Introduction
It is still unclear which factors determine complete remission in some AML cases or no remission whatever. It has been suggested however, that the induction of efficient anti-tumour immunity play a critical role in successful treatment of this disease. Various cell-mediated immune systems such as cytotoxic T lymphocytes (CTL), lymphokine-activated killer (LAK) cells and activated macrophages have been reported to be involved in the immune surveillance against leukaemia, especially - in the first AML phase [1].
CTL as well as LAK cells express CD2 antigen, and adhesion molecule CD58 is the ligand for the CD2 on human T cells and have been shown to provide a costimulatory signal for T cell activation [2, 3, 4]. Antigen-specific T cell activation requires the presence of accessory (antigen-presenting) cells. In vitro studies have demonstrated that AML blasts can act as accessory cells during mitogenic activation and during recognition of alloantigens and leukaemia-specific antigenic epitopes [5]. Two types of signals are usually required to initiate the program of T cell activation: specific recognition of HLA-presented antigenic peptides and antigen non-specific signals collectively called costimulation [6, 7]. According to Brouwer et al [8] the expression increase of such adhesive molecules as CD58 or constimulative B7 molecules on AML myeloblasts raises their sensitivity to cytotoxic T lymphocytes.
The aim of this work was to determine the expression level (intensity) of CD58 and HLA-DR on myeloblasts in AML cases as well the connection of the expression intensity with the achievement of remission in AML patients.
Material and Methods
Patients
Fifty-three newly diagnosed and untreated adult acute myeloid leukaemia M1 (18/53, 11 males, 7 females) and M2 (35/53, 20 males, 15 females) patients were subjected to this study. The diagnosis was based on bone marrow smears, cytochemistry and immunophenotyping according to the FAB group criteria [9]. All patients were treated by intensive combination chemoteraphy (3+7) (cytosine arabinoside (ARA-C) for 7 days plus daunorubicin push in the first 3 days of ARA-C administration). If no CR was achieved after first course, the same therapy was repeated. Remission (CR+) or lack of remission (CR-) was estimated from bone marrow smears and immunophenotyping bone marrow aspirates in flow cytometer. Remission was determined when patients had normal peripheral blood counts and normocellular bone marrow with less than 5% blasts.
Nineteen patients: 10 males and 9 females with a mean age 38 ± 17yr (range 17-72) and with white blood cells (WBC) count 55.9 ± 106.5 x 109/l (range 1.2-393.0) achieved remission. Thirty-four patients: 24 males and 13 females with a mean age 55 ± 15 yr. (range 23-75) and with WBC count 75.9 ± 121.3 x 109/l (range 0.8-450.0) not achieved remission.
Cell preparation and immunofluorescence
Fresh, heparinized bone marrow aspirates and venous blood (2 ml) was diluted with an equal volume of PBS. In 42 cases, bone marrow samples were collected, and in 11cases leukemic cells from peripheral blood were used (bone marrow was not available). Cells were incubated with appropriate monoclonal mouse-anti human antibodies. Cell surface markers have been evaluated by means of direct immunofluorescense using fluorescein (FITC) and phycoerythrine (PE) conjugated monoclonal antibodies against antigens: CD45, CD14, CD13, CD33, CD117, and anti -CD58 and anti -HLA-DR (DAKO, Copenhagen, Denmark, Immunotech Marseille France, Becton Dickinson). Isotype specifics FITC or PE - coupled control antibodies were used as controls.
Flow cytometry studies
The samples were analysed on flow cytometers: CytoronAbsolute (Ortho) or/and FACS Calibur (BD).
To determine antibody expression the percentage of positive events in the leukemic blast population was calculated and compared to the isotype control [10]. For a quantitative analysis of the fluorescence expression intensity histograms of two studied molecules and their isotype control were made and the mean fluorescence intensity, as defined by the mean channel peak fluorescence, was performed. The relative fluorescence intensity - RFI - in arbitrary units (AU) was calculated from these values [10]:
For direct quantitative analysis the QuantiBRITE test (BD) was applied. The mean channels of PE fluorescence were defined and antibody-bounding capacity (ABC) was then calculated using QuantiCALC software [11].
Statistical analysis
Significance of differences was assessed by t-Students test using the program Statistica. Differences were considered significant for p values <0.05. Correlation coefficient was determined by linear regression.
Results
The CD58 (LFA-3) adhesive molecule was present practically in all examined patients on the majority of myeloblasts at diagnosis: only in one AML M2 patient myeloblasts were LFA-3 -negative. However, considered differences in expression intensity of CD58 between individual patients.
After therapy one group of patients achieved remission (CR+) and second group patients no received remission (CR-). The analysis expression intensity of CD58, determined at diagnosis on myeloblasts, showed that in patients easily achieving remission studied intensity was higher than in those with no remission (Table 1, Fig. 1, Fig. 2). The mean CD58 RFI value in CR+ patients calculated as 15,1 AU (arbitrary units) was significantly higher than in CR- patients - 10,2 (Table 1, Fig. 3). The high values expression intensity of CD58 positively correlated with complete remission after inductive therapy (correlation coefficient - r = 0.77).
Parallel, both CR+ and CR- patients had HLA-DR expression on nearly all myeloblasts, but expression intensity of HLA-DR at diagnosis differed between the studied groups of patients (Table 1, Fig. 1, Fig. 2). In CR+ group a significantly higher expression intensity of HLA-DR on myeloblasts was observed as compared to group without remission (mean RFI - 24.4 AU in CR+ group versus mean RFI for HLA-DR - 17.0 in CR- group) (Tab. 1, Fig. 3). The high values of intensity HLA-DR expression positively correlated with complete post-inductive remission (r = 0.72).
A preliminary analysis of antibody bounding capacity (ABC) was performed for the CD58 and HLA-antigens. The direct method of expression intensity measurement also revealed significant differences between the examined groups. In patients with easily acquired remission the CD58 and HLA-DR antigen values on myeloblasts at diagnosis were significantly higher than in those without remission. The range of CD58 ABC values for CR+ group was 4200-6800 and for CR- group was 590-3000 and the range of HLA-DR ABC values was 36 600-88 800 and 5 500-28 000 in CR+ and CR- group respectively (Table. 1).
Apart from the positive correlation between remission and high CD58 and HLA-DR expression intensity, a positive correlation between level of CD58 and HLA-DR expression intensity was also noted (r=0.79).
Discussion
Blocking studies with anti-CD58 monoclonal antibody demonstrated that this pathway could be involved when AML blasts are used as accessory cells. However the final effects of signalling through that pathway differ between AML patients [5, 13, 14] as well as other authors state that nearly all AML blast cells reveal a CD58 expression. In the present study, the differences in the expression intensity determined at diagnosis on myeloblasts between patients with remission and no remission seem to be significant for prognostic purposes, both with the direct and indirect method of determining the intensity.
The our observation, that the high LFA-3 expression intensity (expressed as the RFI coefficient and ABC values) appeared at diagnosis on myeloblasts of CR+ patients was significantly different from the low RFI and ABC values on myeloblasts of CR- patients, is new.
AML blasts from different patients usually express high level of the antigen presenting HLA molecules [6]. In the present study it has been determined that the HLA-DR expression intensity is similar as in the case of CD58, both for patients with easily acquired remission and those with no remission whatever. The RFI of CD58 values on myeloblasts of patients with easily acquired remission were no lower than ca. 13 AU, while the HLA-DR values were ca. 22 AU. There were only several patients, with no remission, where the RFI values for CD58 were 13 and 14 AU and for HLA-DR - 22 AU (Fig. 1). Most probably in such cases it was difficult to achieve remission because other factors were involved.
According to Bruserud et al. studies [5, 6], myeloblasts of AML patients may act as intermediates sending appropriate signals necessary for T cell activation and function as accessory antigen presenting cells, through presence of HLA particles on their surface. He has determined that the presence of adhesive molecule CD58, costimulative CD80/CD86 molecules and HLA is necessary to activate this process. He also found [5] that several problems must be overcome before a successful therapy is introduced. One is to determine the reason for a wide heterogeneity of patients - the therapeutic effect is then more predictable.
In summary, the current studies reveal that not only CD58 and HLA presence on myeloblasts is necessary to anti-tumour immunity, equally important is an appropriate density of these molecules on leukemic cells. Our results show that the easy achieving remission can be expected in patients with high expression intensity both of CD58 adhesive molecule and HLA antigen.
Acknowledgements The authors express their sincere appreciation for physicians` from Department Internal and Haematology (Head - prof. dr hab. med. Lech Konopka) and from Department of Haematology (Head - prof. dr hab. med. Stanisław Maj), Institute Haematology and Blood Transfusion, for clinical data of studied patients.
References
1. Archimbaud E, Thomas X, Campos S, et al. (1992): Expression of surface adhesion molecules CD54 (ICAM-1) and CD58 (LFA-3) in adult acute leukemia: relationship with initial characteristics and prognosis. Leukemia 6: 265-271.
2. Bain BJ, Clark D, Lampert IA, Wilkins BS (2001): Special techniques applicate to bone marrow diagnosis. Blackwell Science 3-edition Oxford.
3. Brouwer RE, Zwinderman KH, Kluin-Nelemans HC, et al. (2000): Expression and induction of costimulatory and adhesion molecules on acute myeloid leukemic cells: Implications for adoptive immunotherapy. Exper Hematol 28: 161-168.
4. Bruserud O (1999): Acute myelogenus leukemia blasts as accessory cells during T lymphocyte activation: possible implications for therapeutic strategies. Leukemia 13: 1175-1187.
5. Bruserud R, Ulvestad E (2000): Acute myelogenus leukemia blasts as accessory cells during in vitro T lymphocyte activation. Cell Immunol 206: 36-50.
6. Hirano N, Takahashi T, Ohtake S, et al. (1996): Expression of costimulatory molecules in human leukemias. Leukemia 10: 1168-1176.
7. Komatsu F, Kajiwara M (2000): CD18/CD54 (+CD102), CD2/CD58 pathway-independent killing of lymphokine-activated killer (LAK) cells against glioblastoma cell lines T98G and U373MG. Oncol Res 12: 17-24.
8. Liu SQ, Golan DE (1999): T-cell stimulation through the T-cell receptor/CD3 complex regulates CD2 lateral mobility by a calcium/calmodulin-dependent mechanism. Bioph J 76: 1679-92.
9. Maeda A, Yamamoto K, Yamashita K, et al. (1998): The expression of co-stimulatory molecules and their relationship to the prognosis of human acute myeloid leukemia: poor prognosis of B7-2-positive leukemia. Brit J Hematol 102: 1257-1262.
10. Miwa H, Mizutani M, Mahmud N, et al. (1998) Biphasic expression of CD4 in acute myelocytic leukemia (AML) cells: AML of monocyte origin and hemopoietic precursor cell origin. Leukemia 12: 44-51.
11. Reuss-Borst MA, Klein G, Walter HD, Müller CA (1996): Differential expression of adhesion molecules in human leukemias. Leukemia 9: 869-874.
12. Yang H, Reinherz EL (2001): Dynamic recruitment of human CD2 into lipid rafts. Linkage to T cell signal transduction. J Biol Chem 276: 18775-85.
Correspondence: Jolanta Woźniak, Department of Physiopathology, Institute of Haematology and Blood Transfusion, Chocimska 5, 00-957 Warsaw, Poland. Phone/fax number: +48 22 849 82 32. E-mail: jolwoz@poczta.onet.pl