ALBERT

Description: Cdc7 is a heterodimeric serine/threonine protein kinase that is a key regulator in the process of initiation of DNA replication and the G1 to S phase transition. Both the kinase and its known substrates are over-expressed in the majority of human cancers. As a result of the recent progress in the areas of pharmacogenetics and high throughput screening technology, identifying specific small molecule inhibitors of cell cycle regulated protein kinases has provided a means not only to study these signal transduction pathways but also to identify potential novel therapeutic agents. To this end, we have developed an assay for Cdc7 kinase inhibitory activity using a highthroughput screening (HTS) approach, screening over 250,000 natural and synthetic small molecules. As a result, we have identified and confirmed seventeen compounds, representing nine different chemical scaffolds, with Cdc7 kinase inhibitory activity. Based on potency, we selected the lead compound (CKI-7) which was further characterized using kinase profiling, microarray experiments, and standard cell based cytotoxicity assays. These latter studies demonstrated that CKI-7 induced cytotoxicity of established leukemia and lymphoma cell lines in culture with inhibitory concentrations (IC50s) in the low nanomolar range. Significantly, CKI-7 likewise induced cytotoxicity of MDR1 overexpressing cell lines with similar IC50s, demonstrating that this novel compound can overcome a major mechanism of chemotherapy resistence in human tumor cells. We additonally demonstrate that CKI-7 induces cytotoxicity of patient-derived primary acute leukemia tumor cells (both chemotherapy naïve and relapsed/refractory samples) in vitro at similarly low nanomolar concentrations. In vivo dose-dependent anti-tumor activity of CKI-7 was subsequently demonstrated in a SCID-Beige mouse systemic tumor model utilzing a recently isolated Philadelphia chromosome positive acute lymphoblastic leukemia cell line (PhALL3.1). Standard cell cycle synchronization studies established that exposure to CKI-7 results in cell cycle dependent caspase 3 activation and apoptotic cell death. This cell death is the direct result of Cdc7 kinase inhibition by CKI-7 as demonstrated using a substrate biomarker assay. In conclusion, our data confirm that Cdc7 is a new promising target for cancer therapy, and that CKI-7, a selective small molecule inhibitor of this enzyme, is an equally promising novel cancer therapeutic agent.

Description: Background: The Philadelphia chromosome negative myeloproliferative neoplasms (MPN) includePolycythemia Vera (PV), Essential Thrombocythemia (ET) and Primary Myelofibrosis (PMF). These stem cell disorders carry a propensity to evolve into acute myeloid leukemia (MPN-blast phase [BP] or post-MPN AML) with a dismal prognosis not meaningfully improved by conventional anti-leukemia therapy. Thus, MPN-BP is an urgent unmet clinical need. Responses in patients with MPN-BP to hypomethylating agents and single agent ruxolitinib have been reported. More recently, combination of ruxolitnib and decitabine has demonstrated synergistic activity in vitro in cells derived from patients with MPN-BP and from a murine model of MPN-BP (Rampal et al PNAS 2014). These observations led us to explore the safety of combined decitabine and dose escalation of ruxolitinib in MPN-BP. Objective: To establish the maximum tolerated dose (MTD) of ruxolitinib in combination with a fixed dose of decitabine (DEC-RUX). Methods: We conducted an open label Phase I trial in patients with MPN acceleration phase (AP) as defined by 10%-19% blasts in the peripheral blood or bone marrow or a diagnosis of MPN-BP as defined by ≥ 20% blasts in the blood or bone marrow, following a previous diagnosis of ET, PV or PMF. Patients were enrolled in a standard 3+3 phase I design with an MTD defined as a dose

Description: Despite recent treatment advances, patients with Philadelphia (Ph) chromosome positive acute lymphoblastic leukemia (ALL) have a dismal prognosis. For this reason, a better understanding of the biology of Ph+ ALL, as well as novel approaches to treat this malignancy are needed. To date, there exist few primary Ph+ ALL tumor cell lines amenable to preclinical investigation of either the biology or treatment sensitivity of Ph+ ALL. In this report we describe the isolation and characterization of a novel Ph+ ALL cell line, designated PhALL3.1, obtained from a 62 year old patient who relapsed in the pleural space while receiving imatinib therapy. Tumor cells obtained from this malignant pleural effusion proliferated in standard complete RPMI media without supplementation with exogenous growth factors. The PhALL3.1 tumor expressed an unmutated p190 bcr-abl transcript, and demonstrated persistent sensitivity to both imatinib and desatinib in in vitro assays. Furthermore, we found that the PhALL3.1 cell line was readily amenable to retroviral gene transfer allowing for the modification of this line to express the green fluorescent protein-fire fly luciferase enzyme (PhALL3.1/GFP-FFL). When injected systemically into immune-compromised SCID-Beige mice, PhALL3.1/GFP-FFL tumor cells mimicked a clinical disease pattern with tumor involvement seen predominantly in the bone marrow, CNS, and spleen. This pattern of disease progression, as demonstrated by bioluminescent imaging, was further confirmed by histologic analysis. Likewise, when injected subcutaneously into immune-compromised mice, this cell line readily and reproducibly generated subcutaneous tumors. We conclude that the PhALL3.1 cell line is a novel, well characterized, cell line which readily proliferates both in vitro and in vivo, and therefore has significant value in future pre-clinical investigations into the biology of Ph+ ALL tumors as well as novel treatment approaches for this disease.

Description: Background: Outcomes are poor for older patients with acute myeloid leukemia (AML), high-risk myelodysplastic syndromes (MDS), or relapsed/refractory disease, and new therapies are needed. Using a chemosensitivity screening assay, we previously demonstrated that combination treatment with thioguanine and decitabine can restore therapeutic efficacy in primary leukemia cells isolated from patients with relapsed/refractory AML. To test the safety and synergistic efficacy of this combination in patients with advanced myeloid malignancies, we performed a Phase I dose-escalation trial of thioguanine given with decitabine. Patients and Methods: Patients with untreated AML ≥60 years of age and ineligible for standard induction, relapsed/refractory AML, and high-risk or relapsed MDS were eligible. Two thioguanine dose levels were evaluated: 80 and 120 mg/m2/day, given on Days 1-12 of induction and Days 1-7 of maintenance. Decitabine at 20mg/m2 was administered on Days 3-12 during induction and on Days 3-7 during maintenance. The primary objective was to determine the maximum tolerated dose (MTD) of thioguanine when given with decitabine. Key secondary objectives were to evaluate the overall response rate (ORR) and progression-free survival (PFS). Patient-specific pharmacodynamic measures to assess the biologic activity of thioguanine-decitabine were also performed. These included an in vitro chemosensitivity assay, BH3 profiling to measure the degree to which the leukemic blasts were primed for apoptosis, and genome-wide analysis of DNA methylation changes. Results: Twelve patients (median age 67; range 56-83) with de novo AML (n=1), secondary AML (n=6), relapsed/refractory AML (n=4), and chronic myelomonocytic leukemia (CMML) (n=1) were treated. Three patients experienced dose-limiting toxicity (DLT), which were acute renal failure requiring hemodialysis (80 mg/m2), persistent grade 4 leukopenia and thrombocytopenia (120 mg/m2), and grade 4 sepsis preventing continued treatment (120 mg/m2). Thioguanine at 80 mg/m2 was determined to be the MTD. Eleven of the 12 patients completed the first induction cycle, and 6 patients completed a second, identical induction cycle. The median number of cycles administered was 3 (range 1-8). One patient experienced a DLT prior to the first response assessment and was removed from study. The ORR in this intent-to-treat study was 67% (8/12). Six patients achieved a CR or CRi, one obtained a morphologic leukemia-free state, and one patient had a PR. Responses were observed in all disease types. Five of the 8 responses, including 4 CR/CRi, were achieved with thioguanine at 80 mg/m2, suggesting no loss of efficacy at the MTD compared with the higher dose level. All 11 evaluable patients had ≥50% reduction in bone marrow blast percentages after induction therapy. Six patients had previously received single-agent hypomethylating therapy, and 5 (83%) of these patients responded, demonstrating that thioguanine-decitabine can rescue prior hypomethylating agent failure. Out of the 8 responders, four (50%) proceeded to allogeneic stem cell transplantation (SCT), two relapsed after CR or CRi, one had a CNS-only relapse after achieving a CR, and one patient experienced DLT and was removed from the study. Of the four patients who proceeded to allogeneic SCT, two patients died in CR from transplant-related toxicity, one relapsed, and one patient remains alive and in remission greater than 2 years. Median PFS in responding patients was 42 weeks (range, 10-not reached, weeks). In vitro pharmacodynamic studies currently have been completed on samples from the first 6 patients treated on this trial. The chemosensitivity assay results on pre-treatment mononuclear cells directly correlated with initial response. In addition, significant apoptotic priming of the blasts, as suggested from BH3 profiling, also corresponded to initial clinical response. Conclusions: Thioguanine-decitabine can be administered safely and induce remission, even among patients who had previously been treated with hypomethylating agents. Intriguingly, preliminary results from the chemosensitivity screening assay and BH3 profiling correlated well with clinical responses. Additional correlative studies including DNA methylation analysis are ongoing to better understand the mechanism of synergy between thioguanine and decitabine. A multi-center Phase II trial is planned. Disclosures Jurcic: Astellas: Research Funding. Letai:Astra-Zeneca: Consultancy, Research Funding; Tetralogic: Consultancy, Research Funding; AbbVie: Consultancy, Research Funding.

Description: Abstract 3618 Poster Board III-554 The potential influence over cytokine activity by soluble receptors and the precise cytokine specificity they display suggests that soluble cytokine receptors may play a significant role in the response of cells to cytokine-mediated signaling. GM-CSF biology is mediated by binding to a heterodimeric receptor consisting of a major binding subunit (GMRα) and a major signaling subunit (β). GM-CSF Receptor can signal an astonishing variety of cellular functions, including protection from apoptosis, differentiation, maturation, activation and homing. GM-CSF Receptor α (GMRα) exists in both trans-membrane (tmGMRα) and soluble isoforms (sGMRα). sGMRα is believed to arise exclusively through alternative splicing of the GMRα gene product which removes the exon encoding the transmembrane domain. The splicing is such that the amino terminal 317 residues of sGMRα are homologous to the extracellular domain of tmGMRα, however, the deletion and subsequent frameshift replaces the transmembrane and cytoplasmic domains of tmGMRα with a unique 16 amino acid tail on sGMRα. It has been demonstrated that normal human monocytes constitutively secrete sGMRα, and sGMRα has been implicated in the biology of acute myeloid leukemias and the tumor-nerve interactions evoking bone cancer pain. sGMRα has been shown to bind directly to GM-CSF. We hypothesized that recombinant sGMRα(r-sGMRα) competitively inhibits the biological properties of GM-CSF. Using a baculoviral expression system, we produced N-terminal-His-tagged r-sGMRα. We exposed the GM-CSF and IL-3 dependent myeloid cell line, Mo7e, to varying concentrations of r-sGMRα and found that r-sGMRα directly inhibits GM-CSF dependent cell survival and proliferation in a dose-dependent manner. We found that r-sGMRα was toxic to Mo7e cells at concentrations as low as the Kd of r-sGMRα for GM-CSF (3nM). The finding that lower concentrations of GM-CSF resulted in lower Mo7e stimulation, but not toxicity, indicated that the mechanism of r-sGMRα toxicity was unlikely to simply be a reduction in effective GM-CSF concentration. To test whether the r-sGMRα toxicity seen in Mo7e cells was a result of GM-CSF neutralization or a result of signaling inhibition, we demonstrated that GM-CSF neutralizing antibody is cytotoxic during GM-CSF withdrawal, but not cytotoxic in the presence of IL-3, whereas r-sGMRα was cytotoxic under both conditions. Previous kinetic studies have demonstrated a time-dependent strengthening of the GMRα : β : GM-CSF complex; forming discrete “loose” and “tight” binding forms over minutes. Our studies indicate that neither pre-incubation of r-sGMRα with GM-CSF (6 hours) nor pre-incubation of cells with GM-CSF (prior to r-sGMRα addition) diminishes r-sGMRα's ability to inhibit GM-CSF-dependent cell survival and proliferation. We also determined that both r-sGMRα and GM-CSF degradation is independent of the other in vitro. To determine if r-sGMRα mediated cell death resulted from mechanisms other than GM-CSF signaling inhibition, we also tested additional myeloid cell lines that do not require exogenous cytokines for survival, including K562, (which does not express GMRα) as well as KG-1 and HL-60 cells (which both co-express GMRα and GMRb). While r-sGMRα does not inhibit survival or proliferation of KG-1 cells, r-sGMRα does inhibit proliferation of both HL-60 and K562 cells, but only at concentrations tenfold higher than the Kd (30nM). This inhibition was independent of GM-CSF's presence. These data indicate that soluble receptor may have biological effects beyond competition for ligand with the membrane bound receptor, and beyond the modification of GM-CSF half-life. The ability of r-sGMRα to inhibit IL-3 dependent survival in Mo7e cells, but not K562 cells, supports the notion that r-sGMRα may be directly inducing cytotoxicity by interacting with surface components that are likely part of the GM-CSF Receptor complex. Collectively, our data represent the first evidence demonstrating the modulation of GM-CSF signaling in human myeloid neoplasms by r-sGMRα and further illustrates the potentially significant role of soluble receptors in the response of cells to cytokine-mediated signaling. Disclosures: No relevant conflicts of interest to declare.

Description: Background: Neutropenia is a main side effect of cancer treatment and leads to increased risk of serious infections. Myeloid growth factors, including the granulocyte colony-stimulating factors (G-CSFs) filgrastim (Neupogen®) and pegfilgrastim (Neulasta®) and the granulocyte-macrophage colony stimulating factor (GM-CSF) sargramostim (Leukine®), stimulate neutrophil production and are commonly used as supportive care with myelosuppressive chemotherapy. GM-CSF also stimulates the production and activity of macrophages and dendritic cells, and it is hypothesized that the additional immune protection conferred by GM-CSF might reduce infection risk compared with the G-CSFs. We tested this hypothesis by comparing infection-related hospitalization rates and costs in patients using sargramostim, filgrastim, and pegfilgrastim for chemotherapy-induced neutropenia (CIN). Methods: This retrospective matched cohort study analyzed a large, nationally representative managed care claims database from over 30 health plans in the US during 2000 to 2007. CIN patients were identified as having ≥2 claims of sargramostim or filgrastim or ≥1 claim of pegfilgrastim; ≥1 cancer claim within 120 days prior to the start of a G/GM-CSF treatment episode (index date); and ≥1 chemotherapy claim within 60 days prior to the index date. The treatment episode began with the first G/GM-CSF claim satisfying the 120 day washout period and ended on the last claim date for sargramostim and filgrastim episodes; pegfilgrastim episodes ended on the last claim date plus a mean therapeutic duration of 19 days due to its long-acting nature. A G/GM-CSF claim more than 28 days after a prior claim was considered to be a new treatment episode. This analysis only considered the first treatment episode. Patients had to be ≥18 years old as of the index date and have continuous enrollment. Sargramostim patients were 1:1 matched with filgrastim and pegfilgrastim patients based on gender and year of birth. Outcomes included infection-related hospitalization rates and the associated cost per patient per month. Hospitalization rates were analyzed using univariate and multivariate Poisson methods. Covariates included the Charlson comorbity index, the number of chemotherapy agents received, whether the patient received myleosuppressive agents, and indicator variables for the presence of heart disease, renal disease, liver disease, metastasis, breast cancer, lung cancer, non-Hodgkin’s lymphoma, history of anemia, and neutropenia diagnosis on index date. Results: A total of 990 sargramostim-filgrastim and 982 sargramostim-pegfilgrastim matched pairs were analyzed. Cohorts had similar baseline characteristics, although differences were observed for the fraction of patients with a diagnosis of neutropenia at index date (sargramostim 65%, filgrastim 57%, pegfilgrastim 45%) and the percentage of patients who received myelosuppressive agents (sargramostim 54%, filgrastim 48%, pegfilgrastim 77%). Sargramostim patients experienced infection-related hospitalizations about half as often as patients using filgrastim (p=0.04) or pegfilgrastim (p=0.06). Multivariate analyses adjusted for confounding factors and found that sargramostim patients were 56% less likely to have infection-related hospitalizations compared to filgrastim and pegfilgrastim patients (p=0.03 for both). Infection-related hospitalization costs for sargramostim patients were $728/patient/month ($8,736/patient/year) and $226/patient/month ($2,712/patient/year) less compared to filgrastim (p=0.04) and pegfilgrastim patients (p=0.01), respectively. Conclusions: Among patients with CIN, use of sargramostim is associated with a reduced risk of infection-related hospitalization and lower associated costs compared to filgrastim or pegfilgrastim. Incidence Rate Ratios and Costs of Infection-Related Hospitalizations Univariate Multivariate IRR (95% CI) p-value Adjusted IRR (95% CI) p-value Sargramostim vs. Filgrastim 0.46 (0.22–0.97) 0.0422 0.44 (0.20–0.94) 0.0333 Sargramostim vs. Pegfilgrastim 0.52 (0.26–1.04) 0.0628 0.44 (0.21–0.90) 0.0256 Cost/patient/month Sargramostim Mean (SD) Comparison Group Mean (SD) Incremental Cost p-value Sargramostim vs. Filgrastim $138 ($2,534) $866 ($22,234) −$728 0.0380 Sargramostim vs. Pegfilgrastim $139 ($2,544) $365 ($5,557) −$226 0.0100

Description: The introduction of JAK inhibitors into clinical practice has produced significant clinical benefits for patients with MPNs, including reduction of splenomegaly and improvement in symptom burden. However, JAK inhibitors have demonstrated limited clinical ability to alter the natural history and biology of disease (such as reversal of fibrosis) in MPNs, and have produced only modest reductions in JAK2 V617F allele burden. This has prompted the evaluation of other therapeutic strategies in MPN patients. Heat Shock Protein 90 (Hsp90) inhibitors have recently emerged as promising potential treatment for MPN patients. Hsp90 physically interacts with JAK2, and inhibition of Hsp90 induces degradation of JAK2. Hsp90 is also a client chaperone for a variety of other proteins involved in pathways known to be critical for cancer cell differentiation, proliferation and survival. Preclinical studies have demonstrated that Hsp90 inhibitor therapy results in dose-dependent degradation of JAK2 and inhibition of downstream signaling pathways including STAT5, STAT3 and MAPK, as well as inhibition of growth of cells expressing mutant JAK2. In vivo treatment with Hsp90 inhibitors in MPN retroviral murine models led to normalized of peripheral blood counts, reduction in organomegaly, and improvement in overall survival. Based on these promising preclinical data, we conducted an open label phase II trial to assess the efficacy and confirm the safety of a novel Hsp90 inhibitor, AUY922 (Novartis) in patients with primary myelofibrosis (PMF), post-polycythemia vera and post-essential thrombocythemia myelofibrosis (IPSS-2 or higher) and in patients with polycythemia vera or essential thrombocythemia, who were refractory to, intolerant of, or ineligible for conventional therapy. The primary objective of the study was to determine the efficacy of AUY922 in this patient population. Secondary objectives included confirmation of safety and tolerability along with exploration of how treatment modified the biology of the disease. From 2012-2014, 6 patients were treated, 4 female, median age 55 (53-72), with MPNs; 4 with primary-MF, 1 with PV and 1 with ET, 4 patients were JAK2 V617F positive. Prior treatments included hydroxyurea in 5, anagrelide in 3, interferon in 1, azacitadine and pamidronate in 1. Median length of treatment was 39.5 months (1-145). 3 patients with MF experienced stable disease (including two patients with blast-phase disease), and one experienced an anemia response as measured by the European LeukemiaNet (ELN) response criteria. The patient with ET experienced stable disease as measured by the Revised International Working Group for Myeloproliferative Neoplasms Research and Treatment (IWG-MRT). Notable (greater than grade 2) adverse events included diarrhea in 5 patients, night blindness in 4, alkaline phosphatase elevation in 2, nausea in 2, emesis in 2, blurred vision in 1 and CPK elevation in 1 patient Serious Adverse Events (SAEs) included nausea, vomiting, diarrhea and altered mental status. As well, three patients experienced gastrointestinal bleeding as an SAE (grade 2 and 3). Two of these events were associated with an ileocecal ulcer. Two events were deemed to be possibly related to exposure to the study drug. The trial was terminated for this reason. Evaluation of the impact on JAK-STAT signaling is underway with evidence of reduction in total JAK2 protein following infusion (figure 1). Further evaluation of JAK-STAT signaling, impact on cytokine production, and JAK2 allele burden is under evaluation and results will be presented. In summary, treatment with AUY922 in patients with has demonstrated response including stable disease (including in blast-phase myelofibrosis patients) as well as anemia response and clinical improvement in one myelofibrosis patient. These data indicate that Hsp90 inhibition has clinical activity in MPN patients, and that further clinical therapeutic efforts targeting Hsp90 warrant investigation in this patient population. Figure 1. Responses to treatment with AUY922 per Revised International Working Group for Myeloproliferative Neoplasms Research and Treatment (IWG-MRT) and European LeukemiaNet criteria (ELN). Figure 1. Responses to treatment with AUY922 per Revised International Working Group for Myeloproliferative Neoplasms Research and Treatment (IWG-MRT) and European LeukemiaNet criteria (ELN). Figure 2. Western Blot analysis of total JAK2 from peripheral blood of patient at baseline, cycle 1 day 2, and end of treatment. Figure 2. Western Blot analysis of total JAK2 from peripheral blood of patient at baseline, cycle 1 day 2, and end of treatment. Disclosures Mauro: Pfizer: Consultancy; Novartis Pharmaceutical Corporation: Consultancy, Research Funding; Bristol-Myers Squibb: Consultancy; Ariad: Consultancy. Levine:Foundation Medicine: Consultancy; CTI BioPharma: Membership on an entity's Board of Directors or advisory committees; Loxo Oncology: Membership on an entity's Board of Directors or advisory committees.