ALBERT

Description: Key Points Multiple myeloma cells inhibit myeloma-specific T cells through expression of carcinoembryonic antigen-related cell adhesion molecule-6.

Description: The inhibition of BCR/ABL kinase activity by imatinib mesylate (IM, STI571, Gleevec®) is the standard therapy for patients with Philadelphia chromosome+ (Ph+) chronic myeloid leukemia (CML). However, the long term treatment with IM or other BCR/ABL kinase inhibitors may be limited due to the development of resistant disease and accumulating side effects. Immunotherapeutical approaches directed against Ph+ CML may overcome these problems. So far, the research to develop an immunotherapy against Ph+ leukemia focused on the BCR-ABL fusion protein, giving the promising opportunity to stimulate cytotoxic T lymphocytes against the joining region segment of p210 BCR/ABL. However, only a limited number of peptides spanning the fusion region is endogenously processed and presented in context with HLA class I molecules. In contrast, the constitutively active BCR/ABL kinase leads to the upregulation and activation of multiple proteins, which may subsequently result in the expression of different, probably leukemia-associated antigens on the cell surface of BCR/ABL-positive malignant cells. In this study, we investigated the immunogenicity of antigens upregulated by BCR-ABL kinase activity, and verified the capacity of these antigens to induce an anti-leukemia T cell response in vitro. We performed CD8+ T-cell stimulations with dendritic cells transfected with mRNA coding BCR/ABL wild-type (WT). Following two stimulations, the proliferating T-cell populations were analyzed for cytokine secretion (IFN-γ) in response to target cells expressing either BCR/ABL WT or a kinase deficient (KD, K1172R) variant of BCR/ABL. With this experimental setting it was possible to compare the immunogenic potential of antigens upregulated by BCR/ABL with the immunogenicity of the BCR/ABL protein itself. We were not able to activate T-cell populations directed against either the breakpoint-region of BCR/ABL or the BCR- or ABL part of the fusion protein. In contrast, we show here that the constitutively active kinase domain of BCR/ABL has a key role in enhancing the immunogenicity of BCR/ABL+ cells. A broad, HLA-dependent T-cell immune response specifically directed against BCR/ABL regulated antigens was detected. The inhibition of BCR/ABL kinase activity by IM significantly impaired the immunogenicity of BCR/ABL+ cells. We found the same T-cell reactivity pattern in several healthy donors and a CML patient. This is the first study demonstrating the major contribution of the BCR/ABL kinase domain to the immunogenicity of BCR/ABL+ cells as T-cell responses against these cells are dominated by BCR/ABL regulated antigens, and not by BCR/ABL itself. These results may contribute to the design of clinical vaccination trials for the treatment CML patients with minimal residual disease, e.g. following successful induction therapy with BCR/ABL inhibitors.

Description: The gene transfer of alpha and beta chains derived from a defined TCR has been successfully applied to endow T cells with specificities directed against tumor-associated antigens. However, it is still unclear if the transfer of TCR genes into T cells that already express an endogenous TCRalpha and beta chain leads to engineered T cells expressing four different TCR complexes on their cell surface. Mixed TCR heterodimers composed of endogenous and exogenous TCR chains may acquire new specificities, which may cause unwanted reactions in patients following adoptive T cell transfer. We examined the possibility of mixed TCR heterodimer formation using defined conditions of single TCR chain transfer into human cytotoxic T cell (CTL) clones specific for CMV and Melan-A, respectively. After stimulation for three days CTLs were retrovirally transduced with the beta chain derived from a gp100-specific TCR. The expression of the exogenous (transduced) and the endogenous beta chain was distinguished by flow cytometry using antibodies against the different Vbeta motives. Indeed, CTLs that had been transduced with the single beta chain expressed this chain on the cell surface indicating the formation of mixed TCRs, because the expression of the exogenous TCRbeta chain requires the pairing with the endogenous TCRalpha chain. Furthermore, we transduced the CTL clones with both the TCRalpha and beta chain derived from a gp100-specific TCR. The transduced T cells were positively stained with an A2/gp100 multimer documenting the correct formation of the exogenous TCR chains. Functionality of transduced CTL clones was tested by antigen-specific IFN-gamma release and cytolytic activity. The TCR-transduced T cells were sorted with the A2/gp100 multimer and expanded for two weeks. Double staining with HLA multimers for the endogenous and the transduced TCRs showed the downregulation of the endogenous TCRs in three different CTL clones. In one CTL clone, the endogenous TCR was even replaced by the exogenous TCR as documented by flowcytometry and antigen-specific T cell function. In conclusion, transfer of single TCR chains in CTL clones can result in the formation of TCR heterodimers. However, our results also show that complete TCRs are predominantly expressed or can even replace other TCRs following transfer. The development of dominant TCRs will facilitate the therapeutical approaches of adoptive transfer regimens based on TCR-transduced T cells, because dominant TCRs can be selected or TCRs can be modified to be more dominant.

Description: Abstract 2942 Tumor-specific cytotoxic T cells are common in tumor patients, but ineffectively react against autologous tumor cells. Here, we demonstrate in multiple myeloma and breast cancer that human tumor cells escape recognition by tumor-specific CD8+ T cells through carcinoembryonic antigen-related cell adhesion molecule-6 (CEACAM-6) expression. We demonstrate for the first time CEACAM-6 expression in primary and established myeloma and examined the effects of altered CEACAM expression on cytotoxic T cell activity and cytokine secretion against myeloma and breast cancer cells in vitro —, and in vivo, using a xenotransplant mouse model. Cytotoxic T cells from multiple myeloma patients reacted against myeloma antigens presented by dendritic cells, but not against autologous myeloma cells, which expressed CEACAM6. Gene knockdown or blocking of CEACAM6 on myeloma cells restored CD8+ T-cell reactivity against malignant plasma cells. SiRNA-mediated CEACAM6 knockdown or inhibition by specific mAbs also restored cytokine secretion, cytotoxic activity, and antigen-specific lysis of CEACAM6-positive breast cancer cells. Moreover, CEACAM-6 inhibition was a prerequisite for efficient treatment of xenotransplanted breast tumors by adoptive T cell transfer. CEACAM6 thus plays an important role in inhibiting CD8+ T-cell responses against hematological and epithelial human tumors. Therapeutic targeting of CEACAM6 may be a promising strategy for improving cancer immunotherapy. Disclosures: No relevant conflicts of interest to declare.

Description: Cancer testis (CT)-antigens belong to a class of tumor antigens that are aberrantly expressed in a variety of hematological malignancies including multiple myeloma. Owing to their restricted gene expression, CT-antigens represent potential target antigens for immunotherapeutical approaches such as vaccination and adoptive T cell transfer. As the CT-antigens are self antigens, the majority of CT-antigen-specific autologous T cells display a low avidity T cell receptor (TCR), which often results in a weak tumor recognition efficiency. Our group has been focusing on the isolation of highly avid T cells against CT-antigens that are expressed in multiple myeloma, in particular MAGE-C1, MAGE-C2, and NY-ESO-1. The experimental approach was based on the stimulation of allo-restricted cytotoxic T cells, because highly avid T cells recognizing peptide epitopes in context with foreign HLA-alleles are not depleted in the thymus. HLA-A2-negative T cells were stimulated with HLA-A2-positive allogeneic dendritic cells that had been exogenously loaded with HLA-A2-binding peptides derived from NY-ESO-1, MAGE-C1 or MAGE-C2. Using this technique we were able to isolate allo-HLA-A2-restricted cytotoxic T lymphocyte (CTL) clones with peptide-dominant binding against known and novel peptide epitopes derived from NY-ESO-1, MAGE-C1 and MAGE-C2. The expanded peptide-specific CTL clones lysed HLA-A2-positive myeloma cell lines expressing NY-ESO-1, MAGE-C1 and MAGE-C2, respectively. Of note, the MAGE-C1-specific T cells crossreacted with the corresponding MAGE-C2 peptide due to the existing sequence homology between MAGE-C1 and MAGE-C2. Current experiments focus on redirecting primary T cells toward myeloma cells by retroviral gene transfer of CT-antigen-specific TCRs. The establishment of a set of high avidity TCRs specific for CT-antigens facilitates the development of adoptive transfer regimens based on TCR-transduced T cells for the treatment of multiple myeloma.

Description: Abstract 412 Background: A large proportion of patients are currently not eligible for genotype-adapted strategies in acute myeloid leukemia (AML), in particular those lacking specific genetic aberrations such as PML-RARA, CBFB-MYH11, RUNX1-RUNX1T1, NPM1 or activating FLT3 mutations. This subgroup of patients accounts for about one-third of all AML patients and mainly includes the large group of AML with myelodysplasia-related changes, AML with recurrent cytogenetic abnormalities [inv(3) or t(3;3), t(9;11), t(v;11q23)] and cytogenetically normal AML (CN-AML) with wild-type NPM1 and FLT3. Prognosis in this subgroup of patients is generally poor. Azacitidine has been shown to be active in AML with low blast counts frequently observed in AML with myelodysplasia-related changes and in CN-AML in the absence of specific gene mutations. Aims: To evaluate clinical efficacy of azacitidine in combination with intensive induction chemotherapy and in maintenance for two years as single agent in patients with AML who are not candidates for genotype-adapted treatment approaches. Methods: Patients with AML in the absence of specific genetic aberrations (PML-RARA, CBFB-MYH11, RUNX1-RUNX1T1, NPM1 mutation, activating FLT3 mutations) who are fit for intensive chemotherapy were eligible. Patients were up-front randomized for induction therapy into one standard arm and three experimental arms; i) ICE (standard arm), idarubicin (12 mg/m2/day, iv, days 1,3,5), cytarabine (100 mg/m2/day, cont. infusion, days 1–7), etoposide (100 mg/m2/day, iv, days 1,2,3); ii) AZA-prior, azacitidine (100 mg/m2/day, sc, days 1–5), idarubicin (12 mg/m2/day, iv, days 6, 8, 10), etoposide 100 mg/m2/day, iv, days 6,7,8); iii) AZA-concurrent, azacitidine (100 mg/m2/day, sc, days 1–5), idarubicin (12 mg/m2/day, iv, days 1,3,5), etoposide 100 mg/m2/day, iv, days 1,2,3); iv) AZA-after, idarubicin (12 mg/m2/day, iv, days 1,3,5), etoposide 100 mg/m2/day, iv, days 1,2,3), azacitidine (100 mg/m2/day, sc, days 4–8). After two induction cycles for patients achieving complete remission (CR), consolidation therapy was prioritized; first priority) allogeneic hematopoietic blood stem cell transplantation (HSCT) from matched related as well as unrelated donors, second priority) 3 courses of high-dose cytarabine followed by two-year maintenance therapy with azacitidine as single agent (50 mg/m2/day, sc, days 1–5, every 4 weeks) in patients initially randomized to experimental treatment. The primary endpoint was achievement of CR. The statistical design of the study was based on the Simon's optimal two-stage design applied for each arm separately. The null hypothesis was CR-rate equal or below 0.40 whereas the alternative hypothesis was a CR rate of at least 0.55 with a power of 80% and a level of significance of 5%. Thus, in each arm at least 12 of 26 patients with response to induction therapy were necessary after the first to proceed to the second stage. Results: During the first stage of the study 104 patients were randomized; median age was 62.5 years (range 18–82), 46% were female. Data on cytogenetics showed intermediate risk karyotype in 67% (n=50) including CN-AML (n=31) and high-risk karyotype in 33% (n=25). The most frequent serious adverse events were grade 3/4 infection with an overall incidence of 25% and ranging from 20 to 34% in the different treatment arms. The number of responding patients in the treatment arms AZA-prior and AZA-concurrent after the first stage of the study were 11 of 26 (42%) and 10 of 26 (38%)Both arms, AZA-prior and were terminated accordingly. In contrast, the treatment arms ICE and AZA-after were carried forward to the second stage of patient recruitment since responding patients at that time were 14 of 26 (54%) in both arms. In total, 100 patients each have been enrolled in both treatment arms, ICE and AZA-after, with CR-rates of 59% and 52%, respectively (p=0.39). To date, 60 patients received an allogeneic HSCT (n=36 matched unrelated donors, n=23 matched related donors, n=1 haploidentical family donor). Maintenance treatment was started in 12 patients. Conclusion: Induction therapy with ICE or idarubicin, etoposide followed by azacitidine (AZA-after) appears equally effective in producing CR in patients with AML who are not candidates for genotype-adapted treatment approaches. An amendment perpetuating the treatment arms ICE and AZA-after within a phase-III concept is planned. Disclosures: Schlenk: Celgene: Research Funding. Off Label Use: Azacitidine combined with intensive chemotherapy.

Description: The Cancer Testis (CT) antigen NY-ESO-1 is one of the most immunogenic cancer antigens eliciting strong humoral and cellular immune responses in tumor patients and therefore it is a promising candidate antigen for successful adoptive T cell transfer. The aim of our studies is the transfer of autologous T cells re-directed towards CT antigens by T cell receptor (TCR) gene transfer. The first precondition for genetic transfer of CT-Ag-specific TCRs is the availability of tumor-reactive CD4+ and CD8+ T cell clones that express a CT-Ag-specific TCR. Therefore, we generated the autologous CD8+ T cell clone ThP2 through stimulating HLA-A2.1− PBMCs with autologous HLA-A2+DCs loaded with synthetic NY-ESO-1157–165. After two restimulations, FACS-sorting and cloning, the T cell line specifically recognized the NY-ESO-1157–165 peptide and also specifically lysed NY-ESO-1157–165 expressing tumor cells. In addition, we generated NY-ESO-1 specific T helper1 clones from HLA-DR1+ and HLA-DR4+ healthy donors by stimulation of CD4+ T cells with autologous dendritic cells (DC) pulsed with the NY-ESO-187–111 peptide. The specificity of CD4+ T helper cell clones was determined by proliferation assays and IFN gamma ELISPOT through screening with the NY-ESO-187–111 peptide. By limiting dilution of the NYESO- 1-specific T cell populations we succeeded to isolate CD4+ T cell clones, which recognized NY-ESO-1-pulsed target cells and DCs pulsed with NY-ESO-1 protein. The second precondition for TCR gene transfer is the availability of efficient vector systems. Using vectors based upon mouse myelo-proliferative sarcoma virus (MPSV), it was possible to achieve a high transgene expression in the TCR-transduced T cells. Therefore, we cloned the TCR of the HL-A2-restricted NY-ESO-1-specific CTL clone ThP2 in the retroviral vector and documented the correct expression of the TCR-chains using peptide/HLA-multimers following retroviral transduction of peripheral PBMCs. Moreover, the NY-ESO-1 specific lysis of HLA-A2+ NY-ESO-1+ tumor cell lines after transduction in primary T cells was as well effective as the primary T cell clone. Because the expression of naive transgenic T cell receptors in recipient human T cells is often insufficient to achieve highly reactive T cell bulks we modified the TCR of the ThP2 CTL clone by, murinisation, codon optimalization or by introducing cysteins into the constant regions. Afterwards we compared the expression efficiency of the three different modifications on naive T cells by tetramer-staining. We were able to show that codon optimalization leads to an increase in the expression levels of the transgenic TCRs in human CD8+ T cells. The next step is the development of T cell transfer regiments, which are based on class-II-restricted TCR-transduced T cells.

Description: In Ph+ chronic myeloid leukemia (CML), the constitutively active Bcr-Abl kinase leads to the up-regulation and activation of multiple genes, which may subsequently result in the expression of leukemia-associated antigens. In this study, we investigated the immunogenicity of Bcr-Abl–regulated antigens by stimulating CD8+ T lymphocytes with autologous dendritic cells transfected with RNA coding for Bcr-Abl wild-type or a kinase-deficient mutant. Significant HLA class I–restricted T-cell responses were detected against antigens regulated by the Bcr-Abl kinase, but not toward the Bcr-Abl protein itself. The T-cell repertoire of a patient with CML in major molecular remission due to imatinib mesylate was also dominated by T cells directed against Bcr-Abl–regulated antigens. These results encourage the development of immunotherapeutic approaches against Bcr-Abl–regulated antigens for the treatment of CML patients with residual disease following therapy with Bcr-Abl kinase inhibitors.