0.25–5 cells in 1×106 normal cells per LP. The median tumor cell concentration was higher in specimens from patients with metastatic disease (median=0.96) than in specimens from patients in the adjuvant and neoadjuvant treament groups (median=0.5 and 0.75). No significant differences between the epithelial cell positive group and the epithelial cell negative group with respect to tumor size, lymph nodes involvement, tumor grade, histological type and receptor were found. We conclude that immunocytochemical staining of cytospin preparation is a sensitive and simple method to detect and quantitate breast cancer cells in LP.