ALBERT

Description: Introduction: Chimeric Antigen Receptor (CAR) based cellular-immunotherapies have demonstrated significant clinical efficacy in haematological malignancies. However, the progress of cellular-immunotherapy for the treatment of Acute Myeloid Leukaemia (AML) has failed to gain momentum due to the lack of targetable tumour specific antigens. CD38 is a transmembrane glycoprotein expressed in lymphoid and myeloid cells with high expression in plasma B-cells, and is a well validated target for anti-CD38 therapy in Myeloma. A recent study has furthermore shown that a proportion of AML patients express CD38 on their leukemic blasts. TNF-related apoptosis-inducing ligand (TRAIL) receptor DR4 is another targetable antigen which has been shown to be expressed in 70% of AML patients. In this study, we investigate the therapeutic efficacy of "affinity-optimized" variant(s) of CD38 CAR and membrane bound TRAIL on NK-cell based platforms which can target AML blasts with high expression of CD38 (CD38high AML). The CAR variant is a CAR which binds with lower affinity to CD38 expressed on healthy immune cells such as CD38positive NK cells, while targeting CD38high AML. The membrane bound TRAIL variant (TRAIL4c9) is a mutant which binds with higher affinity to TRAIL-DR4 on AML cells, whilst avoiding binding to decoy receptors. We hypothesize that genetically modifying NK cells to express "affinity optimized" CD38 CARand/or TRAIL4c9 can effectively eliminate CD38high AML cells. Methods: AML cell lines THP-1, U937, and KG1a were immunophenotyped for CD38 and TRAIL-DR4 expression. Retrovirally transduced CD38 CAR-KHYG1 NK cells were used as immune effector cells and were co-cultured with AML cell lines in cytotoxicity assays. CD38low AML cell line KG1a was pre-treated with 10nM all-trans-retinoic acid (ATRA) to upregulate CD38 expression and were subsequently co-cultured with CD38 CAR-KHYG1 in cytotoxicity assays. CD38 CAR-KHYG1 was also co-cultured with n=4 patient derived AML cells in cytotoxicity assays. Using Maxcyte GT electroporation system primary donor derived IL-2 activated NK cells were either mock electroporated, or electroporated with TRAIL4c9 m-RNA orCD38 CAR m-RNA and subsequently co-cultured with THP-1 or ATRA pre-treated KG1a in a cytotoxicity assay. Expression of pro-apoptotic, anti-apoptotic and ligands for checkpoint inhibitory receptors was analysed by immunoblotting or flowcytometry. Results: Based on immunophenotyping, we classified AML cell lines as CD38high (THP-1), CD38moderate (U937) and CD38low (KG1a). CD38 CAR-KHYG1 was significantly more cytotoxic than MOCK KHYG1 against CD38high THP-1, at E:T ratios of 2.5:1, 5:1 and 10:1. CD38 CAR-KHYG1 were also more cytotoxic than MOCK KHYG1 against CD38moderate U937 at multiple E:T ratios; albeit the increase in cytotoxicity was at a much lower level in comparison to THP-1 (Fig 1a). Pre-treatment of CD38low KG1a cells with 10nM ATRA upregulated the cell surface expression of CD38, which were subsequently eliminated by CD38 CAR KHYG1 at E:T ratios of 2.5:1, 5:1 and 10:1. KG1a was intrinsically resistant to NK cells as compared to THP-1 and U937 (Fig 1b). This could partly be explained by the high intracellular expression of Bcl-xL, and higher cell surface expression of Nectin-1 and Sialic acid which are the ligands for checkpoint inhibitory receptors CD96 and Siglec-7/9 respectively on NK cell (Fig 1c). CD38 CAR-KHYG1 mounted a potent cytotoxic response against primary CD45intermediate AML blasts (n=4 patients) at multiple E:T ratios, and the extent of CAR induced cytotoxicity correlated with the cell surface CD38 expression on the primary AML blasts (R2=0.87) (Fig 1d,e). TRAIL4c9 or CD38 CAR m-RNA electroporated primary donor-derived NK cells were also potent in eliminating THP-1 and ATRA pre-treated KG1a at multiple E:T ratios (Fig 1f). This demonstrates the potential of therapeutically treating AML patients, with high CD38 expression, with a combination of NK cells expressing "affinity-optimized" CD38 CAR and membrane bound TRAIL variant. Conclusion: The study demonstrates the therapeutic potential of an "affinity-optimized" CD38 CAR NK cell-based therapy, which can potentially be combined with membrane bound TRAIL expressing NK cells to target CD38high AML. In patients with CD38low expressing AML blasts, patients could be pre-treated with ATRA followed by the combination therapy of CD38 CAR and TRAIL expressing NK cells. Disclosures Stikvoort: Onkimmune Ltd., Ireland: Research Funding. Kirkham-McCarthy:Onkimmune Ltd., Ireland: Research Funding. Van De Donk:Janssen Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees, Research Funding; Roche: Membership on an entity's Board of Directors or advisory committees; AMGEN: Membership on an entity's Board of Directors or advisory committees, Research Funding; Celgene Corporation: Membership on an entity's Board of Directors or advisory committees, Research Funding; Bristol-Myers Squibb: Membership on an entity's Board of Directors or advisory committees, Research Funding; Bayer: Membership on an entity's Board of Directors or advisory committees; Servier: Membership on an entity's Board of Directors or advisory committees; Takeda: Membership on an entity's Board of Directors or advisory committees. Mutis:Celgene: Research Funding; Janssen Pharmaceuticals: Research Funding; Amgen: Research Funding; BMS: Research Funding; Novartis: Research Funding; Aduro: Research Funding; Onkimmune: Research Funding. Sarkar:Onkimmune: Research Funding. O'Dwyer:Onkimmune: Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Research Funding; Janssen: Membership on an entity's Board of Directors or advisory committees, Research Funding; GlycoMimetics Inc: Research Funding; AbbVie: Consultancy; BMS: Research Funding.

Description: Introduction Platelets have been implicated in promoting metastasis in a number of cancers. In many solid tumours including colon and ovarian cancer, thrombocytosis has been demonstrated as an adverse factor for disease metastasis and survival. A number of mechanisms have been proposed including platelet "cloaking" of tumour cells to evade recognition by immune cells, production of cytokines which promote tumour growth and the role of platelets in enabling tethering and migration of metastatic cells. In vitro studies have demonstrated the dependence on platelets for successful solid tumour metastasis and implicated the role of P-Selectin in tumour adhesion to platelets. There have been no published studies demonstrating the interaction of platelets with Myeloma cells and their role in enabling Multiple Myeloma (MM) cell metastasis. We sought to determine in vitro whether platelets interacted significantly with MM cell lines and whether they preferentially bind to specific sub-populations of MM cells. We also sought to determine whether this interaction is mediated by P-Selectin and whether "platelet coating" can protect MM cells from Natural Killer (NK)-mediated cell death. Methods Platelet rich plasma (PRP) sample was isolated from freshly drawn peripheral blood samples from healthy participants by FICOLL gradient separation. All samples were collected with informed consent and in accordance with the declaration of Helsinki. The MM1S cell line and its derivative HECA-452 MM1S, which is enriched for the sialofucosylated structure sialyl Lewis X (SLex), were used for this study. Cells were co-cultured with PRP at a ratio of 50 platelets/ 1 MM cell, washed and stained with CD41, CD138, and HECA-452. FACS was used to assess degree of platelet/MM cell binding. Platelet/MM cell interaction was further confirmed using Image Stream. In some experiments, a P-Selectin blocking antibody was used to determine the dependency of platelet/MM cell binding on P-Selectin. For functional studies, MM cells were co-cultured with PRP at different platelet/cell ratios (50:1, 100:1 and 500:1) for 24 h and then incubated with either the NK cell line KHYG-1 or freshly isolated autologous NK cells. Cytotoxicity assays were performed after 24 h by FACS using Propidium Iodide (PI) staining. NK cells were discriminated from MM cells by staining them with cell trace violet dye. Results Platelets displayed preferential binding to the HECA-452 MM1S cells compared to parental cells, suggesting a requirement for sialofucosylated structures for efficient binding. The interaction between platelets and HECA-452 MM1S cells occurred as early as 30 min and remained stable at 24 h post incubation. Moreover, platelets bound to the surface of MM cells, as demonstrated by Image Stream (Figure 1). This binding could be significantly blocked by a P-Selectin blocking antibody, indicating a dependence upon P-Selectin (P=0.0043 at 50:1; P=0.0027 at 100:1; P=0.0163 at 500:1 ratio). Importantly, co-incubation of platelets and MM cells at 500:1 ratio significantly reduced KHYG1-mediated cytotoxicity on HECA-452 MM1S cells at 24 h post incubation (P=0.0078 at 0.25:1 and P=0.0072 at 0.5:1 ratio). Platelets could also protect HECA-452 MM1S cells from primary NK-mediated cytotoxicity, although this didn't quite reach statistical significance. Conclusions We demonstrate for the first time a physical interaction between MM cells and platelets. Efficient binding requires sialofucosylated structures, as shown by almost exclusive binding of platelets to the HECA-452 MM1S compared to parental cells. Moreover, platelets were able to protect HECA-452 MM1S cells from NK-mediated cell death, suggesting that in vivo these cells may escape the immune system and promote MM spreading and metastasis. Notably, platelet-MM cell binding was reverted by a P-Selectin blocking antibody, suggesting that this interaction can be therapeutically targeted in order to restrict MM metastasis and re-sensitize MM cells to NK cells. Disclosures O'Dwyer: Abbvie: Membership on an entity's Board of Directors or advisory committees; Glycomimetics: Research Funding; BMS: Research Funding; Janssen: Membership on an entity's Board of Directors or advisory committees, Research Funding; Onkimmune: Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Research Funding; Celgene: Research Funding.

Description: Introduction Multiple myeloma (MM) is a cancer of clonal plasma cells that hijack the bone marrow (BM) niche to create a drug resistant, incurable malignancy. Aberrant sialylation has been linked to immune cell evasion, drug resistance, and metastasis in cancer; indeed sialyltransferases, including ST3GAL1, ST3GAL4 and ST3GAL6, are aberrantly expressed in many cancers (Glavey et al., 2014). We have previously shown that targeting ST3GAL6 in MM cells inhibits their ability to extravasate and colonize the BM in mouse models (Glavey et al., 2014). Moreover, we also showed that a subpopulation of MM cells expresses functional E-Selectin ligands which, upon expansion, gives rise to a more aggressive disease and resistance to bortezomib in mice (Natoni et al., 2017). Based off these findings, we herein investigated whether inhibiting sialylation in E-selectin-enriched MM cells with 3Fax-Neu5Ac, a small molecule sialyltransferase inhibitor, could alter the ability of these cells to home in the BM and restore bortezomib sensitivity in vivo. We hypothesized that inhibiting homing of MM cells to the BM will improve survival and that co-treatment with bortezomib and 3Fax-Neu5Ac will have a synergistic effect. Methods E-selectin ligands enriched MM1S cells (either positive or negative for GFP/Luciferase) were derived from parental cells by cell sorting using the HECA-452 antibody, which recognize sialofucosylated E-selectin ligands. We then determined the 3Fax-Neu5Ac dose and exposure times needed to decrease sialylation on these MM cells without causing toxicity. HECA-452-enriched MM1S cells were pretreated with 3Fax-Neu5Ac or vehicle for 7 days before being injected into SCID-beige mice and then treated with vehicle or bortezomib (0.3 mg/kg twice a week). Mice were analyzed via bioluminescence imaging (BLI) to monitor tumor progression and weighed twice a week. Mice were euthanized when they began to show paralysis under our IACUC protocol. 3Fax-Neu5Ac pretreated HECA-452 MM1S cells were also tested in vitro for their ability to adhere and roll on VCAM-1, MAdCAM-1 and E-Selectin under shear stress and to respond to bortezomib in co-culture with HS5 cells. Results Treatment of HECA-452 MM1S cells with 3Fax-Neu5Ac, at 300 μM for 7 days significantly reduced sialylation on these cells. Importantly, reducing sialylation with 3Fax-Neu5AC reduced tumor burden and increased survival, although this did not reach significance for survival (Figure 1A). Both vehicle- and 3Fax-Neu5Ac-treated cells significantly responded to bortezomib in the first 5 weeks of the in vivo study (Figure 1B). However, the HECA-452 MM1S cells did not show increased survival when treated with bortezomib suggesting an acquired mechanism of resistance in vivo. Importantly, pretreatment of the HECA-452 MM1S with 3Fax-Neu5Ac could improve survival of these mice preventing bortezomib resistance. In vitro, the HS5 stromal cells protected the HECA-452 MM1S cells from bortezomib and pretreatment with 3Fax-Neu5Ac partially reverted this protection. Moreover, the HECA-452 MM1S cells pretreated with 3Fax-Neu5Ac displayed reduced adhesion on MAdCAM-1 and E-selectin. Conclusions Sialylation plays an instrumental role in bone homing, BM colonization, and drug resistance of MM cells. Pretreatment of HECA-452 MM1S cells with 3Fax-Neu5Ac decreased their sialylation, restored sensitivity to bortezomib in vivo and prolonged survival in mice. This is likely because 3Fax-Neu5Ac pretreatment has multiple effects on MM cells including reducing cell adhesion mediated-drug resistance and adhesion to key molecules involved in BM homing such as MAdCAM-1 and E-selectin. The reduced adhesion on E-selectin is most likely due to the disruption of E-selectin ligands on the surface of MM cells as they require Sialyl Lewis X to function. Notably, we also found that de-sialylation impairs adhesion on MAdCAM-1 (3Fax-Neu5Ac vs DMSO P=0.038) which, together with E-selectin, is another critical BM homing receptor. This data suggests for the first time that sialylation may controls the affinity of integrin α4β7 and its counter-receptor MAdCAM-1. In turn, this would reduce BM homing and increase MM cells in the circulation were they are more prone to the cytotoxic effects of bortezomib. This study supports the importance of targeting sialylation in MM and provides a strong rationale for further clinical translation of this novel approach. Disclosures O'Dwyer: Glycomimetics: Research Funding; Celgene: Research Funding; BMS: Research Funding; Abbvie: Membership on an entity's Board of Directors or advisory committees; Janssen: Membership on an entity's Board of Directors or advisory committees, Research Funding; Onkimmune: Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Research Funding.