Bispecific Constructs Research Articles

Abstract 3820: Novel PD-1/GITRL bispecific fusion protein delivering concurrent T cell co-stimulation and checkpoint pathway inhibition

Abstract Glucocorticoid-induced tumor necrosis factor receptor-related protein (GITR) is a co-stimulatory receptor, and programmed cell death protein 1 (PD-1) is an important co-inhibitory receptor involved in controlling T-cell activation and function. Preclinical studies using murine cross-reactive molecules demonstrate that the combination of an αPD-1 monoclonal antibody (mAb) with a GITR-ligand (GITRL) fusion protein can elicit synergistic antitumor activity and long-term survival benefit in syngeneic mouse tumor models. Here we describe the activity of two bispecific fusion proteins, each comprising two GITR ligand trimers fused to the Fc region of an αPD1 mAb, in an IgG1 (MEDI5771) or IgG4P (MEDI3387) format. We show that these bispecific molecules are able to concurrently engage both targets and are effective in cell-based PD-1 and GITR reporter assays, and in a range of human primary T-cell assays. Furthermore, the PD-1/GITRL FP bispecific constructs elicit T-cell responses in vitro that are at least equivalent to a combination of a) the parental molecules and b) MEDI1873 (a GITR agonist currently undergoing clinical evaluation) and an αPD-1 mAb. Moreover, using surrogate molecules we demonstrate that IV administration leads to dose-dependent pharmacodynamic responses on CD4+ and CD8+ T cells for both isotypes of the bispecific molecule. These data demonstrate that the antitumor activity for these novel molecules is equivalent to the combination of GITRL-FP and PD-1 mAb in the B16 model. In summary, the simultaneous dual targeting of PD-1 and GITR pathways mediated by these two novel bispecific molecules induces robust T-cell activity that is at least equivalent to a combination of molecules targeting these pathways, thereby demonstrating strong potential as a novel anticancer therapy. Citation Format: Simon J. Dovedi, Jonathan Seaman, Ines Osma-Garcia, Katharina Deschler, Edmund Poon, Ronald Herbst, Robert Wilkinson. Novel PD-1/GITRL bispecific fusion protein delivering concurrent T cell co-stimulation and checkpoint pathway inhibition [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3820.

Abstract 5624: Development of novel bispecific immune modulating antibodies

Abstract The initial success with checkpoint blocking antibodies has evolved into a tremendous effort in combination therapy to further improve outcome in responsive cancers and to overcome resistance. There is strong rationale for this approach, as the cancer-immunity cycle has many points for intervention, and the immune system has developed multiple pathways to the same end. The use of bispecific antibodies provides opportunities to engage two pathways simultaneously that can result in greater efficacy, improved safety profile, and simplified development compared to separately administered antibodies. To generate bispecific antibodies that blocked the PD-1 signaling pathway, we developed and characterized fully human PD-L1 specific antibodies generated using human Ig transgenic mice. A lead antibody was chosen for its potent activity in functional assays and good manufacturing properties. Our initial bispecific construct couples our CD27 agonist antibody with the PD-L1 antibody (αCD27xαPD-L1) as we have previously shown that this combination led to improved therapeutic activity in the BCL1 disseminated lymphoma model. We found that the αCD27xαPD-L1 was more potent in CD27 costimulation assays such as NFkB activation, and T cell cytokine production compared to the parental antibodies. Similarly, αCD27xαPD-L1 was more efficacious in vivo as measured by Ag-specific T cell IFN-γ production and antitumor activity. Importantly, the bispecific antibody was more effective than the combination of anti-CD27 and anti-PD-L1 antibodies in the BCL1 lymphoma model. The enhanced efficacy was attributed to more efficient cross-linking of the bispecific antibody/CD27 receptor, resulting in stronger T cell activation. In addition, we observed that CD27 signaling increased PD-L1 expression on both tumor and immune cells, which may further enhance the activity of the bispecific antibody. The use of bispecific antibodies can improve the efficacy and profile of immunotherapeutic monoclonal antibodies by combining complementary functionalities and enhancing target cross-linking in the relevant tissues. We are currently exploring this approach with multiple antibodies from our pipeline. Citation Format: Laura A. Vitale, Lawrence J. Thomas, Li-Zhen He, Thomas O'Neill, Jenifer Widger, Andrea Crocker, James Testa, Karuna Sundarapandiyan, Eric Forsberg, James Boyer, James Storey, Joel Goldstein, Henry C. Marsh, Tibor Keler. Development of novel bispecific immune modulating antibodies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5624.

Modular, defined assembly of antibody-based constructs using click chemistry

Abstract Immunoglobulin antibodies contribute to a broad range of research assays and also provide the foundation for a diverse variety of biopharmaceutical agents. Here we describe the use of bio-orthogonal click chemistry reactions to synthesize several common antibody-based structures and demonstrate their use in cell-based applications. First, defined antibody-fluorophore conjugates were synthesized, characterized and functionally tested in receptor occupancy and antibody internalization assays. Next, a double conjugation strategy was employed to synthesize antibody drug conjugates which become brightly fluorescent upon endocytosis. ADC uptake and targeted cell killing were demonstrated using cell line and spheroid models. Then bispecific antibodies were constructed using a click chemistry reaction to create tetravalent heterodimeric antibodies. Structure and function of these bispecific constructs are demonstrated in cell-free and flow cytometry assays.

A bispecific nanobody approach to leverage the potent and widely applicable tumor cytolytic capacity of Vγ9Vδ2-T cells

ABSTRACTThough Vγ9Vδ2-T cells constitute only a small fraction of the total T cell population in human peripheral blood, they play a vital role in tumor defense and are therefore of major interest to explore for cancer immunotherapy. Vγ9Vδ2-T cell-based cancer immunotherapeutic approaches developed so far have been generally well tolerated and were able to induce significant clinical responses. However, overall results were inconsistent, possibly due to the fact that these strategies induced systemic activation of Vγ9Vδ2-T cells without preferential accumulation and targeted activation in the tumor. Here we show that a novel bispecific nanobody-based construct targeting both Vγ9Vδ2-T cells and EGFR induced potent Vγ9Vδ2-T cell activation and subsequent tumor cell lysis both in vitro and in an in vivo mouse xenograft model. Tumor cell lysis was independent of KRAS and BRAF tumor mutation status and common Vγ9Vδ2-T cell receptor sequence variations. In combination with the conserved monomorphic nature of the Vγ9Vδ2-TCR and the facile replacement of the tumor-specific nanobody, this immunotherapeutic approach can be applied to a large group of cancer patients.

Rigidly connected multispecific artificial binders with adjustable geometries

Multivalent binding proteins can gain biological activities beyond what is inherent in the individual binders, by bringing together different target molecules, restricting their conformational flexibility or changing their subcellular localization. In this study, we demonstrate a method to build up rigid multivalent and multispecific scaffolds by exploiting the modular nature of a repeat protein scaffold and avoiding flexible linkers. We use DARPins (Designed Ankyrin Repeat Proteins), synthetic binding proteins based on the Ankyrin-repeat protein scaffold, as binding units. Their ease of in vitro selection, high production yield and stability make them ideal specificity-conferring building blocks for the design of more complex constructs. C- and N-terminal DARPin capping repeats were re-designed to be joined by a shared helix in such a way that rigid connector modules are formed. This allows us to join two or more DARPins in predefined geometries without compromising their binding affinities and specificities. Nine connector modules with distinct geometries were designed; for eight of these we were able to confirm the structure by X-ray crystallography, while only one did not crystallize. The bispecific constructs were all able to bind both target proteins simultaneously.

An FDA oncology analysis of CD3 bispecific constructs and first-in-human dose selection

We retrospectively examined the nonclinical studies conducted with 17 CD3 bispecific constructs in support of first-in-human (FIH) trials in oncology. We also collected information on the design of dose-finding clinical trials. Sponsors have used different MABEL approaches for FIH dose selection. To better assess acceptable approaches, FIH doses were computed from nonclinical studies and compared to the maximum tolerated doses (MTDs) in patients, to the highest human doses (HHDs) when an MTD was not identified, or to the recommended human dose (RHD) for blinatumomab. We concluded that approaches based on receptor occupancy, highest non-severely toxic dose, or no-observed adverse effect level are not acceptable for selecting the FIH dose as they resulted in doses close to or above the MTDs, HHDs, or the RHD. A FIH dose corresponding to 10%–30% pharmacologic activity (PA) was an acceptable approach. A FIH dose corresponding to 50% PA was acceptable for all except one construct, potentially due to its biological or structural properties. The most common toxicities in animals and patients were those related to cytokine release. Doses were better tolerated when intra-animal or intra-patient dose escalation was used. Exposing naïve patients to an MTD achieved with intra-patient dose escalation design may be unsafe.

Abstract 1700: Radiation-induced vegf-targeted 4-1bb costimulation enhances immune control of tumor growth

Abstract Radiotherapy (RT)-potentiated immunomodulation is an emerging field in solid tumor malignancies. RT induces peripheral antitumor immunity that can lead to the reduction of distant, non-irradiated tumor lesions (abscopal response). This has provided rationale for current early phase trials combining existing checkpoint blockade and co-stimulatory mAbs with stereotactic RT. Objective responses remain suboptimal with checkpoint blockade monotherapy, and dose limiting toxicity is widely observed with these modalities. We hypothesize that targeting immunotherapy to the tumor using oligonucleotide aptamers that bind to RT induced tumor stress products will significantly enhance the therapeutic index. We used the aptamer platform to generate a bi-specific construct containing a costimulatory aptamer ligand specific to murine 4-1BB which was conjugated to an aptamer specific to products secreted into the tumor stroma (VEGF or osteopontin). Tumor targeting, local control, abscopal responses and toxicity were compared to unconjugated 4-1BB-VEGF constructs, or mAB based therapies, when used in combination with RT in murine subcutaneous and autochthonous MCA tumor models. RT induced VEGF upregulation was demonstrated in null/low VEGF secreting tumors that led to preferential accumulation of VEGF-41BB aptamers in the irradiated tumor.12Gy x1 elicited optimal tumor targeting of bi-specific VEGF-4-1BB conjugate. Significant anti-tumor immune responses were observed in s.c. murine tumor models. This conjugate showed dramatic reduction in systemic toxicity compared to the equivalent gold standard 4-1BB mAb. RT-VEGF-4-1BB treatment significantly improved local control, overall survival and induced potent abscopal responses compared to unconjugated and non-irradiated controls in the Balb/c 4T1 and MCA murine tumor models. Furthermore, RT-VEGF-4-1BB treatment showed similar anti-tumor effects as CTLA-4 mAB with significantly less systemic toxicity. This approach increased intra-tumoral CD4+ and CD8+ T cells, the CD8+ to Treg ratio, and induced significantly more tumor cell death. In conclusion, this strategy has been shown to induce potent anti-tumor immune responses in relevant murine tumor models and significantly improves the therapeutic index over non-targeted mABs. Furthermore, radiotherapy can expand the scope of tumor targeted immune modulation to virtually all solid malignancies and induce potent abscopal responses. These results provide the rationale for developing clinical grade bispecific aptamers for future testing. Citation Format: Brett Schrand, Bhavna Verma, Agata Levay, Shradha Patel, Iris Castro, Ana Paula Benaduce, Randall Brenneman, Oliver Umland, Hideo Yagita, Eli Gilboa, Adrian Ishkanian. Radiation-induced vegf-targeted 4-1bb costimulation enhances immune control of tumor growth [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1700. doi:10.1158/1538-7445.AM2017-1700

Abstract No. 129 - Development of a novel bispecific antibody for targeted immunotherapy of hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is an ideal target for novel immunotherapies due to its otherwise immunosuppressive intratumoral environment. While the efficacy of bispecific antibodies for immunotherapy is well established for hematological malignancies, their limited circulatory half-lives make them challenging to use for solid tumors. Intratumoral delivery can bypass this issue, potentially making transcatheter-administered bispecific antibody therapy an exciting new paradigm for HCC treatment. We have developed a bispecific antibody that can simultaneously bind to both HCC cells and cytotoxic T cells in order to enhance T-cell mediated tumor lysis. The bispecific antibody targeting the HCC surface antigen glypican-3 (GPC-3) and the T-cell antigen CD-3 was designed as a fusion protein (IDT DNA) and expressed recombinantly in E. coli expression system (Origami Cells) according to previous reports. [1] Specific binding of the antibody against both HCC cells (HepG2) and T cells (Jurkat) was confirmed using cell-surface ELISA assays. [2] Bispecific-triggered activation of T cells (donor CD4 cells) was assessed in vitro using IFN-gamma release ELISA. The bispecific antibody is designed as a 33kD fusion protein consisting of a GPC-3 binding peptide (RLNVGGTYFLTTRQ), previously shown to bind specifically to HCC cells, connected to a CD3-binding single variable fragment domain (ScFV), derived from an OKT3 antibody. [3] The antibody was successfully expressed and purified. It binds to both HCC cells (31.6nM, SD 0.008nM) and T cells (13.2nM, SD 0.006nM) with affinities that are comparable to values reported for other bispecific constructs. [4] The binding to both cell lines were specific and can be inhibited by using excess of either pure GPC-3 binding peptide or full-length OKT3 antibody. The activity of the bispecific antibody is demonstrated by its ability to activate T cells in vitro, resulting in robust IFN-gamma release as measured using ELISA (0.196 Bispecific vs 0.008 Cell Only Control, p = 0.0003). A novel bispecific antibody targeting GPC-3 and CD-3 was successfully synthesized and validated in vitro demonstrating the potential for targeted immunotherapy of HCC.

Frequency of regulatory T cells determines the outcome of the T-cell-engaging antibody blinatumomab in patients with B-precursor ALL

Blinatumomab can induce a complete haematological remission in patients in 46.6% with relapsed/refractory B-precursor acute lymphoblastic leukemia (r/r ALL) resulting in a survival benefit when compared with chemotherapy. Only bone marrow blast counts before therapy have shown a weak prediction of response. Here we investigated the role of regulatory T cells (Tregs), measured by CD4/CD25/FOXP3 expression, in predicting the outcome of immunotherapy with the CD19-directed bispecific T-cell engager construct blinatumomab. Blinatumomab responders (n=22) had an average of 4.82% Tregs (confidence interval (CI): 1.79–8.34%) in the peripheral blood, whereas non-responders (n=20) demonstrated 10.25% Tregs (CI: 3.36–65.9%). All other tested markers showed either no prediction value or an inferior prediction level including blast BM counts and the classical enzyme marker lactate dehydrogenase. With a cutoff of 8.525%, Treg enumeration can identify 100% of all blinatumomab responders and exclude 70% of the non-responders. The effect is facilitated by blinatumomab-activated Tregs, leading to interleukin-10 production, resulting in suppression of T-cell proliferation and reduced CD8-mediated lysis of ALL cells. Proliferation of patients' T cells can be restored by upfront removal of Tregs. Thus, enumeration of Treg identifies r/r ALL patients with a high response rate to blinatumomab. Therapeutic removal of Tregs may convert blinatumomab non-responders to responders.

Redirecting T cells to eradicate B-cell acute lymphoblastic leukemia: bispecific T-cell engagers and chimeric antigen receptors

Recent advances in antibody technology to harness T cells for cancer immunotherapy, particularly in the difficult-to-treat setting of relapsed/refractory acute lymphoblastic leukemia (r/r ALL), have led to innovative methods for directing cytotoxic T cells to specific surface antigens on cancer cells. One approach involves administration of soluble bispecific (or dual-affinity) antibody-based constructs that temporarily bridge T cells and cancer cells. Another approach infuses ex vivo-engineered T cells that express a surface plasma membrane-inserted antibody construct called a chimeric antigen receptor (CAR). Both bispecific antibodies and CARs circumvent natural target cell recognition by creating a physical connection between cytotoxic T cells and target cancer cells to activate a cytolysis signaling pathway; this connection allows essentially all cytotoxic T cells in a patient to be engaged because typical tumor cell resistance mechanisms (such as T-cell receptor specificity, antigen processing and presentation, and major histocompatibility complex context) are bypassed. Both the bispecific T-cell engager (BiTE) antibody construct blinatumomab and CD19-CARs are immunotherapies that have yielded encouraging remission rates in CD19-positive r/r ALL, suggesting that they might serve as definitive treatments or bridging therapies to allogeneic hematopoietic cell transplantation. With the introduction of these immunotherapies, new challenges arise related to unique toxicities and distinctive pathways of resistance. An increasing body of knowledge is being accumulated on how to predict, prevent, and manage such toxicities, which will help to better stratify patient risk and tailor treatments to minimize severe adverse events. A deeper understanding of the precise mechanisms of action and immune resistance, interaction with other novel agents in potential combinations, and optimization in the manufacturing process will help to advance immunotherapy outcomes in the r/r ALL setting.

CD3/Bars: A Novel Bispecific Format for the Treatment of B-Cell Lymphomas

BackgroundAutoantigens are suspected to play an important role in the pathogenesis of different types of B cell neoplasia. Suggestive of this hypothesis is the restricted usage of a stereotyped IgHv repertoire in CLL, MCL and DLBCL. Further evidence supporting this notion is the identification of specific autoantigens as the target of the B-cell receptor from malignant lymphomas and myelomas, such as paratarg-7 as antigenic target for 15% of paraproteins of MGUS and MM patients. ARS2 was previously identified as the autoantigenic target for the B-cell receptor of approximately 25% of DLBCLs of the ABC type, here termed ARS2 reactive lymphomas. We had recently shown that the B-cell receptor antigens can be used to target B-cell lymphoma cells in vitro and in vivo in an approach designated as BARs (B-cell receptor antigens for reverse targeting), the first therapeutic strategy in oncology with absolute and exclusive specificity for the malignant clone. To test whether BARs can substitute the B-cell binding antibody (e.g. CD19) in T-cell engaging bispecific antibodies, we designed a bispecific CD3/BAR product consisting of a recombinant single chain fragment (scFv) against CD3 linked to ARS2 (CD3-ARS2). One arm of this construct should engage the T cell co-receptor CD3 of human T cells, and the other site should bind to the B cell receptor of ARS2 reactive lymphomas thus specifically redirecting and activating T cells against lymphoma cells.Material and methodsVL and VH genes of the CD3 - OKT3 hybridoma and the DNA sequence of the 33 amino acids containing the B-cell receptor binding epitope of ARS2 were cloned into a pcDNA 3.1 vector by standard cloning techniques. VH and VL were linked by a glycine-serine linker, as was VL to the ARS2 epitope resulting in VH-(Gly4Ser1)n-VL-(Gly4Ser1)n-ARS2 peptide chain. The final cloning product was transfected in HEK cells for production of the bispecific construct. Binding capacity to lymphoma cell lines (OCI-Ly3, U2932, HBL-1) and PBMCs was assessed by flow cytometry. Western blot analysis was used for detection of CD3-ARS2 after incubation with the monoclonal anti-His Tag antibody. Cytotoxicity was evaluated by LDH release assay.ResultsThe CD3 - ARS2 bispecific construct bound to CD3 on T cells and the B-cell receptor of ARS2 reactive lymphoma cells. CD3/ARS2 induced rapid cytotoxicity exclusively in ARS2 reactive lymphoma cell lines at concentrations as low as 250 ng/ml with an effector - target ratio of 10:1. Specific T-cell mediated cytotoxicity reached 40% after 4 hours. Lymphoma cell lines with BCRs of a specificity other than ARS2 were not affected.ConclusionThe CD3/BAR construct is a novel therapeutic principle for the treatment of B-cell lymphomas, suggesting that BARs might also be useful as part of CAR-T cells. Compared to CD3/CD19 bispecific antibodies the CD3/BARs are exclusively cytotoxic against the malignant clone and spare normal B-cells. This should considerably reduce the acute toxicity of T-cell engaging bispecific constructs and circumvent long-term B cell depletion. Experiments comparing the cytotoxic capacity of CD3/BARs with CD3/CD19 bispecific antibodies are underway, as are analyses evaluating possible synergisms of these constructs. DisclosuresNo relevant conflicts of interest to declare.

Abstract B088: Generation of immuno-modulatory receptor binding bispecific antibodies to modulate tumor immunity

Abstract Introduction: Monoclonal antibodies blocking CTLA-4 or PD-1 ligand interaction have been shown to induce durable clinical responses in a subset of melanoma and NSCLC patients. The clinical efficacy of these antibodies has been linked to their ability to restore robust anti-tumor immune responses. Based on clinical and preclinical data it is expected that combinations of antibodies blocking inhibitory and/or costimulatory molecules could benefit patients that do not respond to existing immunotherapies. However dual blockade of immuno-modulatory receptors (iMODs) has also been shown to increase immune-related toxicity as has been shown for a combination of CTLA-4 and PD-1 antibodies. Bispecific antibodies are a novel therapeutic modality capable of specifically binding two different targets. As such they may be ideally suited to address dual blockade of iMODs. Previous work on other cell surface targets has shown that bispecific antibodies can exert functional activities that cannot be reproduced by monoclonal antibody combinations. In addition, bispecific antibodies that more selectively target specific cell populations, based on the co-expression of the two targets, could have reduced safety liabilities in patients. We report here the generation and characterization of common light chain (cLC) Fab panels against PD-1, PD-L1, LAG-3, CD137, OX40 and TIM-3 for combination in a full length IgG1 bispecific antibody format and present the functional evaluation of some of these combinations in human in vitro functional assays. Methods: Humanized cLC mice (MeMo®) were immunized with recombinant protein or DNA, and materials harvested from these mice were used to generate Fab regions against these antigens using phage display. Furthermore, large and diverse synthetic cLC Fab-phagemid libraries were used to select iMOD specific Fabs. The resulting cLC Fabs were cloned into bispecific antibody expression constructs for further testing. Recombinant proteins (both wild type and domain swapped orthologs) were generated for each target and used for binding and competition in both ELISA and FACS. Reporter assays using NFAT or NFkB-luc in Jurkat cells were used for each target to measure functional activity of antibodies. Furthermore, primary human assays using CD3 or SEB activation with cytokine release were used as a read-out of T cell activity. Results: Two parallel strategies were applied to generate panels of cLC Fabs against PD-1, PD-L1, LAG-3, CD137, OX40 and TIM-3; mouse immunization and phage selection yielded >1350 specific Fab arms whereas synthetic library selections yielded >850 Fabs. Initial characterization and binning of representative clones were carried out in bispecific IgG format (iMOD specific Fabs combined with a Tetanus toxoid-specific control Fab arm) allowing for stringent ranking of the iMOD panels in a monovalent format for (domain) specificity, cynomolgus cross-reactivity, affinity, stability, and ligand blocking potency. The resulting characteristics of the identified bins were used to generate diverse bispecific antibody panels that combined different iMOD Fabs. These were first tested in reporter assays to select functional clones, that were then confirmed in a SEB driven T cell proliferation assay. Several bispecific candidates have been identified that show significantly increased cytokine production and other functional parameters compared to benchmark antibodies. Conclusion: Large diverse Fab panels binding six different iMODs were generated for use in bispecific antibody combinations. Characterization and screening of large panels of bispecific IgG has been carried out in relevant functional assays. Results from the screening of bispecific antibodies targeting iMOD combinations support the concept of empirical screening to identify potent immunomodulatory combinations. Citation Format: Rinse Klooster, Cecile Geuijen, Mark Throsby. Generation of immuno-modulatory receptor binding bispecific antibodies to modulate tumor immunity [abstract]. In: Proceedings of the Second CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; 2016 Sept 25-28; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2016;4(11 Suppl):Abstract nr B088.

Abstract B34: Quantitative systems pharmacology and immunotherapy: accelerating lead generation and optimization of a PD-1 x TIM-3 biotherapeutic in immuno-oncology

Abstract The goal of this collaboration was to provide early quantitative decision making guidance for the project team by developing and interrogating a quantitative systems pharmacology (QSP) model of the co-modulation inhibitory receptors PD-1 and TIM-3 in immuno-oncology. The QSP model was to: (1) provide predictions of the best-in-class profile for a PD-1 and TIM-3 dual antagonist biologic(s), (2) accelerate project timelines, (3) provide biological insights, and (4) reduce experimental costs. The QSP model was based on first principles as a system of elementary mass-action, mechanistic PKPD, ordinary differential equations. The model parameters and reactions were based on biophysics, and are interpretable. The model reactions include protein synthesis and elimination, ligand-receptor and drug-target formation and turnover, and drug administration and first order clearance. There were four versions of the model: PD-1 monospecific, TIM-3 monospecific, PD-1 x TIM-3 bispecific and fixed dose combination (FDC) targeting PD-1 and TIM-3. The monospecific models were then benchmarked against published data such that model parameter values were set to known values and unknown parameters were estimated. Once benchmarked, the FDC and bispecific models were analyzed by systematically investigating how tuning the model parameters (e.g., affinity, avidity, dose, half-life, target expression, etc.) impacted target inhibition, and to simulate patient variability. The model was in good agreement with published clinical data from nivolumab and pembrolizumab, and data from RMT3-23 in the TIM-3 driven mouse model. QSP model analysis predicted: (1) there would be diminishing returns on very tight binding biologics due to Target Mediated Drug Disposition (TMDD) that offsets potency, and (2) there is no advantage between FDC, 2-2 bispecific, and 2-1 bispecific formats, which are predicted to be roughly equivalent. As a result of these analyses, there was a significant reduction in the number of experiments, and acceleration of project timelines by (1) eliminating rounds of affinity maturation, as drug leads were in predicted optimal drug parameter ranges, and (2) eliminating the need to construct and evaluate bi-specific constructs and proceed with FDCs. Citation Format: Joshua F. Apgar, Jamie Wong, Ryan Phennicie, Mike Briskin, John M. Burke. Quantitative systems pharmacology and immunotherapy: accelerating lead generation and optimization of a PD-1 x TIM-3 biotherapeutic in immuno-oncology. [abstract]. In: Proceedings of the Fourth AACR International Conference on Frontiers in Basic Cancer Research; 2015 Oct 23-26; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2016;76(3 Suppl):Abstract nr B34.

Generation of camelid VHH bispecific constructs via in-cell intein-mediated protein trans-splicing.

Production of various combinations of bispecific variable domain of heavy chain of heavy chain-only antibody (VHH) constructs to evaluate their therapeutic potential usually requires several gene-engineering steps. Here, we present an alternative method of creating bispecific VHH constructs in vivo through protein trans-splicing (PTS) reaction; this method may reduce the number of gene manipulation steps required. As a proof-of-concept, we constructed a bispecific antibody (bsAb) containing an anti-epidermal growth factor receptor VHH and anti-green fluorescent protein VHH, and we evaluated and confirmed its bispecificity. We also tested antibody labeling by fluorescent protein tagging using the PTS reaction. Compared with the conventional gene construction method, bsAb construction via PTS is a promising alternative approach for generating multiple bsAb combinations.

Overcoming the Constraints of Anti-HIV/CD89 Bispecific Antibodies That Limit Viral Inhibition.

Innovative strategies are necessary to maximize the clinical application of HIV neutralizing antibodies. To this end, bispecific constructs of human antibody F240, reactive with well-conserved gp41 epitope and antibody 14A8, reactive with the IgA receptor (CD89) on effector cells, were constructed. A F240 × 14A8 bispecific single chain variable region (scFv) molecule was constructed by linking two scFvs using a conventional GGGGS linker. Despite immunoreactivity with HIV gp41 and neutrophils, this bispecific scFv failed to inhibit HIV infection. This is in sharp contrast to viral inhibition using a chemical conjugate of the Fab of these two antibodies. Therefore, we constructed two novel Fab-like bispecific antibody molecules centered on fusion of the IgG1 CH1 domain or CH1-hinge domain to the C-terminus of F240scFv and fusion of the kappa chain CL domain to the C-terminus of 14A8scFv. Both Bi-Fab antibodies showed significant ADCVI activity for multiple clade B and clade C isolates by arming the neutrophils to inhibit HIV infection. The approach presented in this study is unique for HIV immunotherapy in that the impetus of neutralization is to arm and mobilize PMN to destroy HIV and HIV infected cells.

Abstract B023: Costimulatory T-cell engagement via a novel bispecific anti-CD137 /anti-HER2 protein based on Anticalin® technology

Abstract Background: CD137 is a potent costimulatory immunoreceptor and a member of the TNF-receptor (TNFR) superfamily. The receptor, also known as 4-1BB, is mainly expressed on activated CD4+ and CD8+ T cells, activated B cells, and natural killer (NK) cells. While multiple lines of evidence show that CD137 is a highly promising therapeutic target, current approaches are not designed to achieve a tumor-target driven activation, which may reduce the available therapeutic window via peripheral T cell activation and toxicity. To overcome this limitation, we applied Anticalin® technology to generate a bispecific protein therapeutic binding to CD137 and a differentially expressed tumor target, HER2. Methods: Anticalin® proteins are 18 kD protein therapeutics derived from human lipocalins which enable straight-forward multimeric drug targeting across several formats. We utilized phage display to generate an Anticalin protein binding to CD137 with high affinity and specificity. The CD137-specific Anticalin protein was genetically fused to a Trastuzumab variant, yielding four different constructs covering a range of distances between the binding sites of the T cell-target and the tumor cell target. To minimize Fc-receptor interaction of the resulting bispecific and concomitant potential toxicity towards CD137-positive cells, the backbone of Trastuzumab was switched from IgG1 to an engineered IgG4 isotype. Results: All four bispecific constructs bound the targets CD137 and HER2 with a nearly identical affinity compared to the parental building blocks, and both targets could be simultaneously bound. Compared to non-engineered Trastuzumab, binding to human receptors FcγRI and FcγRIII was significantly reduced, while binding to the neonatal Fc receptor (FcRn) was retained. Functional activity was demonstrated in human T cell activation assays, and shown to be tumor target (HER2) dependent. Conclusion: We report the first bispecific therapeutic protein that targets the potent costimulatory immunoreceptor CD137 in a tumor-target dependent manner, utilizing HER2 as the tumor target. Compared to currently existing CD137-targeting antibodies, this approach has the potential to provide a more localized activation of the immune system with reduced peripheral toxicity. Bispecific T cell engagers based on CD137 and HER2 may have utility in HER2-positive cancers where there is a significant unmet medical need, such as bladder, ovarian and gastric cancer. Citation Format: Marlon J. Hinner, Rachida-Siham Bel Aiba, Alexander Wiedenmann, Corinna Schlosser, Andrea Allersdorfer, Gabriele Matschiner, Christine Rothe, Ulrich Moebius, Shane A. Olwill. Costimulatory T-cell engagement via a novel bispecific anti-CD137 /anti-HER2 protein based on Anticalin® technology. [abstract]. In: Proceedings of the CRI-CIMT-EATI-AACR Inaugural International Cancer Immunotherapy Conference: Translating Science into Survival; September 16-19, 2015; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2016;4(1 Suppl):Abstract nr B023.

Induction of NK and T Cell Immune Responses Against Leukemia Cells By Bispecific NKG2D-CD16 and -CD3 Fusion Proteins

Monoclonal antibodies (mAbs) have by now become an established tool in therapy of many malignancies. The interaction of a mAb's Fc portion with Fcγ receptors (FcγR) on immune effector cells is important for its efficacy, but often insufficient to potently induce antitumor immunity e.g. due to FcγR polymorphisms. Moreover, Fc parts of mAbs may bind to FcγRs expressed on non-cytotoxic cells (e.g., platelets and B cells) and interact with FcγRs that do not trigger cytotoxicity (e.g. CD16b on granulocytes). These shortcomings can be overcome by novel antibody formats like bispecific antibodies allowing for improved activation of a specific pool of effector cells. Here we report on the development and preclinical characterization of two bispecific fusion proteins that target ligands of the immunoreceptor NKG2D (NKG2DL) which are widely expressed on malignant cells but generally absent on healthy tissue. Our fusion proteins consist of the extracellular domain of NKG2D as targeting moiety for tumor-expressed NKG2DL fused to Fab-fragments of either an agonistic CD3 or CD16 antibody. Specific binding of these NKG2D-CD3 and NKG2D-CD16 constructs was confirmed using NKG2DL-transfectants with regard to their target arm and either NK cells or T cells with regard to their different effector parts. Dose titration assays revealed an increased affinity of NKG2D-CD16 to the FcγR on NK cells as compared to our previously described (e.g., Steinbacher et al, 2014) Fc-optimized NKG2D-IgG1 fusion protein, which was mirrored by a potently increased ability to induce lysis of NKG2DL-tranfectants and NKG2DL-positive primary acute myeloid leukemia (AML) cells by allogeneic NK cells. The novel NKG2D-CD3 construct in turn was found to potently activate allogeneic and autologous CD4+ and CD8+ T cells. Next we comparatively analyzed the efficacy of T cells and NK cells to lyse autologous leukemia cells upon treatment with NKG2D-CD3 and NKG2D-CD16, respectively, by using PBMC of AML patients (blast counts 30-70%) directly ex vivo in long term cytotoxicity assays. NKG2D-CD16 potently induced AML cell lysis, and this was, in line with their proliferative and higher effector potential, by far exceeded upon stimulation of T cells with NKG2D-CD3. Taken together, we here introduce novel "antibody-like" bispecific constructs that take advantage of the highly tumor-restricted expression of NKG2DL and potently activate the reactivity of NK or T cells for immunotherapy of leukemia. DisclosuresNo relevant conflicts of interest to declare.

A Novel Bispecific CD38 Antibody Eradicates Multiple Myeloma in a Mouse Model Following Yttrium-90-DOTA Capture

BACKGROUNDDespite high rates of initial response to immunomodulatory agents and proteasome inhibitors, the vast majority patients living with multiple myeloma (MM) will ultimately die of progressive disease that is a function of persistent MM cell clones. The radiosensitivity of malignant plasma cells is well documented in clinical settings, and we have previously reported the therapeutic efficacy of a streptavidin-biotin (SAB) pretargeted radioimmunotherapy (PRIT) system delivering the β-emitter 90Y to CD38 antigen targets (Green, et al. Cancer Res. 2014). While SAB-PRIT is capable of disease eradication in mouse models, SA immunogenicity and interference by endogenous biotin may complicate clinical translation of these findings into human trials. METHODSTo prospectively address these potential limitations and to further enhance therapeutic efficacy of PRIT, we engineered an anti-human CD38 xanti-chelated radiometal fusion protein (FP) specifically designed to trap second-step radiolabeled ligand (Y-DOTA) using a very high affinity anti-Y-DOTA scFv (C825). The bispecific anti-CD38 FP construct (028Fc-C825) expressed in CHO cells was affinity purified, assessed by gel filtration chromatography and binding specificity to both CD38 and Y-DOTA was confirmed. In vivo experiments showed the absence of toxicity of the bispecific construct at a dose range of 0.14-2.8 nmol per mouse (n=5/group). Biodistribution studies were performed in athymic nude mice (n=5/group) bearing s.c. luciferase transduced NCI-H929Luc human MM xenograft tumors. Animals received 2.8 nmol (420 µg) of either 028-Fc-C825 (anti-CD38 FP) or an irrelevant matched negative control bi-specific FP 2H7-Fc-C825 (anti-CD20 FP) with the same modular IgG-scFv structure (NCI-H929Luc are CD20 negative). The bispecific FP was administered 23 hours prior to a synthetic DOTAY-Dextran clearing agent (CA; 5 µg) and 24 hours prior to administration of trace labeled 90Y-DOTA (2 µg). In therapy studies, athymic nude mice (n=10/group) bearing s.c. NCI-H929Luc human MM xenograft tumors received 2.8 nmol of either 028-Fc-C825 (anti-CD38 FP) or 2H7-Fc-C825 (anti-CD20 FP) 23 hours prior to synthetic DOTAY-Dextran CA and 24 hrs prior to 90Y-DOTA (2 µg; 1200 µCi per mouse). An activity range of 800 µCi to 1200 µCi was previously identified as optimal in a MM xenograft model using the high energy beta particle emitter 90Y (t1/2 = 64 hours) [Green, et al. Cancer Res. 2014]. RESULTSSelective tumor targeting of 90Y ligand to CD38+ xenografts was observed following administration of the bispecific anti-CD38 FP. Blood, tumor and nonspecific organ uptakes of 028-Fc-C825 demonstrated tumor-to-normal organ absorbed ratios that were 15:1; 17:1; 9:1; and 13:1, respectively for blood, lung, liver and kidney; compared to 0.7:1; 1:1; 1:1 and 0.5:1, respectively, in mice pretargeted with 2H7-Fc-C825 (control). The bispecific anti-CD38 FP also resulted in tumor-to-normal organ ratios of absorbed radioactivity that were superior to those measured in animals receiving anti-CD38 SAB-PRIT (OKT10-SA chemical conjugate) in the same study (5:1; 8:1; 5:1 and 2:1 for blood, lung, liver and kidney respectively). In therapy studies, no animal in any group lost >7% of initial body weight and by day 11 anti-CD38 bispecific FP treated animals were 102.4% ± 3.4% of baseline weight. All mice in the control bispecific FP group experienced exponential MM tumor growth and 100% of these control animals required euthanasia within 27 days. All mice pretargeted with 028-Fc-C825 (anti-CD38 bispecific FP) demonstrated tumor response by day 6. After 45 days, 100% of the anti-CD38 bispecific FP treated animals remained alive (Figure A) and objective remissions were observed within 18 days in 100% of the mice treated with 028-Fc-C825 followed by 1200 µCi of 90Y-DOTA, including 90% complete remissions (no detectable tumor in 028Fc-C825 treated mice) compared to tumors that were 2040 ± 1389% of initial tumor volume [p<0.0001, Student's t-test] in untreated control animals by day 12 (Figure B). CONCLUSIONBiodistribution studies demonstrate MM tumor specific targeting after anti-CD38 xanti-chelated radiometal FP with minimal levels of yttrium-90 radioactivity detected in normal organs. Tumor responses are very encouraging in this MM xenograft model and results support clinical evaluation of bispecific anti-CD38 FP PRIT in MM. [Display omitted] [Display omitted] DisclosuresNo relevant conflicts of interest to declare.

Peripheral B Cells from Patients with MGUS and Multiple Myeloma (MM) Can Be Stimulated by Paraprotein-Target Specific T-Helper Cells to Produce Paraprotein-Identical Monoclonal Antibodies: Rationale for PARs, a Novel Therapeutic Approach with Ultimate Specificity in MGUS/MM

Background: Paratarg-7 (P-7) is the antigenic target of paraproteins(Preuss et al. Int J Cancer 2009;125:656-61) from 15% of European and 37% of African-American MGUS/MM patients, stronlgy supporting a role of P-7 in the pathogenesis of MGUS/MM via chronic auto-antigenic stimulation. All patients with P-7 specific paraproteins are carriers of the hyperphosphorylated version of p-7 (pP-7). We recently identified pP-7 specific T-helper cells which were restricted by certain "permissive" HLA-DR haplotypes (Neumann et al., Int J Cancer 2015; 137:1076-1084). These HLA-DR subtypes are overrepresented among patients with P-7 specific paraproteins compared to the corresponding normal population indicating that there are two prerequisites for the development of MGUS/MM with a P-7 specific paraprotein: 1st carriership of pP-7 and 2nd a permissive HLA-DR subtype. We now investigated the functional role of the pP-7 specific T-helper cells and their interaction with peripheral B cells with cognate specificity.Methods: Three patients with MGUS or MM, respectively, with a P-7 specific paraprotein and pP-7 specific T-helper cells were included in this study so far. In addition, the B cells from one healthy pP-7 carrying son of one of the patients were also analyzed. In vitro stimulation of antigen-specific peripheral B cells by pP-7 specific T-helper cells followed a modified protocol previously described by Lanzavecchia et al. (Eur J Immunol. 1983; 13:733-738). To this end, CD19+ B cells and CD3+ T cells were magnetically isolated from the proband's PBMC. T cells were replaced by corresponding T-helper cell clones.Results: In all patients studied, the autologous pP-7 specific T-helper cells stimulated the peripheral B cells to produce P-7 specific antibodies. The P-7 specific B-cell responses were monoclonal and the immunoglobulin type was the same as the paraprotein of the corresponding patient. In contrast, B-cell stimulation with CMV-pp65 specific T-helper cells used as controls always induced an antigen-specific, yet polyclonal response. When the peripheral B cells of a pP-7 carrying patient's son were also stimulated with pP-7 specific T-helper cells, they induced - in contrast to the mother - a polyclonal P-7 specific antibody response in his B cells, even though mother and son shared a "permissive" HLA-DR haplotype (HLA-DRB1*1301).Conclusion: In patients with MGUS/MM monoclonal B cells are found in the peripheral blood that can be induced to produce monoclonal antibodies identical to the serum paraprotein by T-helper cells with specificity for the antigenic target of the paraprotein. This does not only support an indispensable role of these T-helper cells in the pathogenesis of MGUS/MM via chronic antigenic stimulation, it also proves that precursors of the malignant plasma cells can be found in the peripheral blood that might fuel the development of malignant plasma cells. Cytogenetic and molecular genetic analyses are underway to determine if these precursor B-cells share the malignant genotype of their malignant plasma cells. These B cells can now be targeted by PARs (p araprotein a ntigens for r everse targeting) conjugated to toxins, as parts of bispecific constructs (PAR/CD3 or PAR/CD16) and/or PAR/CAR T cells. Use of PARs can be envisaged prophylactically for carriers of modified autoantigens like pP-7 with a permissive HLA-DR haplotype and a monoclonal B-cell response in vitro or in MM patients achieving a VGPR or CR after treatment for the prevention of relapse. DisclosuresNo relevant conflicts of interest to declare.

BARs (B -cell receptor a ntigens for r everse targeting): A Novel and Ultimately Specific Treatment Concept for B-Cell Neoplasms

Introduction: The major task of a B-cell receptor is the binding and internalization of its antigenic target, and its processing for antigen presentation to T-cells. Chronic antigenic stimulation has been discussed to play a role in the pathogenesis of malignant B-cell lymphomas. We therefore systematically searched for the antigenic targets of BCRs from various B-cell neoplasms.Methods: Recombinant BCRs were expressed as recombinant Fabs (rFabs) based on corresponding pairs of functional variable region heavy and light chain genes, which had been amplified from isolated genomic DNA lymphoma biopsies. Whenever possible, "natural" Fabs (nFabs) were also obtained by papain digestion of fresh or cultured lymphoma cells. Both nFab and rFab were used to screen for binding to proteins expressed on macroarrays derived from human cDNA expression libraries and identical binding pattern of nFabs and rFabs was demonstrated by an antigen competition assay.Results: Two antigens (paratarg-7 and sumoylated HSP-90 which are hyperphosphorylated and sumoylated, respectively, in patients compared to healthy controls) are the targets of paraproteins from (depending on ethnicity) 30-50% of all multiple myeloma patients; the BCR from 67% of patients with primary CNS lymphoma target hyperglycosylated neurabin, 26% of the BCR from ABC-type DLBCL target hypophosphorylated ARS2 and 45% of all mantle cell BCR target LRPAP1; optineurin is the BCR target of 12% follicular lymphomas and various autoantigens have been identified as the targets of roughly 30% of all CLL cases. For all autoantigens binding to its specific BCR, rapid internalization and induction of proliferation was demonstrated, indicating partial dependence on antigenic stimulation even in cell lines that had been in culture for years. Most importantly, BCR-specific cytotoxicity of recombinant pseudomonas-exotoxin conjugated ARS2 against an ABC-DLBCL cell line with BCR specific for ARS2 (OCI-Ly3) was demonstrated in vitro and in vivo after establishment of OCI-Ly3 lymphomas in SCID beige mice.Conclusions: Assuming that only a minority of BCR targets have been identified to date, the prevalence of posttranslationally modified autoantigens strongly supports a role of chronic antigenic stimulation in many B-cell neoplasms. Due to the predominance of a single or few BCR antigens in each malignant B-cell entity studied, BARs represent an attractive and novel therapeutic concept for a broad spectrum of B-cell neoplasms and are the first therapeutic approach in oncology that targets exclusively the malignant cells. BARs can be used for conjugation with toxins, radionuclides and small molecules as well as for bispecific constructs (e. g. with CD3 or CD16) and CAR T-cells, the toxicity of which should be drastically reduced due to the ultimate specificity of BARs that spares normal B-cells. Supported by Wilhelm-Sander-Stiftung DisclosuresPfreundschuh:Roche, Janssen, Celgene: Honoraria, Research Funding.