ALBERT

Description: Abstract 578 Autologous stem cell transplantation (ASCT) for multiple myeloma (MM) offers a unique setting to explore the role of immunotherapeutic strategies in eradicating residual disease. A fundamental challenge to developing an effective anti-tumor immune response is overcoming the immunosuppressive milieu by which tumor cells evade host immunity. Key elements contributing to tumor-mediated immune suppression are the increased presence of regulatory T cells in patients with malignancy, and upregulation of the PD-1/PDL1 pathway. Tumor expression of PD-L1 promotes T cell tolerance by binding PD-1 on activated T cells and suppressing their capacity to secrete stimulatory cytokines. In addition, the PD-1/PDL-1 pathway has been shown to inhibit T cell-mediated lysis of tumor cells, potentially preventing a clinically meaningful immunologic response to tumor vaccines. We are conducting a clinical trial in which patients with MM are treated with an anti-PD1 antibody (CT-011) alone (Cohort 1) and in combination with a dendritic cell/myeloma fusion cell vaccine (Cohort 2) following ASCT. To date, 27 patients have been enrolled into Cohort 1, in which patients receive three infusions of CT-011 at doses of 3mg/kg given at 6 week intervals beginning 1–3 months following ASCT. Mean age of the patients is 57 years; 61% are male. 11 patients have received at least two infusions of CT-011. The remaining patients are undergoing pre-transplant therapy/transplant. CT-011 has been well tolerated, with possibly related adverse events consisting of transient grade 1–2 leukopenia, diarrhea, fatigue, arthralgia, rash, and peri-orbital edema. One patient developed grade 3 neutropenia, which resolved after two days without growth factor. Immunologic response was determined by quantifying circulating tumor reactive T cells prior to each dose of CT-011 and at 1, 3, 6 months following the last infusion, as defined by the percentage of T cells expressing IFNg in response to ex vivo exposure to autologous tumor lysate. 4 patients have completed 6 months of follow up after the third dose of CT-011, and are evaluable for immune response. CT-011 therapy was associated with the dramatic expansion of myeloma specific T cells. Mean percentage of circulating tumor reactive CD4+ and CD8+ T cells increased from 1.5 and 1.96 respectively prior to the first infusion of CT-011, to 4.26 and 8.28 respectively 1 month following the third infusion. As determined by tetramer staining in the subset of patients who are HLA A2.1, infusion of CT-011 resulted in a mean 9 fold expansion of T cells specific to the MUC1 antigen, which is aberrantly expressed by myeloma cells. Notably, immunologic response to CT-011 persists at 6 months following completion of therapy. Clinical response, as determined by time to disease progression, will be determined with longer follow up, as the median time from transplant is presently 8 months. We are initiating enrollment to Cohort 2, in which patients will be vaccinated with an autologous DC/myeloma fusion vaccine 1 week prior to each dose of CT-011. These data demonstrate that CT-011 results in the expansion of tumor reactive lymphocytes in the early post-transplant period, providing an ideal platform for combination with a tumor vaccine. Disclosures: Rosenblatt: CureTech Ltd.: Research Funding. Schickler:CurTech Ltd.: Employment, Research Funding. Rotem-Yehudar:CureTech Ltd: Employment, Research Funding. Avigan:CureTech Ltd: Research Funding.

Description: Abstract 2866 Introduction: Diffuse large B cell lymphoma (DLBCL) is the most common lymphoid malignancy and is generally responsive to anthracycline-containing chemotherapy. However, 60% of patients (pts) will relapse after their first line treatment. At the time of relapse the only curative approach includes the use of a stem cell transplant (SCT). The incidence of DLBCL increases with age which creates a subset of pts who are not candidates for first line anthracycline-based chemotherapy, and a large subset of pts who are not candidates for SCT due to advanced age and/or co-morbidities. Thus, there is a significant unmet need for therapies with a low toxicity profile in elderly or medically unfit pts with DLBCL. 90Y-ibritumomab tiuxetan (90Y-IT) is an anti-CD20 murine antibody linked with a beta-emitting isotope approved for use in indolent lymphoma. Maintenance rituximab (R) has been reported to increase response rates and prolong remission duration in some lymphomas. We performed a phase II multicenter clinical trial to examine the efficacy of 90Y-IT induction followed by maintenance R in pts with DLBCL. Patients and Methods: Eligible pts were either intolerant of anthracycline-based chemotherapy or had relapsed or refractory CD20+ DLBCL with measurable disease. Pts had to be ineligible for SCT for reasons other than failure to harvest stem cells. Bone marrow involvement by lymphoma of less than 25% based on bilateral bone marrow aspirate and biopsy was required. R 250 mg/m2 was administered IV immediately followed by 111In-ibritumomab tiuxetan. Nuclear scans were performed at 24 and 48 hours to insure there was no altered biodistribution. On day 8 a second infusion of R 250 mg/m2 followed by 0.4 mCi/kg (for pts with a baseline platelet count 〉150,000/mm3) or 0.3 mCi/kg 90Y-IT (for pts with a baseline platelet count 100,000-149,000/mm3) was given. Pts with multiple extranodal sites or prior bone marrow involvement received CNS prophylaxis with intrathecal methotrexate or cytarabine. Maintenance R 375 mg/m2 was given on weeks 3–6, then weekly × 4 every 6 months × 4 cycles or until progression. Results: Between 10/2003 and 9/2009, 25 pts have been treated. During the course of the study, the ownership of the therapeutic agent changed three times and therefore enrollment was interrupted on two occasions. The median age of pts was 79 (range 45–91), 36% pts had a sIPI score 3 or more. The median number of prior regimens is 2 [0-5]. The 90Y-IT treatment regimen produced an overall response rate of 36% [9 pts] with 28% CR [7 pts]. To date, the mean OS is 18 months (median 8.1 months) with a median follow-up of 11.2 months. Among responding pts, the median OS has not been reached with a median follow-up of over 26.2 [0.1-71.4] months. Thirteen pts died within the first year, 6 patients (24%) continue to be in remission greater than 18 months, and 4 patients (16%) remain in long-term remission [39.9-71.4 months]. The most frequently observed toxicity was hematologic. Eleven percent of pts had grade 4 neutropenia with only one patient experiencing febrile neutropenia, and 16% of pts experienced grade 4 thrombocytopenia. There were no unexpected non-hematologic toxicities except for 1 patient that experienced extravasation. One late-occurring case of MDS/AML was reported that is possibly related to the study regimen, and one case of adenocarcinoma of the GI tract that is likely unrelated. Of note, none of the pts that progressed on the chemotherapy preceding this study achieved a response to the study regimen. Conclusions: The 90Y-IT treatment regimen has an acceptable toxicity profile in elderly or heavily pretreated pts with DLBCL. The two week outpatient 90Y-IT infusion produces response rates and durations similar to that of more prolonged cytotoxic chemotherapy regimens. Progression on previous chemotherapy predicts for poor response to 90Y-IT. Treatment with 90Y-IT can provide durable remission to a select subset of pts who are not candidates for SCT, or intensive anthracycline based chemotherapy. Disclosures: Off Label Use: We are describing a phase II study of the use of 90Y-Ibritumomab Tiuxetan for treatment of diffuse large B-cell lymphoma. Current FDA approved use of 90Y-Ibritumomab Tiuxetan includes relapsed or refractory, low-grade or follicular B-cell non-Hodgkin's lymphoma (NHL) or previously untreated follicular NHL who achieve a partial or complete response to first-line chemotherapy. Joyce:Spectrum Pharmaceuticals, Inc.: Research Funding; Cell Therapeutics Inc: Research Funding; Biogen Idec: Research Funding.

Description: Abstract 1883 Introduction: Acute myeloid leukemia (AML) arises from a malignant stem cell (LSC) population that is resistant to cytotoxic therapy and represents a reservoir for disease recurrence. A focus of research interest lies in developing cellular immunotherapy for the treatment of AML, in order to immunologically eradicate disease without the toxicity associated with allogeneic transplantation. In this study, we evaluated whether LSCs are effective targets for leukemia specific immunotherapy. To investigate this issue, we have developed a novel class of human T Cell Receptor-Like (TCRL) recombinant antibodies, which target MHC-peptide complexes mimicking their recognition. TCRL can be utilized to quantify the density of antigen presentation by LSCs; however, being able to distinguish LCSs from normal stem cells immunologically has been challenging. The tumor associated complex glycoprotein MUC1 may be used to define such an immature cell population. Methods and Results: We have previously demonstrated that LSCs strongly express the tumor associated antigen, MUC1, in contrast to normal hematopoietic stem cells. TCRL specific for the HLA-A2/MUC1 complex detectable by flow cytometry were constructed in order to quantify MUC1 antigen presentation. CD34+ cells were isolated from HLA-A2.1 AML patients with active disease and were analyzed for expression of MUC1 protein with the DF3 anti-MUC1 antibody and for HLA presentation of MUC1 peptide with the HLA-A2.1/MUC1 TCRL. Remarkably, despite the strong cell surface expression of the MUC1 protein (mean expression 68%, n=8), minimal levels of HLA bound MUC1 peptide was detected (mean expression 7.6%, n=8). CD34+ leukemia cells strongly expressed HLA-A2, indicating that the lack of TCRL recognition of the HLA-MUC1 complex was due to the lack of processing and presentation of the MUC1 peptide. Similarly, we demonstrated that leukemia progenitors minimally express HLA-A2-WT1 peptide complexes (mean expression 5%, n=3). These findings suggest that leukemia progenitors exhibit an immuno-privileged phenotype, potentially shielding them from leukemia specific immunotherapy. We have developed a MUC1 inhibitor peptide, G0-203, that has been shown to induce differentiation and apoptosis of leukemia progenitors. We evaluated whether, in addition to inducing differentiation of leukemia progenitors, exposure to GO-203 would augment antigen presentation. Isolated CD34+ AML cells were treated with 5μM GO-203 for 24 hours. Expression of HLA-A2/MUC1 peptide complexes, as determined by binding of the MUC1 TCRL antibody, was assessed before and after treatment. Exposure to GO-203 markedly upregulated the mean expression of HLA-A2.1/MUC1 peptide complexes by CD34+ leukemia cells (10% to 62%; n=4, p

Description: Abstract 848 Introduction: Acute myeloid leukemia (AML) arises from a malignant stem cell population that is resistant to cytotoxic therapy and represents a critical reservoir of conferring disease recurrence. A major focus of investigation is the identification of unique markers on leukemia stem cells (LSCs) that differentiate them from normal hematopoietic stem cells and thereby serve as potential therapeutic targets. MUC1 is a high molecular weight transmembrane glycoprotein that is aberrantly expressed in many epithelial tumors and confers cell growth and survival. We have developed an inhibitor of the MUC1-C receptor subunit that blocks oligomer formation and nuclear localization. In the present study, we have examined expression of MUC1 on LSCs as compared to normal hematopoietic stem cells and studied the effect of MUC1-C inhibition on the functional properties of LSCs. Methods and Results: Using multichannel flow cytometric analysis, we isolated the LSC compartment as defined by CD34+/CD38-/lineage- cells from bone marrow specimens obtained from patients with active AML. The majority of LSCs strongly expressed MUC1 with a mean percentage of 77% (n=6). These findings were confirmed by immunocytochemical staining of LSCs isolated by flow cytometric sorting. MUC1 expression was not detectable on the CD34- fraction of AML cells, but was present on the granulocyte-macrophage progenitor (GMP) fraction (CD34+/CD38+ cells) (mean=83%; n=6). In contrast, MUC1 expression was not observed on CD34+ progenitors isolated from normal donors (18%, n=6). In concert with these findings, RT-PCR analysis for MUC1 RNA demonstrated expression in CD34+ cells isolated from AML patients, but not normal volunteers. Notably, we also found that MUC1 expression selectively identifies malignant hematopoietic progenitors in a patient with chimerism between normal and leukemia derived stem cells. The presence of MUC1+CD34+ cells was detected in a patient with AML who achieved a morphologic complete remission following sex mismatched allogeneic transplantation. Using Bioview technology, we found that MUC1 is expressed only in the recipient (XX) CD34+ cells, representing residual malignant cells, whereas the donor (XY) derived CD34+ cells, representing the majority of the progenitors, lacked MUC1 expression. We subsequently examined the effects of MUC1-C inhibition on the capacity of leukemic progenitors to proliferate and support colony formation. MUC1-C inhibition with the GO-203 cell-penetrating peptide resulted in downregulation of the β-catenin pathway, an important modulator of cell division and survival, which is known to support the LSC phenotype. No significant change was detected with a control peptide, or with MUC1-C inhibition of progenitors isolated from a normal control. Furthermore, MUC1-C inhibition resulted in apoptosis, as demonstrated by flow cytometric staining for AnnexinV in AML CD34+ cells, but not in CD34+ progenitors isolated from normal volunteers (mean Annexin positive cells 53% and 5%, respectively, n=4). Consistent with these findings, the MUC1-C inhibitor, but not the control, peptide resulted in cell death of CD34+ cells isolated from AML patients, but not normal controls. Most significantly, exposure of CD34+ AML cells to the MUC1-C inhibitor resulted in loss of their capacity for colony formation in vitro with mean colonies of 4 and 40 for those cells exposed to the MUC1 inhibitor and a control peptide (n=2). In contrast, colony formation by normal hematopoietic stem cells was unaffected. Conclusions: MUC1 is selectively expressed by leukemic progenitors and may be used to differentiate malignant from normal hematopoietic stem cell populations. MUC1-C receptor subunit inhibition results in (i) downregulation of b-catenin signaling, (ii) induction of apoptosis and cell death, and (iii) disruption of the capacity to induce leukemia colony formation. Disclosures: Stone: genzyme: Consultancy; celgene: Consultancy; novartis: Research Funding. Kufe:Genus Oncology: Consultancy, Equity Ownership.

Description: Abstract 3583 Introduction: Acute myeloid leukemia (AML) arises from a population of leukemia initiating cells (LICs) that exhibit chemotherapy resistance and serve as a reservoir for disease recurrence. We have demonstrated that the oncoprotein, MUC1, is expressed on LICs but not normal hematopoietic stem cells. We have developed an inhibitor of the MUC1-C subunit, designated GO-203, which blocks MUC1-C dimerization and its oncogenic function. In a murine NSG (NOD-scid IL2Rg) model, we examined whether MUC1high progenitors are more effective in inducing leukemic engraftment as compared to MUC1low progenitors and evaluated the capacity of GO-203 to eradicate disease. Methods and results: To confirm the hypothesis that LICs are contained within the MUChigh progenitor population, lineage−/MUC1high and lineage−/MUC1low cells were isolated from AML derived bone marrow mononuclear cells by flow cytometric sorting, and transplanted (1×106 cells/mouse) into sub-lethally irradiated NSG mice using retro orbital injections. Leukemia engraftment was observed in all recipient mice (6/6) inoculated with MUC1high cells, with no evidence of normal human engraftment as shown by immunophenotype, cell morphology, and presence of cytogenetic abnormality. In a cohort of 6 mice, treatment with GO-203 was initiated 8 weeks after challenge with MUC1high AML cells following the demonstration of circulating leukemia cells. Leukemia was eradicated in 5/6 mice treated with GO-203. In contrast, 6/6 mice inoculated with MUC1low cells demonstrated engraftment with normal human hematopoetic cells as demonstrated by immunophenotype, cell morphology and absence of cytogenetic abnormality by FISH analysis. Engraftment of normal hematopoietic cells was unaffected by treatment with GO-203 in 4/5 of mice inoculated with MUC1low cells, supporting that GO-203 does not affect non-malignant hematopoiesis. We further determined the effect of GO-203 treatment in a prevention model. NSG mice were treated with 21 days of GO-203 beginning 24 hours after inoculation with CD34+/lineage−/MUC1high cells isolated from patient derived bone marrow mononuclear cells. Treatment with GO-203 prevented leukemia engraftment in 4/5 animals with the 5th animal showing only 1.9% leukemia cells. In contrast, 4/5 animals inoculated with the CD34+/lineage-/MUC1high cells in the absence of GO-203 demonstrated evidence of extensive bone marrow infiltration with leukemia cells comprising a mean of 33% of the bone marrow mononuclear cells. Conclusions: The results demonstrate that LICs are contained within the MUC1high fraction of AML derived progenitors and that MUC1low progenitors contain normal hematopoietic stem cells capable of normal engraftment. Treatment with the MUC1-C inhibitor, GO-203, prevents engraftment of leukemia, and eradicates established disease in a murine model. Our work indicates that GO-203 is a novel therapeutic agent for AML that targets LICs and has the potential to improve disease outcomes. Disclosures: Kufe: Genus Oncology: Consultancy.

Description: Abstract 4066 The fundamental challenge in designing an effective conditioning regimen for allogeneic transplantation involves the prevention of disease relapse while minimizing the risk for Graft versus Host Disease (GVHD). Treatment with total lymphocyte irradiation (TLI) and anti-thymocyte globulin (ATG) has been shown to minimize the risk of GVHD through the biasing of the T cell reconstitution towards an inhibitory phenotype. However, disease relapse remains a significant concern. Clofarabine is a second generation nucleoside analog with potent cytoreductive capacity and demonstrates efficacy in hematological malignancies. In this study, we examined the combination of clofarabine, TLI and ATG with respect to T cell reconstitution, risk for GVHD and transplant outcome. Sequential cohorts of 5 patients were treated with TLI and ATG alone or in conjunction with 20 mg/m2, 30 mg/m2 or 40 mg/m2 of clofarabine for 5 days. Cyclosporine and mycophenolate mofetil were administered as GVHD prophylaxis. Twenty patients have been enrolled (5 AML/MDS, 2 ALL, 6 lymphoma, 2 CLL, 5 myeloma) and received HLA matched peripheral blood stem cells collected from related (N=11) and unrelated donors (N=9). Of 19 evaluable patients, 15 are alive with a median follow up of 665 days. Day 30 and 100 mortality was 0% for TLI and ATG and 0% and 10% for those receiving clofarabine. The maximum tolerated dose (MTD) of clofarabine was 30 mg/m2 as 2 patients experienced treatment related mortality at the 40 mg/m2 dose level. Grade 5 infections and multiorgan failure occurred in both patients. All patients demonstrated engraftment with mean bone marrow donor chimerism of 92.5% at Day 30. The first cohort's ANC did not drop below 500 cells/uL, while median time to neutrophil engraftment in the patients who received clofarabine was 9 days. The median time to platelet recovery was 11 and 12 days for patients receiving TLI and ATG alone or with clofarabine, respectively (p=0.39). T cell reconstitution studies demonstrated a significant decrease in CD4+ cells to (

Description: Abstract 1908 Introduction: Tumor vaccines hold promise as a means of eliciting anti-myeloma immunity and controlling disease that may be resistant to chemotherapy and biologic therapy. We have developed a whole cell tumor vaccine, whereby patient derived plasma cells are transduced with an attenuated vaccinia vector that contains transgenes for the costimulatory molecules B7.1 (CD80), ICAM-1 (CD54), and LFA-3 (CD58), designated TRIad of COstimulatory Molecules (TRICOM). In this manner, a broad array of tumor antigens, including those which may be specific to a given patient, are presented in the context of costimulatory molecules that have been shown to be synergistic in the stimulation of the effector T-cells. In the present study, we evaluated the phenotype and functional characteristics of TRICOM transduced primary myeloma cells. Methods and results: Plasma cells were isolated from bone marrow aspirates obtained from patients with multiple myeloma following Ficoll density centrifugation. Bone marrow derived mononuclear cells were infected with a replication-defective poxviral vector, the modified vaccinia Ankara strain (MVA), encoding TRICOM, or a control empty MVA vector. The expression of costimulatory molecules was assessed using flow cytometric analysis 3 hrs following viral infection. Viral transduction using the TRICOM vector at the dose of 20 MOI (multiplicity of infection) increased the mean percentage of CD38+ cells expressing CD80, CD54 and CD58 from a minimal baseline level (below 5%) to 70%, 56% and 47%, respectively (n=4). Transduction with control MVA vector did not augment expression of costimulatory molecules on plasma cells (mean percent expression of CD80, CD54 and CD58 of 2.6%, 2.7% and 3.8%, respectively, n=4). Of note, compared to CD38+ plasma cells, the CD38 negative fraction of bone marrow derived mononuclear cells demonstrated a significantly lower TRICOM transduction efficiency (mean percent expression of CD80, CD54 and CD58 of 16%, 17% and 16%, respectively, n=4, p

Description: Abstract 948 Patients with acute myeloid leukemia (AML) achieve remission following chemotherapy; however, curative outcomes remain elusive due to relapse with chemotherapy-resistant disease. Allogeneic transplantation remains a potentially curative therapy for AML patients, but is associated with significant morbidity and mortality due to the lack of specificity of the alloreactive response. A promising area of investigation is the development of cancer vaccines that educate host immunity to more selectively target leukemia cells, including the stem cell compartment. Our group has developed a cancer vaccine model in which dendritic cells (DCs) are fused to autologous tumor cells, resulting in the presentation of multiple tumor antigens with the capacity to elicit a broad anti-tumor response. A fundamental challenge to developing a more effective tumor vaccine is overcoming the immunosuppressive milieu by which tumor cells evade host immunity. Key elements contributing to tumor-mediated immune suppression are the increased presence of regulatory T cells in patients with malignancy, and upregulation of the PD-1/PDL1 pathway. Tumor expression of PD-L1 promotes T cell tolerance by binding PD-1 on activated T cells and suppressing their capacity to secrete stimulatory cytokines. In addition, the PD-1/PDL-1 pathway has been shown to inhibit T cell-mediated lysis of tumor cells, potentially preventing a clinically meaningful immunologic response to vaccination. We are conducting a clinical trial in which AML patients who are in a first or second complete remission following chemotherapy receive three monthly doses of DC/AML fusion cells alone (Cohort 1) or in conjunction with anti-PD1 antibody, CT-011 (cohort 2). To date, 16 patients (9 males, 7 females; mean age 55 years) have been enrolled to the first cohort. All patients underwent successful tumor collection from either a bone marrow aspirate (N=12), collection of 20 cc of peripheral blood (N=3), or leukapheresis product (N=1) at the time of presentation with newly diagnosed AML (N=15) or first relapsed AML (N=1). The mean yield was 1.45×108 cells, and the mean viability was 90%. Tumor cells were subjected to immunohistochemical analysis to identify antigens unique to the leukemia fusion partner. Those patients achieving complete remission following 1–2 cycles of induction chemotherapy underwent leukapheresis for dendritic cell generation. Adherent peripheral blood mononuclear cells were isolated, cultured in the presence of GM-CSF and IL-4 for 5–7 days, and then exposed to TNFα for 48–72 hours to generate mature DCs. Mean viability of the DC preparation was 92%. DCs strongly expressed the co-stimulatory molecule CD86 (mean 75% expression). One patient died during remission induction chemotherapy and 3 patients were removed from study after induction chemotherapy to undergo allogeneic transplantation. Vaccine was successfully generated in 9 patients at a dose of 5×106 fusions cells, mean fusion efficiency of 30%, and viability of 87%. As a measure of their activity as antigen presenting cells, the capacity of the fusion cell preparation to stimulate allogeneic T cell proliferation ex vivo was quantified. In contrast to the leukemia preparation (mean stimulation index (SI) 3.7), the DC and fusion cell preparation were potent stimulators (mean SI 20.8 and 13.1, respectively). Vaccination with the DC/leukemia fusion vaccine was initiated within 12 weeks from count recovery following their final cycle of chemotherapy. 4 patients have completed vaccinations and are 2, 4, 5 and 6 months following the final vaccine. One patient was taken off study for disease progression one week after receiving his first vaccine. 4 patients experienced grade 1 vaccine site reactions. Biopsy of a vaccine site reaction demonstrated a dense T cell infiltrate. Additional vaccine related adverse events have included grade 1 ankle pain and edema. The remaining patients are undergoing chemotherapy, and when complete, will initiate vaccination. Peripheral blood samples are being collected prior to each vaccination and at 1, 3, and 6 months following completion of vaccination. Immune response targeting leukemia cells, leukemic stem cells, and leukemia associated antigens will be assessed. Levels of circulating regulatory T cells and T cell expression of PD1 will be measured. Time to disease progression will also be determined. Disclosures: Avigan: Curetech: Research Funding.