Abstract Abstract 928 Background: Lenalidomide (Len) has been shown to be clinically effective against multiple myeloma (MM) as a single agent and in combination for both newly diagnosed and relapsed/refractory settings, and is increasingly utilized in maintenance therapy. Despite the successful use of Len, some patients display a poor response or become refractory after an initial response. As previously published by our group, hyperactivation of Wnt/β-catenin signaling was identified as a mediating factor for Len-specific resistance in MM cell lines. Additionally, a downstream transcriptional product of Wnt/β-catenin signaling is the extracellular matrix binding protein CD44, which has been implicated in cellular-adhesion-mediated drug-resistance (CAM-DR) in several cancer models, including dexamethasone-resistant MM. These early studies established a rationale to target aberrant Wnt/β-catenin signaling and CD44-mediated CAM-DR to overcome Len resistance. Methods: Wild-type (Wt) and lenalidomide-resistant (LR) MM cell lines were used as models to evaluate the effectiveness of targeting the Wnt/β-catenin signaling pathway and CD44-dependent CAM-DR to overcome Len resistance with Wnt/β-catenin antagonists, shRNA knockdown of CD44, and anti-CD44 neutralization. Results: All-trans-retinoic acid (ATRA) inactivates β-catenin nuclear signaling by altering β-catenin cellular distribution. Combination treatment with Len and ATRA produced a significant (p<0.05) anti-proliferative and cytotoxic effect on LR cells. ATRA exposure of both Wt and LR cells also effectively reduced β-catenin-dependent T-cell factor/lymphoid enhancer factor (TCF/LEF) promoter activity by as much as 70%, and reversed the apparent increase in activity stimulated by acute Len treatment. Cell surface and total CD44 levels were effectively reduced by as much as 60% when cells were treated with ATRA, even in the presence of Len. Primary patient samples from individuals displaying Len refractory disease were treated ex vivo with Len and ATRA in combination, resulting in a synergistic reduction in cellular viability. We investigated an alternative approach to β-catenin abrogation using FH535, a small molecule β-catenin antagonist that disrupts the transcriptional-dependent association of β-catenin and TCF. Combinations of Len with FH535 resulted in an additive suppression of cellular proliferation on Wt MM cell lines. Similarly, FH535 reduced the viability and enhanced Annexin-V staining on LR KAS-6 and U266 cells by as much as 80% and 55% respectively when used in combination with Len. In addition to selectively targeting β-catenin to overcome LR MM, we evaluated the role of CD44 in LR cell lines. Flow cytometric analysis revealed that cell surface CD44 levels were increased by as much as 11-fold in LR KAS-6 cells when compared to Wt cells. Acute treatment of Wt ANBL-6 and MM1.S cells with Len moderately enhanced the cell surface levels of CD44 by approximately 1.5-fold. The enhanced CD44 surface levels induced by Len on LR ANBL-6 and KAS-6 corresponded to at least a 2-fold increase in cellular adhesion to bone marrow stromal cells, and enhanced adhesion to increasing concentrations of hyaluranon (HA). Pre-incubation with free HA or CD44 neutralizing antibodies reduced the adhesive properties of both Wt and LR ANBL-6 and KAS-6 cell lines. shRNA knockdown of CD44 sensitized LR cells to Len when compared to a non-specific scrambled shRNA transfection. Interestingly, extensive statistical analysis of banked gene expression data from primary samples located at the Multiple Myeloma Research Consortium (MMRC) database revealed a significant correlation (p=0.01) between high expression levels of CD44 and a poor outcome prognosis, albeit irrespective of treatment regimens. Furthermore, current analysis of surface CD44 levels on primary MM bone marrow aspirates suggests that those patients who have received prior Len therapy have an overall increase in CD44 when compared to Len naïve patients. Conclusions: Collectively, our results suggest that MM cell Len resistance is, at least in part, dependent on β-catenin and CD44 and that selective targeting of these cellular proteins in conjunction with Len treatment represents a rational approach for clinical treatment of Len-resistant MM. Disclosures: Aukerman: Celgene Corporation: Employment. Lopez-Girona:Celgene Corp: Employment, Equity Ownership. Hussein:Celgene Corporation: Employment. Chopra:Celgene Corporation: Employment. Orlowski:Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding.

Abstract Abstract 2845 Lenalidomide is an immunomodulatory agent that has both direct tumoricidal and immunomodulatory activities which are critical for its clinical activity in the treatment of various hematologic malignancies. This activity is at least in part mediated by enhanced T-cell and NK-cell effector function to eliminate tumor B cells, attributed to restoration of impaired T-cell activity and formation of immunologic synapses. Rituximab is an anti-CD20 monoclonal antibody that is active in the treatment of B-cell lymphomas through a variety of mechanisms, including antibody-dependent cellular cytotoxicity (ADCC). Preclinical studies and early clinical trials have shown an enhancement, and potentially synergy, in antitumor activity when lenalidomide is combined with rituximab. In order to further explore the mechanistic basis of this enhancement we investigated the impact of lenalidomide and rituximab on immune synapse formation and ADCC. We have previously shown that the combined use of lenalidomide and rituximab enhances NK cell-mediated immune synapse formation and the resultant cytotoxicity, versus each agent alone. Here we evaluate the molecular events that take place on the cell surface upon exposure of JeKo-1 cells (mantle cell lymphoma) and primary B-CLL cells to lenalidomide alone or lenalidomide plus rituximab. Change in CD20 expression resulting from exposure to vehicle control (0.1% DMSO) or 1 μM lenalidomide for 30 min or 24, 48, 72 hrs was assessed using immunocytochemistry, flow cytometry and isolation of cell membrane-associated proteins followed by Western blotting. At all time points evaluated, levels of cell surface and cell membrane-associated CD20 expression were unchanged in JeKo-1 cells. However, the distribution of CD20 was dramatically altered within 30 minutes after addition of lenalidomide. CD20 redistribution was accompanied by F-actin polymerization and lipid raft aggregation associated with the polarized localization (capping) of a number of proteins including CD20, CD19 and cytoskeleton signaling molecules Rac1 and Vav1, critical regulators of immune synapse formation in effector cells. Of note, other surface proteins involved in signaling such as CD45 were not part of this capping mechanism. By 48 hours of lenalidomide treatment, the majority of JeKo-1 cells (>80%) showed continued capping of CD20. These responses were also seen in primary B-CLL cells, although the effects were variable. In addition, CD20, F-actin and lipid rafts co-localized at the immune synapses formed between JeKo-1 and NK cells treated with either 1 μM lenalidomide for 24 hrs, 0.1% DMSO for 24 hrs followed by 10 μg/ml rituximab for 30 min, or treated sequentially with 1μM lenalidomide for 24 hrs followed by 10 ug/ml rituximab for 30 min. Lenalidomide and rituximab induced similar effects on B-CLL cells and the immune synapses formed between B-CLL and NK cells. We also determined whether formation of lipid rafts and actin cytoskeleton modifications were a prerequisite for CD20 capping. Cholesterol extraction from JeKo-1cells by 5 mM methyl-β-cyclodextrin (MCD) treatment for 30 min led to complete abrogation of lenalidomide-induced capping. The polymerization of the F-actin cytoskeleton and capping of CD20 was also affected, with no impact on cell viability. In addition, MCD treatment inhibited the formation of immunologic synapses between JeKo-1 cells and NK cells treated with 1 μM lenalidomide alone and in cells co-treated with 1 μM lenalidomide and 10 μg/ml rituximab. These data are consistent with a requirement for the integrity of lipid rafts to maintain the capping of CD20 and to potentially mediate lenalidomide enhancement of ADCC by rituximab. Our results further demonstrate that lenalidomide does not down-regulate CD20, but rather induces its polarized localization at the cell surface. The capping of CD20 is accompanied by redistribution of proteins such as Vav1 and Rac1 that become part of the immune synapse complex. Therefore the capping process induced by lenalidomide appears integral to immune synapse formation and may coordinately enhance the clustering of both the CD20 antigen and the attached rituximab, potentially further enhancing its activity, which would support the clinical combination of these agents. Ongoing studies are currently examining the role of the capping process and intracellular signaling cascades in the direct tumoricidal activity of lenalidomide. Disclosures: Gaidarova: Celgene Corporation: Employment, Equity Ownership. Mendy:Celgene Corporation: Employment. Heise:Celgene Corporation: Employment. Aukerman:Celgene Corporation: Employment. Daniel:Celgene Corporation: Employment. Chopra:Celgene Corporation: Employment. Lopez-Girona:Celgene Corporation: Employment, Equity Ownership.

B-cell chronic lymphocytic leukemia (B-CLL) is a lymphoproliferative disorder characterized by the surface expression of CD20, CD5 antigens, as well as the receptor CD40. Activation of CD40 by its ligand (CD40L) induces proliferation and rescues the cells from spontaneous and chemotherapy-induced apoptosis. CD40 activation also induces secretion of cytokines, such as IL-6, IL-10, TNF-alpha, IL-8, and GM-CSF, which are involved in tumor cell survival, migration, and interaction with cells in the tumor microenvironment. Here we demonstrate that in primary B-CLL tumor cells, the novel antagonist anti-CD40 monoclonal antibody, HCD122, inhibits CD40L-induced activation of signaling pathways, proliferation and survival, and secretion of cytokines. Furthermore, HCD122 is also a potent mediator of antibody-dependent cellular cytotoxicity (ADCC), lysing B-CLL cells more efficiently than rituximab in vitro, despite a significantly higher number of cell surface CD20 binding sites compared with CD40. Unlike rituximab, however, HCD122 (formerly CHIR-12.12) does not internalize upon binding to the cells. Our data suggest that HCD122 may inhibit B-CLL growth by blocking CD40 signaling and by ADCC-mediated cell lysis.

556 Although low grade lymphomas generally respond well to treatment with chemotherapeutic agents the disease typically follows a course of recurrent relapse associated with shorter remissions resulting in disease progression. Rituximab (R) treatment in combination with CHOP (Cyclophosphamide, C; Doxorubicin, H; Vincristine, O, Prednisone, P) has been reported to increase the overall response rate in low grade NHL (up to 95%), but most patients eventually relapse highlighting the need for additional innovative strategies to improve the durability of clinical responses. In two Phase I studies we have shown that Proleukin (aldesleukin, rhIL-2) can augment activity of rituximab via expansion and activation of NK cells to mediate antibody-dependent cellular cytotoxicity (ADCC; Gluck et al., 2004. Clin. Cancer Res.). In this preclinical study, we investigated the ability of Proleukin (IL-2) to enhance the overall tumor efficacy and durability of responses when administered in combination with rituximab following initial CHOP/R treatment in a Daudi xenograft model of low grade CD20+ B-cell lymphoma. CHOP/R was modeled on clinical dosage regimens and treatment commenced when tumors were established (150-200mm3). Treatment groups comprised either CHOP alone (C, 40 mg/kg, i.v; H, 3.3 mg/kg, i.v; O, 0.5 mg/kg, i.v all administered on Day 1 and P, 0.2 mg/kg, p.o., days 1-5) or a 4 week cycle of rituximab (R) alone (10 mg/kg, once weekly, i.p) starting day 1 or CHOP in combination with a 4 week cycle of rituximab. Proleukin treatment (1 mg/kg; s.c., thrice weekly) commenced either concomitantly on day 1 or one week later on day 8 for a total of 4 weeks. All treatments were well tolerated. CHOP/R significantly inhibited tumor growth (99% tumor growth inhibition, TGI; 5/10 CR) and was superior to either R (63% TGI, 1 CR) or CHOP (77% TGI, 1 CR) alone. Notably, the addition of Proleukin (IL-2) to CHOP/R commenced either on D1 or D8 induced substantial tumor regression with 7/10 and 9/10 CR respectively and was superior to CHOP/R (p 164 days) compared to CHOP/R alone (90 days). In summary, our preclinical findings indicate that CHOP/R followed by IL-2/R is safe, efficacious and is predicted to add benefit in delaying time to progression compared to CHOP or CHOP/R. The clinical utility of CHOP/R followed by IL-2/R remains to be evaluated.

678 Cancer is a complex disease that involves multiple genetic defects that drive tumor-cell proliferation, suggesting that simultaneously inhibiting multiple cell signaling pathways may lead to a more favorable therapeutic outcome. In this work we examined the activity of CHIR-258, a novel, orally active, small molecule, that exhibits a multitarget kinase selectivity against Class III, IV and V receptor tyrosine kinases (RTKs). Biochemical IC50s of CHIR-258 for FLT3 = 1 nM; cKIT = 2 nM; VEGFR1/2/3 ∼ 8 nM; FGFR1/3 ∼ 10nM;, PDGFRb = 27 nM; and CSF-R1 = 36 nM. The in vivo activity of CHIR-258 was shown to result from direct inhibition of RTK signaling and consequent induction of apoptosis in tumor cells, in addition to potent inhibition of angiogenesis. The anti-tumor efficacy of CHIR-258 was evaluated in a variety of murine and human solid tumor models (colon, prostate, breast, ovarian) and hematological cancers (leukemia and myeloma), and compared with that of more “selective” kinase inhibitors currently in clinical development. In solid tumor models, CHIR-258 exhibited potent tumor inhibition, including regression, and was more effective than more “selective” RTK inhibitors. When compared to a RTK inhibitor with a similar target profile and “broad” selectivity as CHIR-258, efficacy was generally equivalent with continuous daily dosing. However, CHIR-258 exhibited more potent anti-tumor activity when cyclic or intermittent dosing schedules were evaluated. Pharmacodynamic studies have shown that target modulation and inhibition of signaling pathways were maintained for 24 hours followed by induction of tumor cell apoptosis. Our data suggests that biological activity of CHIR-258 is dependent upon degree and duration of response, which likely plays a role in the efficacy observed in the absence of continuous drug treatment and differentiates CHIR-258 from other RTK inhibitors. In conclusion, these data support the development of multitargeted agents like CHIR-258 based on our understanding of the relative contributions of the disease-relevant targets in tumorigenesis.

CD40 is expressed on most B cell malignancies including multiple myeloma and represents an attractive target for antibody therapy. We have generated a novel, highly potent, fully human antagonistic anti-CD40 monoclonal antibody, CHIR-12.12, using XenoMouse® mice (Abgenix, Inc). The antibody can mediate anti-tumor activity potentially by at least two mechanisms: CHIR-12.12 can block CD40-ligand mediated survival signals and it can lyse tumor cells by antibody-dependent cellular cytotoxicity (ADCC). We have previously reported that CHIR-12.12 mediates stronger killing of CD40- and CD20-expressing lymphoma cells than rituximab by ADCC in vitro and significantly inhibits the growth of both rituximab-responsive and rituximab-resistant human lymphoma xenografts in vivo. In this study, we examined in vitro and in vivo efficacy of CHIR-12.12 against human multiple myeloma. The human MM cell line IM-9, which expresses both CD40 and CD20, the target antigen for CHIR-12.12 and rituximab respectively was used for the study. CHIR-12.12 induced lysis of target tumor cells by ADCC in a dose dependent manner reaching maximum cell lysis at 0.1ug/ml concentration. The maximum specific lysis of IM-9 cells by CHIR-12.12 was greater than the lysis induced by rituximab (64% vs 45 %, n=3, p<0.01). In addition, the EC50 of CHIR-12.12 was on average 5.9 picomolar, which was 10-fold lower than the EC50 of rituximab. Greater ADCC by CHIR-12.12 was not due to higher density of CD40 molecules on the target tumor cells compared to CD20 molecules. IM-9 cells expressed 35590 ±8858 CD40 molecules compared to 93783 ± 2247 CD20 molecules. The in vivo CHIR-12.12 efficacy was then evaluated in IM-9 xenograft model. In an un-staged conditional survival model, where treatment began one day after intravenous inoculation of IM-9 tumor cells, CHIR-12.12 significantly prolonged the survival of tumor-bearing mice in a dose-dependent manner with 60% survival in the 0.1 mg/kg CHIR-12.12 treated group and 80% survival in the 1 and 10 mg/kg groups respectively on day 56 (Log Rank Test: P<0.01 and P<0.001, respectively). All animals in the control IgG1 and bortezomib treated groups were terminated between day 18 and day 26 due to severe disease related to tumor development (i.e., hind limb paralysis and significant body weight loss). In a staged subcutaneous model, where treatment began once the tumor volume was 150–200mm3, CHIR-12.12 administered weekly at 0.1, 1 and 10 mg/kg significantly inhibited tumor growth with a tumor volume reduction of 17% (P>0.05), 34% (P<0.01) and 44% (P<0.001) respectively. Bortezomib, when tested at 0.5 mg/kg twice a week did not inhibit tumor growth. At the maximally tolerated dose (MTD) of 1 mg/kg twice a week, bortezomib inhibited tumor growth by 30% (P<0.01). Taken together, these data demonstrate that the anti-CD40 mAb CHIR-12.12 has potent activity against human multiple myeloma in vitro and xenograft models in vivo.

CD40 and CD20 are expressed in several B-cell malignancies and represent attractive targets for therapeutic intervention. The anti-CD20 monoclonal antibody, rituximab, is an approved drug for the treatment of non-Hodgkin’s lymphomas. however, the existence of patients with rituximab-resistant disease limits its clinical utility. We have previously reported that the novel, highly potent, fully human antagonistic anti-CD40 monoclonal antibody, CHIR-12.12, generated from XenoMouse® mice (Abgenix, Inc) has greater anti-tumor activity than rituximab in both rituximab-responsive and rituximab-resistant human NHL models. In this study, we evaluated the potential therapeutic application of combining CHIR-12.12 and rituximab for the treatment of NHL. Namalwa is a Burkitt’s lymphoma cell line that gives rise to aggressive rituximab-resistant tumors when implanted in nude mice. Direct treatment of these tumor cells with CHIR-12.12 or rituximab in culture does not affect cell growth when compared to treatment with an isotype control antibody. Although Namalwa cells express more CD20 than CD40 (average of 10,059 CD20 and 3,138 CD40 molecules per cell respectively, P=0.05), when tested for in vitro ADCC killing using human NK cells as effectors, CHIR-12.12 mediated stronger target cell lysis than rituximab (31.43% vs. 14.15%, P<0.0001). Adding CHIR-12.12 and rituximab together did not enhance the in vitro ADCC killing. When CHIR-12.12 and rituximab were tested in a subcutaneous Namalwa xenograft model, CHIR-12.12 alone caused 60% tumor growth inhibition (P=0.028) whereas rituximab alone at 10 and 20 mg/kg did not inhibit Namalwa tumor growth. When tumor-bearing mice were administered rituximab at 10 mg/kg plus CHIR-12.12 at 5 or 10 mg/kg, synergistic anti-tumor activity was observed in a CHIR-12.12 dose-dependent manner. The mean tumor volume reduction in combination groups is 77% with CHIR-12.12 at 5 mg/kg (P=0.0037) and 83% with CHIR-12.12 at 10 mg/kg (P=0.0018), respectively. The potential interaction between CD40 and CD20 molecules was evaluated in vitro by treating the tumor cells with CHIR-12.12 and assessing the change in CD20 expression and vice versa. The result showed no augmentation of one antigen expression by treating the tumor cells with the other antibody. The mechanism of anti-tumor synergy observed in this combination is under evaluation. Taken together, these data suggest that the combination therapy with anti-CD40 CHIR-12.12 and rituximab has the potential to improve patient outcome in B-cell malignancies co-expressing CD20 and CD40 antigens and support the further development of CHIR-12.12 antibody for treatment of B-cell malignancies.