ALBERT

Description: Tumors often co-exist with T cells that recognize somatically mutated peptides presented by cancer cells on major histocompatibility complex I (MHC-I). However, it is unknown why the immune system fails to eliminate immune-recognizable neoplasms before they manifest as frank disease. To understand the determinants of MHC-I peptide immunogenicity in nascent tumors, we tested the ability of thousands of MHC-I ligands to cause tumor subclone rejection in immunocompetent mice by use of a new ‘PresentER’ antigen presentation platform. Surprisingly, we show that immunogenic tumor antigens do not lead to immune-mediated cell rejection when the fraction of cells bearing each antigen (‘clonal fraction’) is low. Moreover, the clonal fraction necessary to lead to rejection of immunogenic tumor subclones depends on the antigen. These data indicate that tumor neoantigen heterogeneity has an underappreciated impact on immune elimination of cancer cells and has implications for the design of immunotherapeutics such as cancer vaccines.

Description: INTRODUCTION T Cell Receptor mimic (TCRm) antibodies to low-density peptide epitopes from undruggable intracellular proteins presented in the context of major histocompatibility (MHC) molecules are therapeutically effective in mouse models of human cancers. CD47 blockade by use of a high affinity SIRPα variant protein (CV1) has been shown to improve the effects of monoclonal antibodies to high-density antigens in tumor models by enhancement of antibody dependent cellular phagocytosis (ADCP). We asked if combination therapy with a TCRm antibody to Preferentially Expressed Antigen in Melanoma (PRAME) could enhance activity of both drugs in vitro and in vivo. Additionally, we explored the role of macrophage-secreted cytokines in the enhanced in vivo activity. METHODS We performed in vitro ADCP assays with human acute myeloid and acute lymphoid leukemia cell lines containing antigens of interest using the two agents alone and in combination. We performed therapy experiments in NSG mice using the same leukemia cell lines transformed with a luciferase vector and measured tumor burden through bioluminescent imaging. Survival was measured. We examined cell-surface expression of epitopes of interest and HLA on cell lines in vitro after incubation with IFNγ and TNFα using flow cytometry and performed in vitro ADCP assays with the leukemia cell lines after pretreatment with IFNγ. RESULTS CV1 and TCRm antibody showed additive effects in vitro with a statistically significant increase in phagocytosis in both antigen positive cell lines with combination therapy versus single agent therapy. CV1 and TCRm antibody showed greater than additive therapeutic effects in vivo with a 3-log reduction in leukemia burden relative to control untreated mice and a 5-10 fold reduction relative to single agent groups. After therapy was stopped, mice treated with the combination had statistically significant increases in survival (p

Description: Background: The use of short-ranged (50-80 µm), high-energy (~100 keV/µm) α particle-emitting isotopes for radioimmunotherapy may result in more specific tumor cell kill and less damage to normal tissues than β-emitters. 225Ac-lintuzumab consists of a radiometal that emits four α-particles linked to an anti-CD33 antibody. A phase I trial showed that 225Ac-lintuzumab is safe at doses ≤ 3 µCi/kg and has anti-tumor activity against relapsed/refractory AML across all dose levels studied (Jurcic et al. ASH, 2011). We are conducting a multicenter, phase I dose-escalation trial to determine the maximum tolerated dose (MTD), toxicity, and biological activity of fractionated-dose 225Ac-lintuzumab in combination with LDAC. Patients and Methods: Patients ≥ 60 years with untreated AML not suitable for standard induction chemotherapy (e.g., antecedent hematologic disorder, unfavorable cytogenetic or molecular abnormalities, and significant comorbidities) were eligible. Patients received LDAC 20 mg twice daily for 10 days every 4-6 weeks for up to 12 cycles. During Cycle 1, two fractions of 225Ac-lintuzumab were given one week apart, beginning 4-7 days following completion of LDAC. To prevent radiation-induced nephrotoxicity, patients were given furosemide while receiving 225Ac-lintuzumab and spironolactone for one year afterward. 225Ac doses were escalated using a 3+3 design. Four dose levels were studied with a total accrual of up to 24 patients. In planned analyses, dose escalation proceeded if 〈 33% of patients in a cohort experienced dose-limiting toxicity (DLT). Results: Fourteen patients (median age, 77 years; range, 68-87 years) completed therapy. An additional patient received only one of two planned fractions of 225Ac-lintuzumab due to technical issues and is excluded from analysis. Nine (64%) had prior myelodysplastic syndrome, for which seven received prior therapy with hypomethylating agents (n=6) or allogeneic hematopoietic cell transplantation (n=1). One patient (7%) had chronic myeloid leukemia in molecular remission prior to development of AML. Nine patients (64%) had intermediate-risk and five (36%) had unfavorable cytogenetics. Median CD33 expression was 81% (range, 45-100%). 225Ac-lintuzumab was given at 0.5 (n=3), 1 (n=6), 1.5 (n=3), or 2 (n=2) μCi/kg/fraction. Up to 4 cycles of LDAC were administered. DLT was seen in one patient at 1 µCi/kg/fraction who had grade 4 thrombocytopenia with marrow aplasia for more than 6 weeks following therapy. Hematologic toxicities included grade 4 neutropenia (n=4) and thrombocytopenia (n=6). Grade 3/4 non-hematologic toxicities included febrile neutropenia (n=7), pneumonia (n=4), bacteremia (n=1), cellulitis (n=1), transient creatinine increase (n=1), hypokalemia (n=1), rectal hemorrhage (n=1), and generalized weakness (n=2). Eight of 11 patients (73%) evaluated after Cycle 1 had bone marrow blast reductions (mean reduction, 72%; range, 34-100%). Seven (64%) had blast reductions of at least 50%. Objective responses (1 CR, 1 CRp, 2 CRi) were seen in four of the 14 patients (29%) after one cycle of therapy (Table 1). Responses were seen only at doses ≥ 1 µCi/kg/fraction (4 of 11 patients, 36%). Median progression-free survival (PFS) was 2.7 months (range, 1.7-16.9 months). Median overall survival (OS) was 5.5 months (range, 2.2-24 months). Conclusions: Fractionated-dose 225Ac-linutuzmab can be safely combined with LDAC and produce remission in older patients with untreated AML. Dose escalation continues to define the MTD. Additional patients will be treated at the MTD in the phase II portion of this trial to determine response rate, PFS, and OS. Table 1. Objective Responses Response Dose Level (μCi/kg/fraction) Total (n=14) 0.5 (n=3) 1 (n=6) 1.5 (n=3) 2 (n=2) CR 0 0 1 (33%) 0 1 (7%) CRp 0 0 0 1 (50%) 1 (7%) CRi 0 1 (17%) 1 (33%) 0 2 (14%) Overall Response 0 1 (17%) 2 (67%) 1 (50%) 4 (29%) Abbreviations: CR, complete remission; CRp, CR with incomplete platelet recovery; CRi, CR with incomplete count recovery. Disclosures Jurcic: Ambit Biosciences: Research Funding; Astellas Pharma US, Invc.: Research Funding; Tetralogic Pharmaceuticals: Research Funding; Sunesis Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees; Novartis Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees; Amgen: Consultancy; Bayer Pharmaceuticals: Consultancy; Merck and Co.: Consultancy; Celgene Corp.: Research Funding; Actrinium Pharmaceuticals, Inc.: Membership on an entity's Board of Directors or advisory committees, Research Funding. Off Label Use: Ac-225-lintuzumab is an investigational agent being developed for the treatment of acute myeloid leukemia.. Pagel:Actinium Pharmacetuicals, Inc.: Equity Ownership. Park:Actinium Pharmaceuticals, Inc.: Research Funding; Juno Therapeutics: Consultancy. Levy:Takeda: Consultancy. Perl:Actinium Pharmaceuticals, Inc.: Research Funding. Earle:Actinium Pharmaceuticals, Inc.: Employment. Cicic:Actinium Pharmaceuticals, Inc.: Employment, Equity Ownership. Scheinberg:Actinium Pharmaceuticals, Inc.: Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties.

Description: Acute myeloid leukemia (AML) is an aggressive hematological malignancy with a 5-year overall survival rate of less than 30% which causes over 10,000 deaths per year in the United States. Treatment options for this disease increasingly include epigenetic drugs, such as hypomethylating agents (e.g. decitabine) or histone deacetylase (HDAC) inhibitors (e.g. pracinostat) which can function via direct cytotoxic mechanisms and also through altered differentiation of AML blasts; immunomodulatory effects like reactivation and presentation of cancer testis antigens in context of human leukocyte antigen (HLA) complexes have been reported as well, which may result in clearance of cells via the adaptive immune system. However, the landscape of immunogenic T cell epitopes induced by these drugs might be even broader than reported since standard analyses only consider exonic protein sequences and do not take into account typically untranslated genomic regions. Recently, it has been shown that single and combination treatment of decitabine and pracinostat can induce cryptic transcription start sites in generally epigenetically repressed solitary long-terminal repeats (LTRs) of the LTR12C family which give rise to novel mRNAs and resulting protein variants. We hypothesized that the intronic parts of these gene products might provide a source of cryptic T cell epitopes with high immunogenic potential, which are induced through epigenetic drug treatment. To test this hypothesis, we treated 5 different AML cell lines (HL-60, U937, OCI-AML02, MOLM13, AML14) with (1) DMSO, (2) 500 nM decitabine or (3) a combination of 500 nM decitabine and 100 nM pracinostat for 72 hours to induce transcription of non-annotated transcription start sites. Subsequently, HLA class I complexes were immunopurified and peptides presented by these complexes isolated and analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The activation of silenced genes by epigenetic drug treatment with either decitabine alone or the combination treatment yielded increases of about two-fold in the identified unique HLA ligands. This increase in peptide identifications also led to improved detection of cancer testis antigen-derived epitopes, as has been reported before. Intriguingly, by adding LTR12C derived sequences stretching from the published GATA2 specific binding site until the next genomic exon to the peptide search analyses we were able to identify several cryptic peptides from 4 out of 5 AML cell lines derived from these usually untranscribed genomic regions. The identifications were exclusively dependent on previous treatment with either decitabine alone or in combination with pracinostat. Though the immunogenicity of these HLA ligands has not been determined yet, we assume that due to their genetically repressed state in untreated cells, these new peptide sequences represent a new class of neoepitopes, with potential to be novel targets of existing T cells within patients or after augmentation by other immunotherapies. In summary, we demonstrated for the first time the induced presentation of epitopes from normally untranscribed LTR12C regions through epigenetic drug treatment and therefore provide a previously undescribed source of potential targets for immunotherapy in AML. Disclosures Scheinberg: Eureka: Consultancy; Ensyce: Consultancy.

Description: Preferentially expressed antigen in melanoma (PRAME) is a well-validated target for T cell-based immunotherapy in leukemias and solid tumors. PRAME is a retinoic acid receptor binding protein that prevents retinoic acid-mediated differentiation, proliferation arrest, and apoptosis. As a cancer-testis antigen, PRAME has limited expression in healthy adult tissue that is restricted to the testes, ovaries, and endometrium. However, PRAME is over-expressed in multiple cancers including ALL, AML, melanomas, and breast cancers, making it a specific and highly attractive therapeutic target. PRAME is an intracellular protein making it impossible to target using traditional antibodies and it is not currently druggable. After proteasomal processing, the PRAME300-309 peptide is presented on the cell surface in the context of HLA*A02:01 molecules, for recognition by CD8 T cells. We therefore hypothesized that a TCR-mimic (TCRm) monoclonal antibody that recognizes surface PRAME300-309 presented by HLA*A02:01 could have therapeutic activity. Here, we describe Pr20, a therapeutic TCRm antibody, specific for the PRAME300-309 peptide in complex with HLA*A02:01, identified through a human scFv phage display library screen. Pr20 was engineered into full length human IgG1. Pr20 exhibited specific binding to PRAME300-309 -pulsed TAP-deficient T2 cells and bound PRAME+/ HLA*A02:01+ Ph+ ALL and AML, demonstrating that endogenously presented PRAME300-309 could be recognized by Pr20. Pr20 was determined to have 4 nM binding affinity by scatchard plot analysis. The specific epitope was mapped using alanine substitutions of non-anchor residues in the PRAME300-309 peptide and determined to primarily require the C-terminal residues. Pr20M, an afucosylated form of the antibody with enhanced Fc binding, mediated antibody-dependent cellular cytotoxicity (ADCC) in-vitro in a PRAME+/ HLA*A02:01+ restricted manner. Pharmacokinetic studies in C57BL/6 mice indicated that Pr20M was stable in-vivo and biodistribution studies in HLA*A02:01 transgenic mice suggested that there was no significant antibody sink. Pr20M was therapeutically active in established xenograft leukemia models in NSG mice (T, B, and NK-deficient). Interestingly, Pr20 binding to PRAME+/HLA*A02:01+ melanomas was minimally detectable, but was dramatically increased upon treatment with IFNγ, which also led to an increased sensitivity to ADCC. The data provide rationale for developing TCRm antibodies against intracellular oncoproteins as therapeutics. Disclosures No relevant conflicts of interest to declare.

Description: Oncogenic co-operation between c-Myc and activated phosphoinositide 3-kinase (PI3K) signaling pathways is crucial in lymphomagenesis, providing an opportunity for developing mechanism-based therapy to disrupt this co-operative survival mechanism. Combining constitutive c-Myc expression with constitutive PI3K activity in mouse germinal center B (GCB) cells resulted in Burkitt lymphoma-like tumors. Furthermore, analysis of primary human Burkitt lymphoma (BL) tissue sections revealed that two-thirds of the cases expressed high levels of phosphorylated AKT and S6 proteins, indicative of PI3K and mTORC1 activation. Prior attempts to develop small molecule inhibitors that specifically and directly target c-Myc protein have been unsuccessful. However, c-Myc cellular protein abundance can be decreased by using epigenetic modifying drugs (HDAC inhibitors or bromodomain/BET inhibitors) that are known to inhibit c-Myc transcription. Several investigators attempted disrupting c-Myc and PI3K cooperation by combining HDAC inhibitors and PI3K pathway inhibitors and synergic activity was demonstrated in DLBCL irrespective of subtype. In this study, we assessed the efficacy of CUDC-907, a new rationally designed dual inhibitor of PI3K and HDACs, in a panel of lymphoma cell lines. CUDC-907 treatment resulted in a dose- and time-dependent growth inhibition and cell death of DLBCL cell lines, irrespective of the cell of origin. CUDC-907 treatment down-regulated the phosphorylation of PI3K downstream targets, including AKT, PRAS40, S6, and 4EBP1, increased histone 3 acetylation, and decreased Myc protein levels. SILAC-based quantitative mass spectrometry demonstrated that CUDC-907 treatment decreased the protein levels of several components of the B cell receptor (BCR) and Toll like receptor (TLR) pathways, including BTK, SYK, and MyD88 proteins. These cellular changes were associated with an inhibition of NF-kB activation. CUDC-907 demonstrated in vivo efficacy with no significant toxicity in a human DLBCL xenograft mouse model. Collectively, these data provide a mechanistic rationale for evaluating CUDC-907 for the treatment of patients with c-Myc and PI3K-dependent lymphomas. Disclosures No relevant conflicts of interest to declare.