Abstract
Background: The occurrence of disseminated tumour cells in bone marrow of patients with breast cancer is linked to a worse prognosis. We present a method for DTC detection from bone marrow samples based on immunocytochemistry, using breast cancer-associated glycosylation molecules as markers for detection and characterization. Materials and Methods: A double immunofluorescence staining of a pan-cytokeratin (CK) marker and either Tn or O-Acetyl-GD3 was carried out in artificial and patient bone marrow samples. Results: The sample in which most cells stained positive for CK/Tn and CK/O-AC-GD3, was obtained from a patient who certainly had remote metastases. All other bone marrow samples showed heterogenous staining, so no correlation to tumour characteristics could be revealed. Conclusion: A certain characterization of tumour cells can be achieved by a double staining of bone marrow samples with CK and a glycosylation marker. For future studies, analysis should be extended to a larger patient collective and further examination of more glycosylation markers should be carried out.
As one out of eight women is diagnosed with breast cancer during her lifetime (1), this malignancy is still the most frequent cancer type worldwide and the leading cause of cancer-related death in women. Thereby, the ultimate cause for death is not the primary tumour itself, but remote metastasis of the tumour in different organs. These metastases arise, according to current models, from single cells, that detach from the primary tumour, travel through the blood stream and lymphatic vessels, and settle down at different sites of the body (2). These cells are called circulating tumour cells (CTCs) (3, 4) as long as they are within blood or lymph, and turn into disseminated tumour cells (DTCs) when they invade bone marrow (5). In the bone marrow these cells can form tumour reservoirs, meaning they can survive there for a long time periods in a dormant state (2, 6). If these cells are activated by any stimulus, they become aggressive, forming metastases not only within bone marrow but also spreading from there (7). Formerly, this event had been regarded as one of the late events during cancer progression, but nowadays it is known, that this activation of sleeping tumour cells can also happen early during tumorigenesis (8). Therefore, the appearance of CTCs and DTCs in patients with primary epithelial tumours is generally linked to a poor prognosis and a worse outcome (9-11). The prognostic relevance of DTCs (12-14) is also supported by the fact, that their presence in bone marrow was associated with a higher tumour stage, lymph node metastasis and low hormone receptor expression levels (15). The predictive value, in contrast, is still to be clarified (16). Another requirement for the use of DTCs in tumour evaluation is the refinement of DTC detection methods, and discovery of a small number of DTCs in a background of large amount of blood or bone marrow cells (1 tumour cell per 1×106-7 surrounding cells; (17)), being still a technical challenge. Therefore, in most DTC-detection methods the first step is enrichment of tumour cells, mostly based on the epithelial surface marker EpCAM (18), but the outstanding disadvantage of this marker is its heterogenous expression in tumour samples (10, 19) and can become down-regulated by processes like epithelial-mesenchymal transition (18) and thereby escape enrichment. In the on-hand study we present an immunocytochemical method for DTC-detection, based on density gradient centrifugation for tumour cell enrichment, followed by double staining with fluorescently-labelled antibodies against tumour specific glycosylation molecules and a pan-cytokeratin marker. The anti-pan-cytokeratin-marker used here is also routinely used in the APAAP-staining of tissue samples for the detection of tumour cells (20, 21). The Thomsen nouvelle (Tn)-antigen, a glycosylation variant bound to Mucin (22) which is a precursor of the MN-blood group antigens (23), is known to be expressed exclusively in breast cancer tissue (24) and is involved in cell adhesion and metastasis (22). It had been shown, that an increased amount of Tn is correlated to a reduced 5-year survival, a higher TNM-staging and grading (25) and is therefore an important molecule for cancer diagnosis, prognosis and therapy. The other molecule used in the on hand study was 9-O-Ac-GD3, a ganglioside, which has higher expression levels in tumour tissues than in normal tissues (26, 27). Its use as a target for creating tumour vaccines remains in question (28). Its levels seem to decrease with the loss of differentiation, so a down-regulation of 9-O-Ac-GD3 could be regarded as a marker for poor prognosis (29). The methodology allows a fast and cost-efficient detection and characterization of disseminated tumour cells, which might help in the selection of the appropriate therapeutical approach.
Materials and Methods
Blood samples. For verification of the staining method and antibody function, immunocytochemistry was first carried out on blood samples with and without addition of breast cancer cell line cells and on samples consisting only of cell line cells.
Therefore, 10 ml blood were drawn from healthy donors, diluted with PBS (Biochrom, Berlin, Germany), carefully layered onto Histopaque 1077 and centrifuged for 30 min. at 400 × g. Afterwards buffy coat was aspired, filled up with PBS and washed by another centrifugation step at 250 × g for 10 min. Supernatant was removed, cell pellet was resuspended in 2 ml PBS and counted using a Neubauer counting chamber. Cytospins were prepared from 500,000 leucocytes or 500,000 leucocytes with addition of 200 tumor cells by down-centrifugation to cover slips at 500 × g for 5 min. The slides were air-dried and kept on −80°C until use.
Cell lines. Breast cancer cell lines Cama-1, MCF-7 and ZR75-1 (ATCC: HTB-21: mammary gland adenocarcinoma, 57136: mammary gland adenocarcinoma; CRL-1500, ductal carcinoma; ATCC, Wesel, Germany) were cultured according to data sheet of the provider. Cells were grown to sub-confluency and splitted twice a week. By addition of trypsin/EDTA (Biochrom, Berlin, Germany) cells could be detached from the ground of the cell culture flask and counted in a hemocytometer. Thereafter 200 tumor cells (an equal mixture of the three cell lines) was added to a processed blood sample, or were used for preparation of slides containing tumour cells only.
Bone marrow samples. Bone marrow samples of breast cancer patients were collected during primary breast cancer surgery. An ethical vote, conform to the Declaration of Helsinki, allowing this procedure, was available (LMU 148-12), and written consent was obtained from the patients.
Two-three ml of bone marrow were used for the experiments. Samples were filled with 25 ml Hanks balanced salt solution (Biochrom, Berlin, Germany) and centrifuged at 170 × g and 9°C for 10 min. Afterwards the upper lipid phase was carefully aspired and discarded, the fraction containing bone marrow cells was transferred in a new tube, together with 8ml Ficoll Paque (Invitrogen, Darmstadt, Germany) and centrifuged again at 1,105 × g for 20 min. Buffy Coat was aspired, diluted with PBS and washed by centrifugation at 535 × g for 10 min. Then supernatant was removed, harvested bone marrow cells were resuspended in 5ml PBS and counted in a Neubauer counting chamber. A total of 1×106 bone marrow cells were used for preparation of a cytospin by down-centrifugation to a cover slip at 535 × g for 5 min. Samples were air-dried and kept at −80°C until use.
Immunocytochemistry. For the immunocytochemical staining coverslips containing either only leucocytes, leucocytes with breast cancer cell line cells, cell line cells alone or bone marrow cells were thawed carefully and fixed in aceton (Merck, Darmstadt, Germany) for about 5min. Then slides were washed several times in PBS, blocked with a 5% BSA (Invitrogen) in PBS solution for 15 min. and primary antibody was applied for 45 min. in a humid chamber in the best-suited dilution (Anti-PanCytokeratin mouse IgG 1:100 (AS Diagnostics, Hueckeswagen, Germany), anti-Thomsen-nouvelle mouse IgM (ThermoFisher, Rockford, IL, USA) 1:100 or anti-O-Ac-GD3 mouse IgM (Antibodies online GmbH, Aachen, Germany) 1:500, respectively, diluted in DAKO S3022 (DAKO, Hamburg, Germany)). The primary antibody was then washed-off with PBS and the secondary, fluorescence-coupled antibody was layered onto the slides (goat-anti mouse IgG-Cy3 (Millipore, Darmstadt, Germany), 1:400, goat-anti mouse IgM-FITC (Santa Cruz Biotechnology Inc., Heidelberg, Germany), 1:20 with DAKO S3022 (DAKO, Hamburg, Germany)) and incubated for 30 min. in a humid chamber. Slides were again washed with PBS and blocked with an Ultra UV Block (ThermoFisher, Rockford, IL, USA) for 15 min, before the secondary antibody was applied in the same manner. Afterwards slides were air-dried, one drop of Vectashield mounting medium with DAPI (Axxora, Lörrach, Germany) was applied and the sample is overcast with a coverslip and sealed with nail polish.
Evaluation. Fluorescent staining was inspected on a Zeiss Axioskop Epifluorescence Microskope with a high resolution Axio Cam MRm CCD-camera (Zeiss, Wetzlar, Germany). Pictures were taken via Axio Vision Release 4.8.2 (06-2010) software and processed by Photoshop (CS 5). Stained cells were counted manually by two independent scientists and counted numbers were averaged. The statistical analysis, examining for a coincidence of stained cells and tumour characteristics was carried out by SPSS V. 20.0.
Results
For immunocytochemical experiments, it is crucial to have obtained positive and negative control stained samples before beginning analysis of precious patient samples. The controls consist of blood samples from healthy donors with and without addition of breast cancer cell line cells and samples prepared from cell line cells only, which also serve for determination of appropriate antibody dilution to be used for further experiments. In our study positive control samples (only tumour cells) stained at about 99%, while negative control samples (lymphocytes only) did not yield any staining. In samples, in which 200 cells from cell lines were mixed to half a million of blood cells, almost all tumor cells could be recovered by immunocytochemical staining, resulting in recovery rates of about 87%.
Immunocytochemical double fluorescent staining of bone marrow samples from different breast cancer patients for CK and the Tn-antigen included in the on hand study. Merged image: blue, DAPI; red, CK; green, Tn.
Immunocytochemical double fluorescent staining of bone marrow samples from different breast cancer patients for CK and the O-Ac-GD3-antigen included in the on hand study. Merged image: blue, DAPI; red, CK; green=O-Ac-GD3.
Based on these findings, 27 bone marrow samples from breast cancer patients were double-stained with antibodies against CK and Tn (Figure 1) or CK and O-Ac-GD3 (Figure 2). Numbers of stained cells in every sample were noted and then set into correlation with respective tumour characteristics (Table I). It was interesting to see, that the only patient, who showed up with metastasis formation (patient 8) had a rather large number of stained cells, so there might be a coherence. All other samples showed more heterogenous staining, for example many stained cells were found in some pT2 samples (sample 10, 20), but similar cell counts were found in a pT1c sample (sample 22) and the pT3 sample (sample 12) has rather low number of stained cells. Also the two patients suffering from triple-negative breast cancer (sample 18 and 24) did not show a similar tendency for the number of stained cells counted, as cell counts for patient 18 are rather low and in the bone marrow sample for patient 24 a higher numer of stained cells could be observed. In a small number of patient samples Ki67 was measured as a cell proliferation marker, but also cell proliferation does not seem to influence the number of DTC, for example in patient samples 19 and 20 Ki67 was measured with 10%, numbers of stained DTCs in turn vary a lot. In sample 21 Ki67 is at 20%, DTC-counts in contrast are a lot lower than for the two 10%-Ki67 samples. Unfortunately for patient 9 no data concerning the tumour were available. So up to now no concrete relation between tumour characteristics and the number of CK/Tn or CK/GD3-stained cells could be figured out. Dividing the patients into sub-groups concerning age at primary diagnosis, tumour size, menopausal stage, lymph node affection, grading and estrogen- or progesterone receptor positivity, no statistical significant difference could be found as all p-values were higher than the significance level of 0.05 (Table II).
Number of CK-/Tn/O-Ac-GD3 stained cells in bone marrow samples in comparison to tumour characteristics of the respective patient.
Statistical analysis of patient subgroups in correlation to number of stained cells. No statistical significant differences in stainings could be seen for different patient sub-groups.
Discussion
The aim of the on-hand study was to present a method for immunocytochemical detection and characterization of DTCs in breast cancer. Staining was first tested in an artificial system consisting of blood samples from healthy donors spiked with carcinoma cell line cells. As recovery rates were acceptable, staining was carried out on bone marrow samples of breast cancer patients. The results presented within this study showed a tendency for appearance of a high number of stained cells (DTCs) in metastatic breast cancer patients. Due to the small number of bone marrow samples available for study, this result has to be considered preliminary. A larger number of patient samples would also be helpful in statistical analysis of patient sub-groups. Furthermore it might be a benefit to the study, if simultaneously to the immunocytochemical staining a gold standard method for CTC and DTC detection, like for example the CellSearch® system (30) was run. Thereby results of both methods could be compared and maybe thereby results of the staining procedure could be clarified more in detail. By these improvements the immunocytochemical staining could turn into a fast and cost-efficient method for detection and furthermore a characterization of DTCs from bone marrow samples. A recent approach for DTC-detection comparing qPCR and immunocytochemical staining for DTC-detection form bone marrow of primary breast cancer patients ended in favour of the multimarker-qPCR, for which cytokeratin 19, mamaglobin and TWIST were used as markers. For immunocytochemistry, in turn, only anti-cytokeratin-antibodies were used, and both methods had a low congruency (31), demonstrating the necessity of testing further markers for DTC-detection, especially in immunohistochemical approaches, as it was done in the on-hand study. The detection of DTCs is rather important as their occurrence is predictive for a locoregional relapse. It was presumed that DTCs might be able to recirculate from the bone marrow to the site of the primary tumor, what might lead to recurrence of the disease (32). DTCs also play an important role during breast cancer treatment, as their appearance might be a surrogate marker for insufficiently treated patients (33), so that a second line therapy with, e.g. bisphosphonates, has to be considered(34). Therefore DTC-detection could be useful for diagnostic procedures, helping to adjust therapeutical strategies to each individual patient.
Acknowledgements
The Authors thank T. Thormeyer for helping with statistical analysis.
- Received April 18, 2016.
- Revision received May 20, 2016.
- Accepted May 23, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.