Conversion of AML-blasts to leukemia-derived dendritic cells (DCleu) in 'DC-culture-media' shifts correlations of released chemokines with antileukemic T-cell reactions.

Dendritic cells (DC) and T-cells are mediators of CTL-responses. Autologous (from patients with acute myeloid leukaemia (AML) or myelodysplasia (MDS)) or allogeneic (donor)-T-cells stimulated by DCleu, gain an efficient lysis of naive blasts, although not in every case. CXCL8, -9, -10, CCL2, -5 and Interleukin (IL-12) were quantified by Cytometric Bead Array (CBA) in supernatants from 5 DC-generating methods and correlated with AML-/MDS-patients' serum-values, DC-/T-cell-interactions/antileukemic T-cell-reactions after mixed lymphocyte culture (MLC) and patients' clinical course. The blast-lytic activity of T-cells stimulated with DC or mononuclear cells (MNC) was quantified in a cytotoxicity assay. Despite great variations of chemokine-levels, correlations with post-stimulation (after stimulating T-cells with DC in MLC) improved antileukemic T-cell activity were seen: higher released chemokine-values correlated with improved T-cells' antileukemic activity (compared to stimulation with blast-containing MNC) - whereas with respect to the corresponding serum values higher CXCL8-, -9-, and -10- but lower CCL5- and -2-release correlated with improved antileukemic activity of DC-stimulated (vs. blast-stimulated) T-cells. In DC-culture supernatants higher chemokine-values correlated with post-stimulation improved antileukemic T-cell reactivity, whereas higher serum-values of CXCL8, -9, and -10 but lower serum-values of CCL5 and -2 correlated with post-stimulation improved antileukemic T-cell-reactivity. In a context of 'DC'-stimulation (vs serum) this might point to a change of (CCL5 and -2-associated) functionality from a more 'inflammatory' or 'tumor-promoting' to a more 'antitumor'-reactive functionality. This knowledge could contribute to develop immune-modifying strategies that promote antileukemic (adaptive) immune-responses.

Expression profiles of HMGB1 on B-CLL related leukocytes contribute to prediction of relapse.

BACKGROUND: The High Mobility Group Box 1 (HMGB1) is a nuclear protein that is frequently overexpressed in hematologic diseases and might be of relevance in immunogenic cancer control thus correlating with patients' (pts.) prognosis in diseases such as acute myeloid, acute lymphatic and chronic lymphocytic leukemia. MATERIALS AND METHODS: Expression profiles of blasts from AML (n = 21), ALL (n = 16) and of B-lymphocytes of CLL (n = 9) pts. were analyzed for surface expression of HMGB1 using flow cytometry. Expression was quantified and correlated with clinically and prognostically relevant markers. RESULTS: Expression profiling of HMGB1 in blasts of AML and ALL subtypes did not show differences between primary vs. secondary disease development and gender related differences. In ALL pts. however, age groups at initial diagnosis between ≥20 vs. <20 years were compared and showed significant differences (≥20 vs. <20 years; 89% vs. 49%, p  <0.05) with higher expression in higher age. In AML and CLL these differences were not visible. To evaluate the prognostic significance of HMGB1 expression, expression quantity was correlated with established and prognostic classification systems (in AML ELN, in ALL GMALL) and probability to relapse. No significant correlation was seen in these entities. However, when AML pts. were analyzed for remission rates after first anthracycline based induction therapy, in those who did not experience a complete remission significantly enhanced HMGB1 surface expression was seen (98 vs. 94%; p < 0.05; n = 20). Furthermore, for CLL it was shown that higher HMGB1 expression was found in pretreated patients with relapsed or/and refractory disease (1 vs. more relapses; 94 vs. 98%; p  <0.05; n = 9). CONCLUSION: HMGB1 is frequently expressed in hematologic malignancies. In this study it was shown that HMGB1 surface expression on AML blasts can be used as predictors for treatment response. In CLL it may be a marker for advanced disease. In order to implement this marker in FACS routine it could be a useful and practical tool for prognostic assessment and treatment planning.

Serum Chemokine-release Profiles in AML-patients Might Contribute to Predict the Clinical Course of the Disease.

In cancer or hematologic disorders, chemokines act as growth- or survival factors, regulating hematopoiesis and angiogenesis, determining metastatic spread and controlling leukocyte infiltration into tumors to inhibit antitumor immune responses. The aim was to quantify the release of CXCL8, -9, -10, CCL2, -5, and IL-12 in AML/MDS-pts' serum by cytometric bead array and to correlate data with clinical subtypes and courses. Minimal differences in serum-levels subdivided into various groups (e.g. age groups, FAB-types, blast-proportions, cytogenetic-risk-groups) were seen, but higher release of CXCL8, -9, -10 and lower release of CCL2 and -5 tendentially correlated with more favorable subtypes (<50 years of age, <80% blasts in PB). Comparing different stages of the disease higher CCL5-release in persisting disease and a significantly higher CCL2-release at relapse were found compared to first diagnosis - pointing to a change of 'disease activity' on a chemokine level. Correlations with later on achieved response to immunotherapy and occurrence of GVHD were seen: Higher values of CXCL8, -9, -10 and CCL2 and lower CCL5-values correlated with achieved response to immunotherapy. Predictive cut-off-values were evaluated separating the groups in 'responders' and 'non-responders'. Higher levels of CCL2 and -5 but lower levels of CXCL8, -9, -10 correlated with occurrence of GVHD. We conclude, that in AML-pts' serum higher values of CXCL8, -9, -10 and lower values of CCL5 and in part of CCL2 correlate with more favorable subtypes and improved antitumor'-reactive function. This knowledge can contribute to develop immune-modifying strategies that promote antileukemic adaptive immune responses.

Minor histocompatibility antigen UTY as target for graft-versus-leukemia and graft-versus-haematopoiesis in the canine model.

Male patients with female-stem-cell donors have better prognosis compared to female-to-male combinations due to Y-encoded minor histocompatibility antigens recognized by female-alloimmune-effector lymphocytes in the context of a graft-versus-leukemia (GvL) effect. We provide data in a dog-model that the minor histocompatibility antigen UTY might be a promising target to further improve GvL-immune reactions after allogeneic-stem-cell transplantations. Female-canine-UTY-specific T cells (CTLs) were stimulated in vitro using autologous-DCs loaded with three HLA-A2-restricted-UTY-derived peptides (3-fold-expansion), and specific T cell responses were determined in 3/6 female dogs. CTLs specifically recognized/lysed autologous-female-peptide-loaded DCs, but not naïve-autologous-female DCs and monocytes. They mainly recognized bone-marrow (BM) and to a lower extent DCs, monocytes, PBMCs and B-cells from DLA-identical-male littermates and peptide-loaded T2-cells in an MHC-I-restricted manner. A UTY-/male-specific reactivity was also obtained in vivo after stimulation of a female dog with DLA-identical-male PBMCs. In summary, we demonstrated natural UTY processing and presentation in dogs. We showed that female-dog CTLs were specifically stimulated by HLA-A2-restricted-UTY peptides, thereby enabling recognition of DLA-identical-male cells, mainly BM cells. These observations suggest UTY as a promising candidate-antigen to improve GvL-reactions in the course of immunotherapy.

Serum levels of sCD137 (4-1BB) ligand are prognostic factors for progression in acute myeloid leukemia but not in non-Hodgkin's lymphoma.

CD178 (Fas/APO-1 ligand) and CD137 ligand (CD137L) have previously been described in sera of patients with various malignancies and play an important role in the pathogenesis of various diseases. Recently, we demonstrated that low levels of soluble (s) CD137L and high levels of sCD178 correlate significantly with a long progression free survival in patients with myelodysplastic syndrome (MDS). In this study, we correlated sCD137L and sCD178 levels in sera of 42 samples of patients with acute myeloid leukemia (AML) and 46 samples of patients with non-Hodgkin's lymphoma (NHL) with stages, subtypes, and the clinical course of the diseases and determined cut-off values with maximum probability for significant differentiation between cases with higher/lower probability for progress free survival. In contrast to patients with MDS, surprisingly no correlation between sCD178 levels and different subtypes and stages or with prognosis in AML or NHL were observed. Regarding sCD137L, NHL-patients displayed lower levels compared with AML. Statistically significant higher median levels of sCD137L are present in patients with undifferentiated AML (M1/M2, 1,470 pg/mL), poor cytogenetic risk (288 pg/mL) and higher levels of BM-blasts (186 pg/mL) compared with patients with monocytoid AML (M4/M5, 89 pg/mL), intermediate cytogenetic risk (59 pg/mL) and lower levels of BM-blasts (14 pg/mL) respectively. Furthermore, in AML patients sCD137L levels correlate significantly with the probabilities to achieve complete remission (CR), stay in CR or with progress of the disease. Taken together, our data demonstrate that sCD137L can be used as a prognostic factor not only in MDS but also in AML.

Expression of poliovirus receptor-related proteins PRR1 and PRR2 in acute myeloid leukemia: first report of surface marker analysis, contribution to diagnosis, prognosis and implications for future therapeutical strategies.

Poliovirus receptor-related (PRR) proteins belong to the Nectin-adhesion molecules' group, are expressed on endothelial cells and on CD34(+) stem cells and mediate the organization of endothelial and epithelial junctions. There is evidence to suggest, that those receptors could have a role in leukemia. We have studied the expression of PRR molecules PRR1 and PRR2 on mononuclear bone marrow (BM) cells of 55 patients with acute myeloid leukemia (AML) at first diagnosis by FACS-analysis using directly Phycoerythrin-labeled markers (PRR1 clone R1.302.12; PRR2 clone R2.477.1) in combination with other Fluorescein conjugated antibodies to evaluate the blast phenotype in AML. The leukemic gate included blasts and residual monocytes and lymphocytes. A case was defined as positive, if more than 20% of the gated cells expressed the regarding receptor. We could demonstrate, that on average 35% PRR1(+) or 45% PRR2(+) cells in AML were found. Within FAB-types we observed a high PRR1 expression in cases with M3 and M4 and lowest expressions in M0 and M5; a high PRR2 expression was found in cases with M3, M4, M5 and M1 and lowest expressions in M0 and M2. Separating our patients' cohorts in cytogenetic risk groups we could detect a significant higher proportion of PRR1(+) cases (73% vs. 25% of cases, P = 0.009) or PRR1(+) cells (57% vs. 18% of cases, P = 0.001) in the cytogenetic favorable risk vs. poor risk group (75% vs. 32% PRR2(+) cases). Moreover cut-off-values with a maximum probability for a significant differentiation between cases with higher or lower levels of these markers could be found: cases with >78% PRR1(+) and cases with >77% PRR2(+) cells were characterized by a tendency for longer relapse free survival times. Qui-square analyses showed, that 3 of 4 cases with FAB-type M3 (P = 0.03) or a favorable karyotype (P = 0.04) were found in the group with >7% PRR1(+) cells, due to only few cases available a similar correlation, however, could not be found in cases with >78% PRR2(+) cells. We can conclude, that blasts in AML regularly express PRR1 and PRR2. Cases with a high expression of PRR1 or PRR2 are characterized by a more favorable prognosis. With respect to the individual PRR-status the benefit of biological response modifiers as priming agents, differentiation mediators or factors influencing cellular metabolisms inducing factors can be discussed under a new point of view.

Leukaemia-derived dendritic cells can be generated from blood or bone marrow cells from patients with myelodysplasia: a methodological approach under serum-free culture conditions.

Functional dendritic cells (DC) are professional antigen-presenting cells (APC) and can be generated in vitro from healthy as well as from leukaemic cells from AML patients giving rise to APC of leukaemic origin presenting leukaemic antigens. In a comparative methodological analysis of 50 AML samples, we could already show that leukaemia-derived DC can regularly be generated under serum-free culture conditions. In this study, we describe the generation and characterization of DC from different mononuclear cell (MNC) fractions from 24 myelodysplastic syndrome (MDS) patients under those different serum-free culture conditions, determine the optimal culture conditions and compare the results with that from 23 healthy donors. In parallel cultures, we compared DC harvests after 7- or 14-day culture, with total or adherent MNC or T-cell-depleted MNC or PB or BM-MNC, thawn or fresh MNC, in Xvivo or CellGro serum-free media, +/-10% autologous plasma or +/-FL. In detail, we could show that MDS-DC harvests compared to healthy DC were higher after 10- to 14-day culture; total or adherent PB or BM-MNC fractions yield comparable DC counts; however, from MACS-depleted MNC fractions or thawn MNC lower DC counts can be generated. Whereas the addition of FL increases the DC harvest, the addition of autologous plasma in many cases has inhibitory influence on DC maturation, CellGro and Xvivo media yield comparable DC counts. Optimal harvest of vital and mature DC from MDS samples was obtained with a GM-CSF, IL-4, FL and TNF-alpha containing serum-free Xvivo medium after 10-14 days of culture (18/26% DC; 54/64% vital DC; 59/51% mature DC were generated from MDS/healthy MNC samples). Surface marker profiles (e.g. costimulatory antigen expression) of DC obtained from MDS samples were comparable with that of healthy DC. The leukaemic derivation of MDS-DC was demonstrated by the persistence of the clonal cytogenetic aberration in the DC or by coexpression of leukaemic antigens on DC. Autologous T-cell activation of leukaemia-derived DC was demonstrated in cases with MDS. Autologous T cells proliferate and upregulate DC-contact-relevant antigens. We are the first who demonstrate that the generation of leukaemia-derived DC is feasible not only in AML but also in MDS under serum-free culture conditions giving rise to DC with comparable characteristics as healthy DC and offering an antileukaemia-directed immunotherapeutical vaccination strategy in AML and MDS.

Leukemia-derived dendritic cells can be generated from blood or bone marrow cells from patients with acute myeloid leukaemia: a methodological approach under serum-free culture conditions.

Functional dendritic cells (DC) are professional antigen-presenting cells (APC) and can be generated in vitro from healthy as well as from leukaemic cells from acute myeloid leukemia (AML) patients giving rise to APC of leukaemic origin-presenting leukaemic antigens. We describe the generation and characterization of DC from different mononuclear cell (MNC) fractions from 50 AML patients under different serum-free culture conditions, determine the optimal culture conditions and compare the results with that from 23 healthy donors. In parallel cultures, we compared DC harvests after 7- or 14-day culture, with total or adherent MNC or T-cell depleted MNC or peripheral blood (PB) or bone marrow-MNC (BM-MNC), thawn or fresh MNC, in Xvivo or CellGro serum-free media, +/-10% autologous plasma or +/-FL. In detail, we could show that AML-DC harvests were higher after 10-14 days culture (healthy DC: 7 days); total or adherent PB or BM-MNC fractions yield comparable DC counts, however, from magnetic cell sorting (MACS)-depleted MNC fractions or thawn MNC lower DC counts can be generated. Whereas the addition of FL increases the DC harvest, the addition of autologous plasma in many cases has inhibitory influence on DC maturation. CellGro and Xvivo media yield comparable DC counts. Optimal harvest of vital and mature DC from AML samples was obtained with a granulocyte/macrophage-colony stimulating factor, interleukin-4, FL and tumour necrosis factor-alpha-containing serum-free Xvivo medium after 10-14 days of culture (36/26% DC; 38/64% vital DC; 46/51% mature DC were generated from AML/healthy MNC samples). Surface marker profiles (e.g. costimulatory antigen expressing) of DC obtained from AML samples were comparable with that of healthy DC. The leukaemic derivation of AML-DC was demonstrated by the persistence of the clonal cytogenetic aberration in the DC or by coexpression of leukaemic antigens on DC. Autologous T-cell activation of leukaemia-derived DC was demonstrated in cases with AML. Autologous T cells proliferate and upregulate DC-contact-relevant antigens. We demonstrate that the generation of leukaemia-derived DC is feasable in AML under serum-free culture conditions giving rise to DC with comparable characteristics as healthy DC and offering an anti-leukaemia-directed immunotherapeutical vaccination strategy in AML.

Serum-free generation and quantification of functionally active Leukemia-derived DC is possible from malignant blasts in acute myeloid leukemia and myelodysplastic syndromes.

Functional dendritic cells (DC) are professional antigen presenting cells (APC) and can be generated in vitro from leukemic cells from acute myeloid leukemia AML patients, giving rise to APC of leukemic origin presenting leukemic antigens (DC(leu)). We have already shown that DC can be successfully generated from AML and myeloplastic syndromes (MDS) cells in serum-free 'standard' medium (X-vivo + GM-CSF + IL-4 +TNFalpha + FL) in 10-14 days. In this study, we present that DC counts generated from mononuclear cells (MNC) varied between 20% (from 55 MDS samples), 34% (from 100 AML samples) and 25% (from 38 healthy MNC samples) medium. Between 53% and 58% of DC are mature CD83+ DC. DC harvests were highest in monocytoid FAB types (AML-M4/M5, MDS-CMML) and independent from cytogenetic risk groups, demonstrating that DC-based strategies can be applied for patients with all cytogenetic risk groups. Proof of the clonal derivation of DC generated was obtained in five AML and four MDS cases with a combined FISH/immunophenotype analysis (FISH-IPA): The clonal numerical chromosome aberrations of the diseases were regularly codetectable with DC markers; however, not with all clonal cells being convertible to leukemia-derived DC(leu) (on average, 53% of blasts in AML or MDS). To the contrary, not all DC generated carried the clonal aberration (on average, 51% of DC). In 41 AML and 13 MDS cases with a suitable antigen expression, we could confirm FISH-IPA data by Flow cytometry: although DC(leu) are regularly detectable, on average only 57% of blasts in AML and 64% of blasts in MDS were converted to DC(leu). After coculture with DC in mixed lymphocyte reactions (MLR), autologous T cells from AML and MDS patients proliferate and upregulate costimulatory receptors. The specific lysis of leukemic cells by autologous T cells could be demonstrated in three cases with AML in a Fluorolysis assay. In six cases with only few DC(leu) or few vital T cells available after the DC/MLR procedure, no lysis of allogeneic or autologous leukemic cells was seen, pointing to the crucial role of both partners in the lysis process. We conclude: (1) the generation of DC is regularly possible in AML and also in MDS under serum-free conditions. (2) Clonal/leukemia-derived DC(leu) can be regularly generated from MDS and AML-MNC; however, not with all blasts being converted to DC(leu) and not all DC generated carrying leukemic markers. We recommend to select DC(leu) for vaccinations or ex vivo T-cell activations to avoid contaminations with non-converted blasts and non-leukemia-derived DC and to improve the harvest of specific, anti-leukemic T cells. DC and DC-primed T cells could provide a practical strategy for the immunotherapy of AML and MDS.

High expression of costimulatory molecules correlates with low relapse-free survival probability in acute myeloid leukemia (AML).

Costimulatory molecules such as lymphocyte function-associated antigen (LFA)-1 (CD11a), LFA-3 (CD58), intercellular adhesion molecule (ICAM)-1 (CD54), neuronal cell adhesion molecule (NCAM) (CD56), B7-1 (CD80), or B7-2 (CD86) are important regulatory elements in healthy immunological cascades, but their role in acute myeloid leukemia (AML) has only been rarely investigated. We studied their expression on mononuclear bone marrow (BM) cells from 105 patients with AML at initial diagnosis and evaluated their prognostic significance. Fluorescence-activated cell sorter (FACS) analyses were performed using antibodies directly conjugated with fluorescein. A BM sample was considered positive if more than 20% of the cells in the blast containing gate expressed the respective marker. The surface expression of CD11a (27 of 29 cases positive with an average of 71% positive blasts; 27(+)/29, 71%), CD54 (23(+)/33, 37%), CD56 (24(+)/93, 20%), CD58 (29(+)/29, 95%), CD80 (13(+)/28, 30%), and CD86 (19(+)/29, 39%) was measured. The expression of these markers in different French-American-British (FAB) classification types (M0-M5) was heterogeneous, except for CD56, which showed a higher proportion of positive cells in monocytic subtypes of AML. In addition, cases with a "poor risk" karyotype as well as patients succumbing to "early death" after double induction therapy according to the AML Cooperative Group (CG) protocol were characterized by a high expression of CD56. Relapse-free survival analyses demonstrated that patients with more than 8% CD56(+) cells in the BM relapsed significantly sooner. CD54 was preferentially expressed in AML M4(eo) and in addition in "favorable" cytogenetic risk groups and in cases that had responded to AML-CG therapy. Only very high proportions (>60%) of CD54(+) cells were associated with a lower probability for relapse-free survival. CD80 and CD86 expressions were similar in all FAB types. Patients who had responded to AML-CG therapy showed higher CD80 proportions and lower CD86 proportions compared to the "nonresponder" group. Whereas cases with more than 15% CD80(+) cells had a significantly lower probability for relapse-free survival, only cases with more than 65% CD86(+) were characterized by a significantly lower probability for relapse-free survival. Expression profiles of CD11a and CD58 were not associated with specific FAB types or prognostically relevant groups. We can conclude: (1) Expression of costimulatory molecules in AML is very variable. This reflects the great diversity of immunophenotypes in AML. (2) CD56 is mainly expressed in monocytic subtypes of AML. CD56(+) subtypes of AML seem to be a separate entity with a worse prognosis independent of the karyotype. (3) High expression of some costimulatory molecules correlates with a worse prognosis concerning relapse-free survival times.

Low L-selectin (CD62L) expression in acute myeloid leukemia correlates with a bad cytogenetic risk.

OBJECTIVES: Interactions between hemopoietic cells and the stromal microenvironment or immunoreactive cells are mediated by specific cell surface receptors. The expression of those molecules may alter the adhesive qualities (mobility and homing) as well as immune response behavior of leukemic blasts. L-Selectin (CD62L) is suggested to play a role in the redistribution and homing of hemopoietic progenitor cells to the bone marrow (BM). Down-regulation of L-selectin is responsible for mobilization of blasts from the BM into the circulation and ligation of L-selectin stimulates proliferation of progenitor cells. This could have an influence on the process of leukemia. METHOD: We have studied the expression of L-selectin on mononuclear BM cells of 36 acute myeloid leukemia (AML) patients at first diagnosis by FACS analysis using a directly fluorescein isothiocyanate conjugated antibody (clone DRE G56). RESULTS: On average the patients presented with 88% blasts in the BM. The expression tended to be higher in primary (p) AML compared with secondary (s) AML. L-Selectin was very heterogenously expressed in all FAB groups. Highest expression was found in cases with AML-M4 with four of nine cases presenting with an inv(16) karyotype. Separating our patient cohort in cytogenetic risk groups we could detect a significantly higher expression of L-selectin in cases with a 'good risk' karyotype and a very low expression in cases with a 'bad risk' karyotype (P = 0.037). Comparing patients who achieved remission after double induction therapy (responders) with patients who showed persisting disease (non-responders) we found a higher percentage of L-selectin+ cases or cells in the responder group than in the non-responder group, although the differences were not significant because of only five cases in the 'non-responder' group. Evaluating cut-off points greatest differences in relapse-free survival probabilities were found in patients who presented with > or = 30% L-selectin+ BM cells compared with cases with < 30%: 86% of cases with > or = 30% L-selectin+ cells were still in remission after a mean follow up time of only 8 months compared with only 46% in the group with < 30% L-selectin+ cells. CONCLUSIONS: We can conclude that (i) expression of L-selectin on AML blasts is variable. This reveals the great diversitiy of immunophenotypes in AML and might contribute to identify individual blast phenotypes in order to detect minimal residual disease in remission. (ii) Low L-selectin expression correlates with a bad cytogenetic risk, with a lower probability to achieve remission and with a shorter relapse-free survival time. This might reflect a decreased homing of the blasts to the BM as well as an impaired cytotoxic T-cell reaction against leukemic cells. The expression of L-selectin on leukemic blasts might be influenced by different cytokine therapies (e.g. with interferon alpha) and this might result in an altered hematologic reconstitution after cytotoxic therapies as well as in an altered immunologic recognition of blasts.

Soluble CD137 (4-1BB) ligand is released following leukocyte activation and is found in sera of patients with hematological malignancies.

Expression of CD137 ligand (4-1BBL), a member of the TNF family of proteins, has been reported on several types of APCs, various carcinoma cells, and can be induced on activated T cells. In this study, we report that the soluble ligand was released constitutively at low levels from leukocytes and at higher levels following cellular activation. Release from cells was blocked by addition of a metalloproteinase inhibitor which concomitantly caused the accumulation of 4-1BBL on the cell surface. In addition, we show that a soluble form of 4-1BBL was present at high levels in the sera of some patients with various hematological diseases, but only at low levels in healthy donors. Soluble 4-1BBL was active in that it competed with recombinant 4-1BBL for binding to the 4-1BB receptor and was able to costimulate IL-2 and IFN-gamma release from peripheral T cells. These results indicate that the release of soluble 4-1BBL from the cell surface is mediated by one or more sheddases and likely regulates 4-1BB-4-1BBL interactions between cells in vivo. Cleavage of 4-1BBL to an active soluble form would alter both proximal and distal cellular responses, including cell survival and costimulatory or inflammatory responses, that are mediated through the 4-1BB pathway. This, in turn, would likely alter disease progression or outcome.

Influence of cytokines and autologous lymphokine-activated killer cells on leukemic bone marrow cells and colonies in AML.

We have already shown that cytokine cocktails (IL-1beta, IL-3, IL-6, SCF, GM-CSF) and/or lymphokine-activated killer (LAK) cells can reduce the amounts of clonal, CD34-positive mononuclear bone marrow cells (BM-MNC) in acute myeloid leukemia (AML). In addition, the influence of those cocktails and/or LAK cells on the clonogenic potential of AML BM-MNC was investigated. BM colonies cultured in agar during different stages of the disease were immunophenotyped in situ: 17 patients at diagnosis, 14 patients in complete remission, 8 patients at relapse, 8 healthy donors. A significant reduction in leukemic cells and colonies positive for CD34 after in vitro culture of BM-MNC with cytokine cocktails was achieved with all samples obtained at diagnosis (n = 8, p < 0.01), in 6 of 8 cases in complete remission but only in 2 of 6 cases at relapse. Cytokine cocktails stimulated granulopoiesis as well as B and T lymphopoiesis. Colonies with leukemic phenotype could never be detected in healthy BM. A significant reduction in leukemic colonies was achieved by coculture of BM-MNC (uncultured or cytokine precultured) with autologous LAK cells in all 4 cases at diagnosis and in 1 case at relapse. An additive effect of in vitro cytokine preincubation of BM samples on the leukemia-reducing effect of LAK cells could be demonstrated in all samples studied (p < 0.001; diagnosis: n = 10, relapse: n = 3, complete remission: n = 7). Patients had a better prognosis if CD34-positive colonies in AML could be reduced by cytokine incubation (p = 0.03) or coculture with autologous LAK cells in vitro (p = 0.04). Our data show that cytokines as well as LAK cells alone and in combination can reduce, however not eliminate clonogenic AML cells. Such mechanisms might be responsible for maintaining stable remissions in AML.

Not the presence but the amount of clonal DNA detectable in remission of acute myeloid leukemia is predictive for relapse.

OBJECTIVES: The persistence of clonal cells after chemotherapy, or a re-emerging of clonal cells in remission (CR) or at relapse in patients with acute myeloid leukemia (AML) was studied to assess the prognostic significance of the amount of clonal DNA in predicting the clinical outcome. METHODS: Clonal rearrangements in the gene sequences of retinoic acid receptor (RAR) alpha, major breakpoint cluster region (M-bcr), immunoglobulin (Ig)-JH, T-cell receptor (TcR) beta, myeloid lymphoid leukemia or cytokines (GM-CSF, G-CSF, IL-3) detected in bone marrow samples from 37 patients with primary AML (pAML) or secondary AML (sAML) were investigated. A relative increase or decrease of clonal DNA in the course of AML was evaluated by comparing the optical densities of DNA bands of the rearranged genes and the total amount of DNA. RESULTS: High amounts of clonal DNA were detectable at diagnosis, during persisting disease and at relapse (Ø 39%, 35%, or 38% of total DNA, respectively), compared to 20% in complete remission (CR). Amounts of clonal DNA (except for Ig-JH gene rearrangements) were of prognostic significance at diagnosis, patients with less than 33% clonal DNA were characterized by significantly longer relapse-free survival times (all cases: p = 0.01; pAML: p = 0.002). Patients in CR exhibiting less than 5% (all cases) or 15% (pAML) clonal DNA showed longer relapse-free survival times (p = 0.08 or p = 0.03, respectively). Vice versa, significantly higher amounts of clonal DNA (all cases 51% vs. pAML 54%) could be detected in cases studied at diagnosis who relapsed in the following 5 months (all cases p = 0.01) or 14 months (pAML p = 0.007). Significantly higher amounts of clonal DNA (33%) could be detected in cases studied in CR who relapsed in the following 4 months (all cases p = 0.002 or pAML p = 0.006, respectively). Moreover, we could prove disease progression on a cellular level months before the clinical onset of sAML after a period of MDS. CONCLUSIONS: Clonal, gene-rearranged DNA is regularly detectable at diagnosis and during persisting AML, in CR and at relapse. However, the presence, rather than the amount of clonal DNA detectable in CR is predictive for relapse. These data might indicate the significance of gene rearrangement analyses in the course of AML to identify cases with a high risk of relapse, independently from the karyotype.

Adoptive transfer of human natural killer cells in mice with severe combined immunodeficiency inhibits growth of Hsp70-expressing tumors.

In vitro, tumor-selective Hsp70 plasma membrane localization correlates with increased sensitivity to lysis mediated by a subpopulation of human natural killer (NK) cells that adhere to plastic following cytokine stimulation. In the present study, we analyzed the capacity of adoptively transferred human NK cells in SCID/beige mice for local tumor control and prevention of metastatic dissemination of Hsp70-expressing CX(+) and non-expressing CX(-) tumors following orthotopic (o.t.) injection. Both tumor sublines were derived by cell sorting of the original cell line, CX2, and thus exhibit an identical MHC and adhesion molecule expression pattern but differ with respect to Hsp70 plasma membrane expression. Viability of adherent, human NK cells in SCID/beige mice up to 18 days and the capacity to migrate have been demonstrated. Growth of Hsp70-expressing and non-expressing CX(+) and CX(-) tumor cells was completely suppressed when 10 x 10(6) NK cells were injected into the i.p. cavity on day 4 after inoculation of 2.5 x 10(6) tumor cells. Although a single injection of 5 or 2.5 x 10(6) NK cells was not sufficient to suppress tumor growth completely in all mice, the reduction in size of CX(+) tumors was significantly greater than that of CX(-) tumors. To mimic the clinical situation, ex vivo stimulated NK cells were injected i.v. on day 4 after o.t. injection of tumor cells. Under these conditions, growth of Hsp70-expressing primary tumors and metastases was suppressed. If CX(-) tumor cells were injected, 3 of 9 mice developed Hsp70-negative primary tumors. However, none of these mice developed distant metastases. In summary, our data indicate that Hsp70 acts as a recognition structure for adherent NK cells in a SCID/beige mouse model.

Cytokines can reduce clonal, CD34-positive cells in acute myeloid leukemia in vitro.

We studied the influence of cytokine mixes on the survival of acute myeloid leukemia (AML) bone-marrow (BM) cells in a 14-day culture assay in vitro. Southern-blot analysis using a panel of different probes in combination with densitometry and flow cytometry were used to detect and compare the amount of clonal or CD34-positive BM cells before and after the culturing procedure. A significant reduction of CD34-positive cells after incubation with a cytokine mix [interleukin (IL)-1beta, IL-3, IL-6, stem cell factor (SCF), erythropoietin (EP) with granulocyte macrophage/colony-stimulating factor (GM-CSF, Cytok1) could be achieved in all 16 cases with a CD34-positive blast phenotype studied at diagnosis (P<0.001), in 3 of 10 cases at relapse, and in 8 of 18 cases in complete remission. In healthy donors, an increase of CD34-positive cells was demonstrated in 5 of 5 samples. A reduction of clonal DNA through incubation with Cytok1 was achieved in 5 of 5 (100%) cases studied at diagnosis, in 1 of 4 (25%) cases at relapse, and in 7 of 9 cases (78%) in complete remission. Cytokine cocktails with GM-CSF (Cytok1) were more efficient in reducing (clonal) CD34-positive cells than cocktails without GM-CSF (Cytok2). AML patients at diagnosis and in complete remission had a better survival probability if their CD34-positive or clonal cells could be reduced in vitro by cytokine cultivation (P<0.05). Vitality of BM cells was not influenced by 14-day cytokine treatment; however, the total cell count could be increased by Cytok1 and Cytok2 by 55-174%, but not by the control medium. Our data show that: (1) clonal cell populations can be regularly detected at diagnosis, during complete remission, and at relapse; (2) CD34-positive cells in AML can be demonstrated to be clonal, gene-rearranged cells; (3) incubation of AML BM-cells with Cytokl leads to a reduction of the CD34-positive, clonal cell load in all cases at diagnosis and in 78% of the cases in complete remission of AML, but in only 25% of the cases at relapse; (4) in all healthy BM samples, proportions of 'healthy' CD34-positive cells were increased. Moreover, absolute cell counts were increased by cytokine incubation of cells obtained at diagnosis, relapse, or complete remission of AML and from healthy donors indicating a selective stimulation of healthy, but not of leukemic CD34-positive cells; (5) cytokine cocktails containing GM-CSF are more efficient in reducing leukemic cells than cocktails without GM-CSF; and (6) in vitro reactivity of clonal or CD34-positive BM cells against Cytokl has clinical relevance. We conclude, that Southern-blot analysis and flow cytometry are suitable methods to detect and quantify leukemic disease and to distinguish between clonal or non-clonal CD34-positive cells. The ex vivo or clinical application of specific combinations of cytokines might be a feasible and successful application of immunotherapy in AML that merits further investigations.

Gene rearrangements in bone marrow cells of patients with acute myelogenous leukemia.

At diagnosis, clonal gene rearrangement probes [retinoic acid receptor (RAR)-alpha, major breakpoint cluster region (M-bcr), immunoglobulin (Ig)-JH, T cell receptor (TcR)-beta, myeloid lymphoid leukemia (MLL) or cytokine genes (GM-CSF, G-CSF, IL-3)] were detected in bone marrow samples from 71 of 153 patients with acute myelogenous leukemia (AML) (46%): in 41 patients with primary AML (pAML) (58%) and in 30 patients with secondary AML (42%). In all cases with promyelocytic leukemia (AML-M3) RAR-alpha gene rearrangements were detected (n = 9). Gene rearrangements in the Ig-JH or the TcR-beta or GM-CSF or IL-3 or MLL gene were detected in 12, 10, 16 and 12% of the cases, respectively, whereas only few cases showed gene rearrangements in the M-bcr (6%) or G-CSF gene (3%). Survival of pAML patients with TcR-beta gene rearrangements was longer and survival of pAML patients with IL-3 or GM-CSF gene rearrangement was shorter than in patients without those rearrangements. No worse survival outcome was seen in patients with rearrangements in the MLL, Ig-JH or M-bcr gene. In remission of AML (CR), clonal gene rearrangements were detected in 23 of 48 cases (48%) if samples were taken once in CR, in 23 of 26 cases (88%) if samples were taken twice in CR and in 23 of 23 cases (100%) if samples were studied three times in CR. All cases with gene rearrangements at diagnosis showed the same kind of rearrangement at relapse of the disease (n = 12). Our data show that (1) populations with clonal gene rearrangements can be regularly detected at diagnosis, in CR and at relapse of AML. (2) Certain gene rearrangements that are detectable at diagnosis have a prognostic significance for the patients' outcome. Our results point out the significance of gene rearrangement analyses at diagnosis of AML in order to identify 'poor risk' patients - independently of the karyotype. Moreover, the persistence of clonal cells in the further course of AML can be studied by gene rearrangement analysis.