Search

Epidermal growth factor receptor (EGFR)-targeted cancer therapy such as anti-EGFR monoclonal antibodies and tyrosine kinase inhibitors have demonstrated clinical efficacy. However, there remains a medical need addressing limitations of these therapies, which include a narrow therapeutic window mainly due to skin and organ toxicity, and primary and secondary resistance mechanisms of the EGFR-signaling cascade (e.g., RAS-mutated colorectal cancer). Using the redirected optimized cell killing (ROCK®) antibody platform, we have developed AFM24, a novel bispecific, IgG1-scFv fusion antibody targeting CD16A on innate immune cells, and EGFR on tumor cells. We herein demonstrate binding of AFM24 to CD16A on natural killer (NK) cells and macrophages with KD values in the low nanomolar range and to various EGFR-expressing tumor cells. AFM24 was highly potent and effective for antibody-dependent cell-mediated cytotoxicity via NK cells, and also mediated antibody-dependent cellular phagocytosis via macrophages in vitro. Importantly, AFM24 was effective toward a variety of EGFR-expressing tumor cells, regardless of EGFR expression level and KRAS/BRAF mutational status. In vivo, AFM24 was well tolerated up to the highest dose (75 mg/kg) when administered to cynomolgus monkeys once weekly for 28 days. Notably, skin and other toxicities were not observed. A transient elevation of interleukin-6 levels was detected at all dose levels, 2–4 hours post-dose, which returned to baseline levels after 24 hours. These results emphasize the promise of bispecific innate cell engagers as an alternative cancer therapy and demonstrate the potential for AFM24 to effectively target tumors expressing varying levels of EGFR, regardless of their mutational status. Abbreviations: ADA: antidrug antibody; ADCC: antibody-dependent cell-mediated cytotoxicity; ADCP: antibody-dependent cellular phagocytosis; AUC: area under the curve; CAR: chimeric-antigen receptor; CD: Cluster of differentiation; CRC :colorectal cancer; ECD: extracellular domain; EGF: epidermal growth factorEGFR epidermal growth factor receptor; ELISA: enzyme-linked immunosorbent assay; FACS: fluorescence-activated cell sorting; Fc: fragment, crystallizableFv variable fragment; HNSCC: head and neck squamous carcinomaIL interleukinm; Ab monoclonal antibody; MOA: mechanism of action; NK :natural killer; NSCLC: non-small cell lung cancer; PBMC: peripheral blood mononuclear cell; PBS: phosphate-buffered saline; PD: pharmacodynamic; ROCK: redirected optimized cell killing; RSV: respiratory syncytial virus; SABC: specific antibody binding capacity; SD: standard deviation; TAM: tumor-associated macrophage; TKI: tyrosine kinase inhibitor; WT: wildtype

Acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) are common forms of adult hematologic malignancies with incidences of approximately 1-6/100.000/year. Progress in the management of AML and high-risk MDS has lagged behind other hematological malignancies, and, in particular, treatment for relapsed or refractory (R/R) disease and minimal residual disease (MRD) are needed to induce long-term remission. Here we describe the preclinical development of AFM28, a novel bispecific Innate Cell Engager (ICE ®) developed using the Redirected Optimized Cell Killing (ROCK ®) platform, designed to specifically activate NK cells and induce depletion of leukemic blasts and LSCs. Natural killer (NK) cell-based innate immunotherapy is emerging as a promising treatment option in AML and MDS based on the demonstrated susceptibility of leukemic blasts for NK cell killing and clinical activity of allogeneic NK cell therapy in R/R disease. Depletion of leukemic stem cells (LSCs) alongside leukemic blasts is considered a critical goal of treatment in order to eradicate MRD and prevent relapse. Accordingly, treatments that effectively target both cell types hold the potential to induce long-term remission in patients with AML or MDS. AFM28 binds the surface antigen CD123, which is almost universally expressed on blasts and LSCs in AML and MDS, and selectively targets CD16A expressed on NK cells, with no detectable binding to CD16B expressed on neutrophils. Through high affinity, bivalent binding to CD16A in a region distinct from the Fc binding site of IgGs, AFM28 remains stably bound to NK cells for extended periods and potently induces antibody-dependent cell-mediated cytotoxicity (ADCC) towards CD123+ AML and MDS cells. When compared with conventional anti-CD123 antibodies, including Fc-enhanced IgG1, AFM28 mediates more potent lysis of tumor cells even with low levels of CD123 expression. Furthermore, in contrast to conventional monoclonal CD123 antibodies, AFM28-induced lysis is not affected by the addition of competing IgG, which otherwise could impact efficacy. Similarly, ex vivo analysis of NK cell activation and inflammatory cytokine release in the ID.Flow human whole-blood loop system demonstrates that AFM28, compared with Fc-enhanced IgG1, produces 100-fold more potent NK cell activation associated with dose-dependent release of pro-inflammatory IL-6, TNFα, and IFNγ, concomitant with depletion of CD123+ plasmacytoid dendritic cells (pDCs) and basophils. In contrast, cytokine release was substantially lower than observed with a CD123-directed T cell-engaging bispecific antibody suggesting a low risk of cytokine release syndrome with AFM28. In vivo, AFM28 demonstrates a monoclonal antibody-like pharmacokinetic profile in mice and cynomolgus monkeys. A pre-clinical toxicology model in cynomolgus monkey suggested that AFM28 is safe and well tolerated and demonstrated pharmacodynamic activity. In summary, these data demonstrate AFM28 could be a novel treatment modality for AML and MDS that has good tolerability and the potential to effectively target MRD by depleting leukemic blasts and LSCs. AFM28 is currently being prepared for first-in-human clinical investigation. High affinity binding, potent induction of NK cell activation, and extended cell surface retention suggest AFM28 endows NK cells with CAR-like properties and may hold particular promise when combined with allogeneic NK cell therapy, for example within a pre-complexed NK cell product. Disclosures Götz: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Pahl: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Schmitt: Affimed GmbH: Research Funding. Müller: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Haneke: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Kozlowska: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Sarlang: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Knackmuss: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Peters: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Reusch: Affimed: Current Employment, Current holder of stock options in a privately-held company. Ross: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company. Nowak: AbbVie: Other: Investigator on funded clinical trial; Tolero Pharma, Pharmaxis, Apogenix: Research Funding; Affimed: Research Funding; Pharmaxis: Current holder of individual stocks in a privately-held company, Research Funding; Celgene: Honoraria; Takeda: Honoraria. Hofmann: Affimed GmbH: Research Funding. Merz: Affimed GmbH: Current Employment, Current holder of stock options in a privately-held company.

RO7297089 is a bispecific antibody (IgG-scFv) targeting Bcell maturation antigen (BCMA) and CD16a (FcγRIIIA) that is being developed for the treatment of multiple myeloma (MM). BCMA is exclusively expressed on plasmablasts and differentiated plasma cells (PCs), and is overexpressed on malignant PCs in MM patients. CD16a is expressed on natural killer (NK) cells, monocytes, mast cells, and macrophages. Herein, we characterized the mode of action and safety profile of RO7297089 in vitro and in vivo. RO7297089 showed potent cell killing when using BCMA+ MM tumor cell lines as target cells and human peripheral blood mononuclear cells, NK cells or macrophages as effector cells. Minimal increases in TNFα (2x) and IFNγ (4x), but not other cytokines/chemokines, were observed compared to the vehicle control treatment only in the presence of the BCMA+ MM tumor cell line up to the concentration tested. This suggests that, unlike T-cell engagers, the risk of cytokine release syndrome in patients receiving RO7297089 is low. Cynomolgus monkey is the only relevant nonclinical species for RO7297089 as it showed binding to both recombinant CD16 and BCMA with comparable affinity to human antigens. Following five weekly intravenous administrations to monkeys at 0, 15, and 50 mg/kg, RO7297089 was well tolerated. In line with the mechanism of action, there were no test article-related cytokine increases or adverse findings observed in both dose levels. Systemic exposure of RO7297089 was approximately dose proportional from 15 to 50 mg/kg. Anti-drug antibodies (ADA) were observed in some animals at both dose levels, and ADA-related decreases in concentrations were observed at only 15 mg/kg. To evaluate in vivo activity, RO7297089-related effects on total plasma sBCMA and PCs were assessed. Elevations of sBCMA levels (100x) were observed post dose at both dose levels, and these effects returned to predose levels in animals that did not maintain concentrations at 15 mg/kg, suggesting that RO7297089 bound to and stabilized circulating cynomolgus sBCMA. Time- and dose-dependent reductions in serum IgM levels were observed at both dose levels. Changes in PC numbers were not detected by immunophenotyping; however, gene expression analysis of PC markers was included and demonstrated clear reductions in mRNA expression levels of PC markers including BCMA and J-chain in blood at both dose levels, suggesting reductions in BCMA+ cells. Collectively, these studies suggest that RO7297089 selectively kills BCMA+ cells by engaging CD16a-positive immune cells and has a favorable safety profile. Citation Format: Satoko Kakiuchi-Kiyota, Melissa M. Schutten, Adeyemi O. Adedeji, Hao Cai, Robert Hendricks, Luna Liu, Sivan Cohen, Aaron M. Fullerton, Nicholas Corr, Lanlan Yu, Denise de Almeida Nagata, Shelly Zhong, Michael Dillon, Christoph Spiess, Steve R. Leong, Bing Zheng, Susanne Wingert, Uwe Reusch, Stefan Knackmuss, Thorsten Ross, Andrew Polson, Ayse M. Ovacik. Preclinical pharmacology and safety of RO7297089, a novel anti-BCMA/CD16a bispecific antibody for the treatment of multiple myeloma [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4556.

The epidermal growth factor receptor (EGFR) is a prime target for cancer therapy across many solid tumor types, including CRC, NSCLC, breast, esophageal cancer, and SCCHN. Binding of EGFR by its ligand EGF induces cell proliferation, a process which is uncontrolled or constitutively active in many cancers. The development of anti-EGFR therapies such as monoclonal antibodies and tyrosine kinase inhibitors enabled clinicians to treat specific patient populations. Nevertheless, prognoses for patients remain unfavorable due to occurrence of intrinsic or acquired resistance via mutations of EGFR and/or downstream signaling factors such as KRAS and BRAF which are drivers of tumorigenesis. Inhibitors that target specific KRAS mutations, such as the G12C mutation (i.e. 14% of the most frequently observed in NSCLC and up to 5% in CRC), have shown early promising efficacy. While such drugs hold promise for specific mutations, there continues to be a need for new treatments with more differentiated and novel modes of action that utilize the power of the innate immune system, such as antibody dependent cellular cytotoxicity (ADCC) and antibody dependent cellular phagocytosis (ADCP). Affimed has engineered the tetravalent bispecific EGFR- and CD16A-binding antibody AFM24 to engage and redirect innate immune cells such as NK cells and macrophages to EGFR+ tumor cells. In contrast to currently available EGFR-targeted therapies (e.g. cetuximab), this innate cell engager has been designed to exhibit a significantly reduced inhibition of EGFR signaling aiming for an improved safety profile, i.e. avoiding skin and gastro-intestinal toxicities. In vitro, AFM24 demonstrated high affinity binding to immune cells regardless of CD16A variants and potent killing of multiple EGFR+ tumor cell lines, carrying relevant KRAS or BRAF mutations via a different mode of action, namely ADCC and ADCP, rather than EGFR signaling inhibition. It was shown that AFM24's activity is independent of both, the EGFR cell surface expression level and the cellular origin of the respective cell lines. Cynomolgus toxicity studies revealed a very good safety profile, and no toxicity was observed after repeated administration (q7x28d, in total 5 infusions) of AFM24 up to 75 mg/kg. In contrast, EGFR-targeting mAbs (e.g. cetuximab) exhibited severe toxicities particularly in skin in toxicology studies when investigated with a similar dosing schedule. Due to its potent activation of innate immunity and its beneficial safety/tolerability profile, AFM24 has the potential to become a novel treatment option for patients suffering from EGFR-expressing cancers and to overcome limitations of available EGFR-targeted therapies such as resistance and/or associated toxicities. AFM24 has cleared the IND and clinical investigations are planned in patients with cancers known to express EGFR. Citation Format: Uwe Reusch, Michael Damrat, Susanne Wingert, Stefan H.J. Knackmuss, Thomas Mueller, Ivica Fucek, Ute Schniegler-Mattox, Kristina Ellwanger, Torsten Haneke, Andras Strassz, Wolfgang Fischer, Erich Rajkovic, Michael Tesar. AFM24, a bispecific EGFR/CD16A innate cell engager with the potential to overcome resistance to current targeted treatments for EGFR-positive malignancies [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5659.

To harness the potent tumor-killing capacity of T cells for the treatment of CD19(+) malignancies, we constructed AFM11, a humanized tetravalent bispecific CD19/CD3 tandem diabody (TandAb) consisting solely of Fv domains. The molecule exhibits good manufacturability and stability properties. AFM11 has 2 binding sites for CD3 and 2 for CD19, an antigen that is expressed from early B cell development through differentiation into plasma cells, and is an attractive alternative to CD20 as a target for the development of therapeutic antibodies to treat B cell malignancies. Comparison of the binding and cytotoxicity of AFM11 with those of a tandem scFv bispecific T cell engager (BiTE) molecule targeting the same antigens revealed that AFM11 elicited more potent in vitro B cell lysis. Though possessing high affinity to CD3, the TandAb mediates serial-killing of CD19(+) cells with little dependence of potency or efficacy upon effector:target ratio, unlike the BiTE. The advantage of the TandAb over the BiTE was most pronounced at lower effector:target ratios. AFM11 mediated strictly target-dependent T cell activation evidenced by CD25 and CD69 induction, proliferation, and cytokine release, notwithstanding bivalent CD3 engagement. In a NOD/scid xenograft model, AFM11 induced dose-dependent growth inhibition of Raji tumors in vivo, and radiolabeled TandAb exhibited excellent localization to tumor but not to normal tissue. After intravenous administration in mice, half-life ranged from 18.4 to 22.9 h. In a human ex vivo B-cell chronic lymphocytic leukemia study, AFM11 exhibited substantial cytotoxic activity in an autologous setting. Thus, AFM11 may represent a promising therapeutic for treatment of CD19(+) malignancies with an advantageous safety risk profile and anticipated dosing regimen.

To harness the cytotoxic capacity of immune cells for the treatment of solid tumors, we developed tetravalent, bispecific tandem diabody (TandAb) antibodies that recognize EGFRvIII, the deletion variant III of the epidermal growth factor receptor (EGFR), and CD3 on T-cells, thereby directing immune cells to eliminate EGFRvIII-positive tumor cells. Using phage display, we identified scFv antibodies selectively binding to EGFRvIII. These highly EGFRvIII-specific, fully human scFv were substantially improved by affinity maturation, achieving KDs in the picomolar range, and were used to construct a set of bispecific EGFRvIII-targeting TandAbs with a broad range of binding and cytotoxic properties. These antibodies exhibited an exquisite specificity for a distinguished epitope in the N-terminal portion of EGFRvIII, as shown on recombinant antigen in Western Blot, SPR, and ELISA, as well as on antigen-expressing cells in FACS assays, and did not bind to the wild-type EGFR. High-affinity EGFRvIII/CD3 TandAbs were most potent in killing assays, displaying cytotoxicity toward EGFRvIII-expressing CHO, F98 glioma, or human DK-MG cells with EC50 values in the range of 1–10 pM in vitro. They also demonstrated dose-dependent growth control in vivo in an EGFRvIII-positive subcutaneous xenograft tumor model. Together with the tumor-exclusive expression of EGFRvIII, the EGFRvIII/CD3 TandAbs’ high specificity and strictly target-dependent activation with no off-target activity provide an opportunity to target tumor cells and spare normal tissues, thereby reducing the side effects associated with other anti-EGFR therapies. In summary, EGFRvIII/CD3 TandAbs are highly attractive therapeutic antibody candidates for selective immunotherapy of EGFRvIII-positive tumors.

To improve recruitment and activation of natural killer (NK) cells to lyse tumor cells, we isolated a human anti-CD16A antibody with similar affinity for the CD16A 158F/V allotypes, but no binding to the CD16B isoform. Using CD16A-targeting Fv domains, we constructed a tetravalent bispecific CD30/CD16A tandem diabody (TandAb®) consisting solely of Fv domains. This TandAb has two binding sites for CD16A and two for CD30, the antigen identifying Hodgkin lymphoma cells. The binding and cytotoxicity of the TandAb were compared with antibodies with identical anti-CD30 domains: (1) a native IgG, (2) an IgG optimized for binding to Fc receptors, and (3) a bivalent bispecific CD30/CD16A diabody. Due to its CD16A-bivalency and reduced koff, the TandAb was retained longer on the surface of NK cells than the IgGs or the diabody. This contributed to the higher potency and efficacy of the TandAb relative to those of the other anti-CD30 antibodies. TandAb cytotoxicity was independent of the CD16A allotype, whereas the anti-CD30 IgGs were substantially less cytotoxic when NK cells with low affinity CD16A allotype were employed. TandAb activation of NK cells was strictly dependent on the presence of CD30(+) target cells. Therefore, the CD30/CD16A TandAb may represent a promising therapeutic for the treatment of Hodgkin's lymphoma; further, anti-CD16A TandAbs may function as potent immunotherapeutics that specifically recruit NK cells to destroy cancer cells.

Affimed has engineered AFM24 based on its Redirected Optimized Cell Killing (ROCK®) platform to redirect innate immune cells to EGFR-positive tumor cells with the potential for activation of a broad anti-tumor immune response. This fully human tetravalent bispecific EGFR and CD16A binding antibody was designed for the treatment of EGFR-expressing malignancies with the potential to overcome the resistance to other EGFR-targeting agents and offering a better safety profile. The efficacy of current standard of care (SOC), including EGFR-targeting monoclonal antibodies (mAbs) and tyrosine kinase inhibitors (TKIs), depends on intrinsic or acquired mutations that impact EGFR signal transduction. AFM24 exerts a mutation-independent mechanism of action (MOA), redirecting innate immunity via NK cells or macrophages to EGFR-positive tumor cells. PK analyses of AFM24 revealed a half-life of ~14 days in mice and ~5 days in cynomolgus monkeys. Using radioactively labeled AFM24 in mice, tumor-specific accumulation in subcutaneous human tumors was observed. While binding to CD16A by classical mAbs has low affinity and is further impaired by competition with serum IgG, no substantial changes in the very high binding affinity of AFM24 to NK cells were observed in the presence of IgG. Anti-tumor efficacy has been shown in in vitro cytotoxicity assays and in in vivo tumor models. Due to its MOA, AFM24 was also active against tumor cell lines expressing mutated Ras in in vitro assays. Antibody-dependent cellular phagocytosis (ADCP) of tumor cell lines expressing high levels of EGFR was comparable between AFM24 and classical or Fc-enhanced mAbs, but ADCP of tumor cells with lower EGFR expression mediated by AFM24 was more pronounced. Owing to their MOA of blocking EGFR-induced signaling, the therapeutic use of current SOC is associated with severe side effects such as skin toxicity, potentially limiting its use due to dosing interruptions and/or treatment termination. AFM24 is designed to exhibit a reduced potential to inhibit EGFR signaling, aiming for an improved safety profile. In a 4-week toxicology study in cynomolgus monkeys, no skin or organ toxicity was observed at AFM24 doses between 24 and 75 mg/kg i.v. once weekly. In toxicology studies investigating EGFR-targeting mAbs, e.g. cetuximab, with a similar experimental design, severe toxicities have been observed at those dosages. In summary, AFM24 is a novel, differentiated bispecific innate immune cell engager binding to EGFR and CD16A, thereby redirecting tumor cell killing by NK cells and macrophages to EGFR-expressing tumors. Its preclinical characterization described here, encourages AFM24’s development for the treatment of EGFR-expressing tumors with the potential to overcome resistance to current SOC while conferring a favorable safety profile. IND-enabling preclinical development is ongoing. Citation Format: Michael Kluge, Uwe Reusch, Stefan Knackmuss, Torsten Haneke, Susanne Wingert, Michael Damrat, Michael Tesar, Martin Treder. Preclinical characterization of the bispecific EGFR/CD16A innate immune cell engager AFM24 for the treatment of EGFR-expressing solid tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 559.

Purpose: Randomized studies with gemtuzumab ozogamicin have validated CD33 as a target for antigen-specific immunotherapy of acute myelogenous leukemia (AML). Here, we investigated the potential of CD33/CD3-directed tandem diabodies (TandAbs) as novel treatment approach for AML. These tetravalent bispecific antibodies provide two binding sites for each antigen to maintain the avidity of a bivalent antibody and have a molecular weight exceeding the renal clearance threshold, thus offering a longer half-life compared to smaller antibody constructs. Experimental Design: We constructed a series of TandAbs composed of anti-CD33 and anti-CD3 variable domains of diverse binding affinities and profiled their functional properties in CD33+ human leukemia cell lines, xenograft models, and AML patient samples. Results: Our studies demonstrated that several CD33/CD3 TandAbs could induce potent, dose-dependent cytolysis of CD33+ AML cell lines. This effect was modulated by the effector-to-target cell ratio and strictly required the presence of T cells. Activation and proliferation of T cells and maximal AML cell cytolysis correlated with high avidity to both CD33 and CD3. High-avidity TandAbs were broadly active in primary specimens from patients with newly diagnosed or relapsed/refractory AML in vitro, with cytotoxic properties independent of CD33 receptor density and cytogenetic risk. Tumor growth delay and inhibition were observed in both prophylactic and established HL-60 xenograft models in immunodeficient mice. Conclusions: Our data show high efficacy of CD33/CD3 TandAbs in various preclinical models of human AML. Together, these findings support further study of CD33/CD3 TandAbs as novel immunotherapeutics for patients with AML. Clin Cancer Res; 22(23); 5829–38. ©2016 AACR.

Background AFM11 is a bispecific, tetravalent T cell-engaging antibody construct binding to CD19 on B cell origin malignant cells such as B-precursor acute lymphoblastic leukemia (B-ALL) and to CD3 on T cells. By engaging CD3-positive T cells, AFM11 elicits T cell-mediated killing of CD19-positive (CD19+) leukemia and lymphoma cells. In vivo anti-tumor activity of AFM11 was investigated in a Raji tumor xenograft model in non-obese diabetic/severe combined immunodeficiency (NOD/scid) mice reconstituted with human peripheral blood mononuclear cells (PBMC). Tumor growth in all AFM11-treated animal groups was significantly reduced. In the highest dose group, all animals achieved complete tumor remission. (Reusch et al., 2015). An ongoing Phase 1 study assesses safety, tolerability, pharmacokinetics (PK), pharmacodynamics (PD) and preliminary efficacy of AFM11 in patients with relapsed/refractory (R/R) CD19+ B-ALL. Methods Patients (pts) with relapsed or refractory CD19+ B-ALL are being enrolled into escalating cohorts of 1-6 pts. The primary objective of the study is to determine the maximum tolerated dose (MTD) of AFM11 administered as a 2-week continuous intravenous (CIV) infusion. AFM11 is administered over the first 2 weeks (wks) of each 4 wk cycle for up to 3 cycles. Pts with rapidly progressing disease receive pre-treatment with 200 mg cyclophosphamide and 10 mg/m2 dexamethasone over 3-5 days before initiating AFM11. A lower starting dose is employed during the first wk of cycle 1 and escalated to a target dose during the second wk of cycle 1 and all subsequent cycles. An accelerated titration design is used until toxicity is observed, followed by a classical 3+3 design. PD activity is assessed by flow cytometry of peripheral blood lymphocytes and serum cytokine measurements. Tumor response is evaluated by local bone marrow and peripheral blood laboratory results between days 15 and 18 of each cycle. Results As of June 29, 2018, fourteen pts (8 female/6 male) have been treated in 5 cohorts. The median age is 41.5 years (range 19-67) and the median number of prior therapies is 5 (range 1-12). AFM11 was well-tolerated with no dose-limiting toxicities (DLTs) observed in the first 5 cohorts. The study switched to 3+3 design in cohort 3 due to the occurrence of grade 2 AFM11-related events in cohort 2. Enrollment into cohort 6 is ongoing and the MTD has not yet been reached. The most frequent (≥2 pts) AFM11-related adverse events were pyrexia (29%), myalgia (14%), and tremor (14%). Most of the events were Common Terminology Criteria for Adverse Events (CTCAE) v4.03 grade 1-2, with one grade 3 and no grade 4 events observed. Transient and reversible neurological events occurred in 3 of 14 (21%) pts: grade 1 paresthesia (n=1), grade 1/grade 2 tremor (n=1 each), grade 2 lethargy (n=1); and grade 3 altered state of consciousness (n=1). The grade 3 event occurred during the third cycle of treatment and was managed with treatment interruption and steroids and resolved within 48 hours. The pt then completed the third cycle at a reduced AFM11 dose without incident. Peripheral B cell reductions were observed in multiple pts in cohorts 4 and 5 and notable cytokine release was detected in two pts. Two pts achieved complete remission with complete hematological recovery (CR): one pt in cohort 4 achieved CR after the first cycle and progressed 2 wks later; one pt in cohort 5 achieved CR after the first cycle which was sustained after the second and third cycles and was assessed as minimal residual disease (MRD) negative after cycle 3. As a result, this pt became eligible to receive stem cell transplantation upon study completion. Both CR pts had peripheral B-cell aplasia after the first few days of treatment. Updated data will be presented. Conclusions The CD19/CD3-targeting tetravalent bispecific T cell engager AFM11 was well-tolerated in pts with R/R B-ALL across the first 5 cohorts of an ongoing Phase 1 study, and the MTD has not been reached. Pyrexia was the most frequently observed related adverse event. Transient neurological events were observed and were consistent with those associated with other CD19-targeted therapies. Peripheral B cell reductions and complete remissions observed in pts treated in the higher dose cohorts suggest that AFM11 is active in pts with R/R B-ALL and that the study is nearing determination of the therapeutic dose and schedule. Disclosures Salogub: Affimed: Research Funding. Mayer:Novartis: Research Funding; Roche: Research Funding; Eisai: Research Funding; Johnson & Johnson: Research Funding; Affimed: Research Funding. Folber:Affimed: Research Funding. Grosicki:Affimed: Research Funding. Skotnicki:Affimed: Research Funding. Prochwicz:Affimed: Research Funding. Myasnikov:Affimed: Research Funding. Gural:Affimed: Research Funding. Schoenborn-Kellenberger:Affimed: Consultancy. Brindley:Affimed: Consultancy. Knackmuss:Affimed: Employment. Schwarz:Affimed: Employment. Schmich:Affimed: Employment. Choe-Juliak:Affimed: Employment. Strassz:Affimed: Employment. Alland:Affimed: Employment. Doubek:AbbVie: Consultancy, Research Funding; Gilead: Consultancy, Honoraria, Research Funding; Janssen: Consultancy, Honoraria; Roche: Consultancy, Honoraria; Affimed: Research Funding; Novartis: Consultancy.

Natural killer (NK) cells are crucial effector cells of the innate immune system capable of rapidly recognizing and eliminating infected, stressed and malignant cells. NK cells are also the prime mediators of antibody-dependent cell-mediated cytotoxicity (ADCC), a potent mechanism of anti-viral immunity that has been applied to cancer therapy by targeting tumor-expressed surface antigens using monoclonal antibodies (mAbs). Classical ADCC is mediated by low affinity Fc-mediated engagement of NK cells via FcγRIIIA (CD16A) and is modulated by differences in target antigen expression levels. While high potency of therapeutic mAbs is achieved when target antigen is available at high density, potency and efficacy decrease substantially when copy numbers are low. Classical ADCC also needs to overcome the inhibitory effect of competing serum IgG and is negatively affected by a low affinity polymorphism of CD16A (158F) that is prevalent in approximately 8 of 10 individuals. Hence, classical Fc-mediated ADCC does not fully utilize the therapeutic potential of NK cell cytotoxicity. B cell maturation antigen (BCMA) has emerged as a promising target for treatment of multiple myeloma (MM) due to its near universal expression on tumor cells and restricted expression in non-malignant tissues. Numerous therapeutic approaches are currently investigated clinically and pre-clinically and target BCMA, however, none of these are aimed at fully utilizing NK cell-mediated ADCC. Low copy numbers of BCMA (approx. 40-15,000) might limit the activity of classical mAbs against BCMA, especially in the presence of high serum concentrations of paraprotein. In addition, NK cells may be ideally suited to target minimal residual disease immediately before or after autologous stem cell transplantation. Consequently, by efficiently redirecting NK cell cytotoxicity to BCMA+ myeloma, AFM26 used alone or in combination with other approaches may provide a novel, optimized treatment strategy. Here we describe development of AFM26, a BCMA and CD16A-directed tetravalent bispecific antibody that selectively engages CD16A+ effector cells, including NK cells, and is designed to overcome the limitations of classical ADCC. AFM26 is based on the recently launched Redirected, Optimized Cell Killing (ROCK) antibody platform and combines high affinity CD16A-directed effector cell engagement with IgG-like pharmacokinetics and manufacturability. We demonstrate that AFM26 interacts with NK cells with high avidity independently of CD16A polymorphism and in presence of competing IgG. NK cell-mediated lysis of BCMA+ target cell lines induced by AFM26 is largely independent of BCMA expression levels with high potency and efficacy observed at low copy numbers ( AFM26 therefore is a promising agent currently in preclinical development to fully unlock NK cell cytotoxicity for BCMA-directed immunotherapy of MM. Disclosures Ross: Affimed: Employment. Reusch:Affimed: Employment. Wingert:Affimed: Employment. Haneke:Affimed: Employment. Klausz:Affimed: Research Funding. Otte:Affimed: Research Funding. Schub:Affimed: Research Funding. Knackmuss:Affimed: Employment. Müller:Affimed: Employment. Ellwanger:Affimed: Employment. Fucek:Affimed: Employment. Schniegler-Mattox:Affimed: Employment. Koch:Affimed GmbH: Employment. Valerius:Affimed: Research Funding. Gramatzki:Affimed: Research Funding. Peipp:Affimed: Research Funding. Tesar:Affimed: Employment. Rajkovic:Affimed: Employment. Treder:Affimed GmbH: Employment.

The epidermal growth factor receptor (EGFR) is a validated target for the treatment of several solid tumor types. Current EGFR-targeting monoclonal antibodies (mAbs) and tyrosine kinase inhibitors function mainly through blocking of signal-transduction. Moreover, treatment with these agents is dependent on the receptor's mutational status which may cause treatment resistance in a large number of patients. In addition, EGFR-targeting therapies have been associated with side effects considered to impact prescription rates, in particular in the U.S. A natural killer (NK) cell-based EGFR-targeting approach has the potential to widen the therapeutic window and overcome intrinsic and acquired resistance. Therefore, it offers a promising and well differentiated therapeutic option. We describe the engineering of different bispecific NK cell engagers designed to redirect NK cell-mediated killing to EGFR-positive tumors. Different bispecific EGFR/CD16A antibody constructs targeting human and cynomolgus EGFR and CD16A were designed using novel human Fv antibody binding domains from a toolbox of formats. The Fv domains were evaluated in terms of their pharmacokinetic (PK) properties. Furthermore, the antibody constructs were characterized in antigen and cell binding assays, as well as in in vitro cytotoxicity assays against tumor cell lines expressing Raswt or mutated Ras. PK parameters were determined in CD1 mice and in vivo efficacy of selected antibodies was investigated in an A-431 tumor model in humanized mice. We generated antibody constructs with varying half-lives, including antibodies with a PK profile which allows for dosing comparable to IgG-based antibodies, while aiming at an improved safety profile compared to other EGFR-targeting therapies. Different high affinity tetravalent, bispecific antibodies binding to CD16A and the extracellular domain of EGFR were characterized. While binding to CD16A of classical mAbs is impaired by serum IgG, no substantial changes in binding affinity of our constructs to NK cells were observed. The antibodies showed superior potency and efficacy compared to mAbs or Fc-enhanced antibodies when tested in cytotoxicity assays. In vivo efficacy of selected antibodies was demonstrated in an A-431 tumor model in humanized mice. We have successfully designed and characterized several novel, highly potent and differentiated tetravalent bispecific antibody constructs to redirect NK cell-mediated cell killing to EGFR-positive tumor cells. These drug candidates are suitable for the treatment of EGFR-expressing malignancies with the potential to overcome resistance to other EGFR-targeting agents and offering a potentially improved safety profile. These candidates are currently in development as monotherapy and are also explored in combination with immune activating agents. Citation Format: Michael Kluge, Michael Tesar, Uwe Reusch, Stefan Knackmuss, Torsten Haneke, Kristina Ellwanger, Ivica Fucek, Thomas Mueller, Ute Schniegler-Mattox, Martin Treder. Pharmacokinetics and in vitro/in vivo characterization of high-affinity bispecific EGFR/CD16A NK cell engagers for the treatment of EGFR-expressing tumors [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 2786.

Recombinant antibodies isolated from human antibody libraries have excellent affinities and high target specificity. As full-length IgGs are cleared inadequately slowly from the circulation, the aim of this work was to figure out which kind of recombinant antibody fragment proves to be appropriate for imaging epithelial cell adhesion molecule (EpCAM)-expressing tumours with the short-living radioisotope (68)Ga.In order to combine the promising tumour targeting properties of antibodies with (68)Ga, four antibody variants with the same specificity and origin only differing in molecular weight were constructed for comparison. Therefore, the binding domains of a single-chain fragment variable (scFv) isolated from a human naïve antibody library were modified genetically to construct the respective full-length IgG, the tria- and diabody variants. These molecules were conjugated with the bifunctional chelating agent N,N'-bis[2-hydroxy-5-(carboxyethyl)benzyl]ethylenediamine-N,N'-diacetic acid (HBED-CC) to enable (68)Ga labelling at ambient temperature and compared in biodistribution and immuno-PET imaging experiments.The antibody variants with identical specificity proved to have the correct molecular weight, high binding affinity and specificity to their antigen, EpCAM. Radiometal complexation was efficiently performed at room temperature leading to (68)Ga-labelled antibodies with unchanged binding properties compared to the original antibody variants. The best targeting properties were obtained with the scFv and especially with the diabody. The triabody showed higher absolute tumour uptake but only moderate clearance from circulation.The antibody variants differed considerably in normal organ uptake, clearance from circulation and tumour accumulation. The data demonstrate the feasibility of imaging solid tumours with the (68)Ga-labelled diabody format. This type of recombinant protein might be a promising carrier even for the short-lived radiometal (68)Ga to support e.g. the management of immunotherapy which may provide important information regarding receptor expression of solid tumours.

The 37-kDa/67-kDa laminin receptor precursor/laminin receptor (LRP/LR) acting as a receptor for prions and viruses is overexpressed in various cancer cell lines, and their metastatic potential correlates with LRP/LR levels. We analyzed the tumorigenic fibrosarcoma cell line HT1080 regarding 37-kDa/67-kDa LRP/LR levels and its invasive potential. Compared to the less invasive embryonic fibroblast cell line NIH3T3, the tumorigenic HT1080 cells display approximately 1.6-fold higher cell-surface levels of LRP/LR. We show that anti-LRP/LR tools interfere with the invasive potential of HT1080 cells. Anti-LRP/LR single-chain variable fragment antibody (scFv) iS18 generated by chain shuffling from parental scFv S18 and its full-length version immunoglobulin G1-iS18 reduced the invasive potential of HT1080 cells significantly by 37% and 38%, respectively. HT1080 cells transfected with lentiviral plasmids expressing small interfering RNAs directed against LRP mRNA showed reduced LRP levels by approximately 44%, concomitant with a significant decrease in the invasive potential by approximately 37%. The polysulfated glycans HM2602 and pentosan polysulfate (SP-54), both capable of blocking LRP/LR, reduced the invasive potential by 20% and 35%, respectively. Adhesion of HT1080 cells to laminin-1 was significantly impeded by scFv iS18 and immunoglobulin G1-iS18 by 60% and 68%, respectively, and by SP-54 and HM2602 by 80%, suggesting that the reduced invasive capacity achieved by these tools is due to the perturbation of the LRP/LR-laminin interaction on the cell surface. Our in vitro data suggest that reagents directed against LRP/LR or LRP mRNA such as antibodies, polysulfated glycans, or small interfering RNAs, previously shown to encompass an anti-prion activity by blocking or downregulating the prion receptor LRP/LR, might also be potential cancer therapeutics blocking metastasis by interfering with the LRP/LR-laminin interaction in neoplastic tissues.

The neuronal perturbations in Alzheimer’s disease are attributed to the formation of extracellular amyloid-β (Aβ) neuritic plaques, composed predominantly of the neurotoxic Aβ42 isoform. Although the plaques have demonstrated a role in synaptic dysfunction, neuronal cytotoxicity has been attributed to soluble Aβ42 oligomers. The 37kDa/67kDa laminin receptor has been implicated in Aβ42 shedding and Aβ42-induced neuronal cytotoxicity, as well as internalization of this neurotoxic peptide. As the cellular prion protein binds to both LRP/LR and Aβ42, the mechanism underlying this cytotoxicity may be indirectly due to the PrPc-Aβ42 interaction with LRP/LR. The effects of this interaction were investigated by 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide assays. PrPc overexpression significantly enhanced Aβ42 cytotoxicity in vitro, while PrP–/– cells were more resistant to the cytotoxic effects of Aβ42 and exhibited significantly less cell death than PrPc expressing N2a cells. Although anti-LRP/LR specific antibody IgG1-iS18 significantly enhanced cell viability in both pSFV1-huPrP1-253 transfected and non-transfected cells treated with exogenous Aβ42, it failed to have any cell rescuing effect in PrP–/– HpL3-4 cells. These results suggest that LRP/LR plays a significant role in Aβ42-PrPc mediated cytotoxicity and that anti-LRP/LR specific antibodies may serve as potential therapeutic tools for Alzheimer’s disease.

Cancer has become a major problem worldwide due to its increasing incidence and mortality rates. Both the 37kDa/67kDa laminin receptor (LRP/LR) and telomerase are overexpressed in cancer cells. LRP/LR enhances the invasiveness of cancer cells thereby promoting metastasis, supporting angiogenesis and hampering apoptosis. An essential component of telomerase, hTERT is overexpressed in 85–90% of most cancers. hTERT expression and increased telomerase activity are associated with tumor progression. As LRP/LR and hTERT both play a role in cancer progression, we investigated a possible correlation between LRP/LR and telomerase. LRP/LR and hTERT co-localized in the perinuclear compartment of tumorigenic breast cancer (MDA_MB231) cells and non-tumorigenic human embryonic kidney (HEK293) cells. FLAG® Co-immunoprecipitation assays confirmed an interaction between LRP/LR and hTERT. In addition, flow cytometry revealed that both cell lines displayed high cell surface and intracellular LRP/LR and hTERT levels. Knock-down of LRP/LR by RNAi technology significantly reduced telomerase activity. These results suggest for the first time a novel function of LRP/LR in contributing to telomerase activity. siRNAs targeting LRP/LR may act as a potential alternative therapeutic tool for cancer treatment by (i) blocking metastasis (ii) promoting angiogenesis (iii) inducing apoptosis and (iv) impeding telomerase activity.

e14001 Background: The epidermal growth factor receptor is a target for the treatment of solid tumors, like CRC, HNSCC and NSCLC. Therapeutic success of EGFR-targeting agents depends on mutations defining resistance to these agents, thus an effective EGFR-targeting therapy overcoming resistance would offer an attractive new treatment option. Natural killer (NK-) cells play a central role in the innate immune system and can destroy cancer cells. NK-cell-engaging bispecific antibodies may offer a safe and effective therapy to target EGFR-expressing tumor cells irrespective of their mutational status. Methods: AFM24 was generated using proprietary human anti-EGFR and anti-CD16A variable domains and characterized for binding, stability, manufacturability, efficacy and safety in biophysical and functional assays in vitro and in vivo. Assays for binding to target/effector cells and cytotoxicity were performed in the presence of physiologically relevant IgG levels using EGFR+ tumor cell lines with and without RAS. The safety profile of AFM24 was investigated in vitro for off-target activity and in vivo in cynomolgus monkeys. Results: AFM24 showed excellent biophysical properties and picomolar EC50 values in cytotoxicity assays. It was more potent than cetuximab, or an antibody with cetuximab-derived anti-EGFR variable domain. Importantly, AFM24 was less prone to IgG interference and eliminated cell lines harboring negative predictive biomarkers in in vitro cytotoxicity assays, even in the presence of relevant serum IgG levels. AFM24 demonstrated in vivo efficacy in a humanized hu-NOG mouse model. Furthermore, AFM24 exhibited a favorable safety profile in cynomolgus monkeys. It was well tolerated up to 94 mg/kg in a single dose, maximum tolerated dose study (MTD) when administered as a 2 hour infusion. Conclusions: Our data suggest that AFM24 is a highly potent and potentially safer drug candidate suitable for the treatment of EGFR-expressing cancers with the potential to overcome resistance to other EGFR-targeting agents and to avoid toxic side effects seen with other anti-EGFR therapeutics.

Neuronal loss is a major neuropathological hallmark of Alzheimer's disease (AD). The associations between soluble Aβ oligomers and cellular components cause this neurotoxicity. The 37 kDa/67 kDa laminin receptor (LRP/LR) has recently been implicated in Aβ pathogenesis. In this study the mechanism underlying the pathological role of LRP/LR was elucidated. Forsters Resonance Energy Transfer (FRET) revealed that LRP/LR and Aβ form a biologically relevant interaction. The ability of LRP/LR to form stable associations with endogenously shed Aβ was confirmed by pull down assays and Aβ-ELISAs. Antibody blockade of this association significantly lowered Aβ42 induced apoptosis. Furthermore, antibody blockade and shRNA mediated downregulation of LRP/LR significantly hampered Aβ42 internalization. These results suggest that LRP/LR is a receptor for Aβ42 internalization, mediating its endocytosis and contributing to the cytotoxicity of the neuropeptide by facilitating intra-cellular Aβ42 accumulation. These findings recommend anti-LRP/LR specific antibodies and shRNAs as potential therapeutic tools for AD treatment.

Two key events, namely adhesion and invasion, are pivotal to the occurrence of metastasis. Importantly, the 37 kDa/67 kDa laminin receptor (LRP/LR) has been implicated in enhancing these two events thus facilitating cancer progression. In the current study, the role of LRP/LR in the adhesion and invasion of liver cancer (HUH-7) and leukaemia (K562) cells was investigated. Flow cytometry revealed that the HUH-7 cells displayed significantly higher cell surface LRP/LR levels compared to the poorly-invasive breast cancer (MCF-7) control cells, whilst the K562 cells displayed significantly lower cell surface LRP/LR levels in comparison to the MCF-7 control cells. However, Western blotting and densitometric analysis revealed that all three tumorigenic cell lines did not differ significantly with regards to total LRP/LR levels. Furthermore, treatment of liver cancer cells with anti-LRP/LR specific antibody IgG1-iS18 (0.2 mg/ml) significantly reduced the adhesive potential of cells to laminin-1 and the invasive potential of cells through the ECM-like Matrigel, whilst leukaemia cells showed no significant differences in both instances. Additionally, Pearson's correlation coefficients suggested direct proportionality between cell surface LRP/LR levels and the adhesive and invasive potential of liver cancer and leukaemia cells. These findings suggest the potential use of anti-LRP/LR specific antibody IgG1-iS18 as an alternative therapeutic tool for metastatic liver cancer through impediment of the LRP/LR- laminin-1 interaction.

Alzheimer's disease (AD) is the most prevalent form of dementia. The amyloid beta (Aβ) peptide is the predominant candidate aetiological agent and is generated through the sequential proteolytic cleavage of the Amyloid Precursor Protein (APP) by beta (β) and gamma (γ) secretases. Since the cellular prion protein (PrP(c)) has been shown to regulate Aβ shedding, we investigated whether the cellular receptor for PrP(c), namely the 37 kDa/67 kDa Laminin Receptor (LRP/LR) played a role in Aβ shedding. Here we show that LRP/LR co-localises with the AD relevant proteins APP, β- and γ-secretase, respectively. Antibody blockage and shRNA knock-down of LRP/LR reduces Aβ shedding, due to impediment of β-secretase activity, rather than alteration of APP, β- and γ-secretase levels. These findings indicate that LRP/LR contributes to Aβ shedding and recommend anti-LRP/LR specific antibodies and shRNAs as novel therapeutic tools for AD treatment.

Adhesion and invasion have been identified as the two key components of metastasis. The 37 kDa/67 kDa laminin receptor (LRP/LR) is thought to enhance these two processes thus endorsing the progression of cancer. Here we report on LRP/LR and the metastatic potential of MDA-MB 231 breast and WHCO1 oesophageal cancer cells. Western blot analysis revealed a significant increase in total laminin receptor precursor (LRP) levels of breast and oesophageal cancer cells in comparison to non-invasive MCF-7 breast cancer cells, whereas LRP/LR cell surface levels in both cell lines were not significantly different to those of MCF-7 cells as analysed by flow cytometry. Incubation of breast and oesophageal cancer cells with the anti-LRP/LR specific antibody, IgG1-iS18, resulted in significant reduction in the adhesive potential of WHCO1 and MDA-MB 231 cells by 92% and 16%, respectively. Moreover, invasion was significantly impeded by 98% and 25% for WHCO1 and MDA-MB 231 cells, respectively. Pearson's correlation coefficients proved a positive correlation between total LRP/LR levels and invasive potential as well as between the adhesive and invasive potential of breast and oesophageal cancer cells. Our findings suggest that through interference of the LRP/LR-laminin-1 interaction, anti-LRP/LR specific antibody IgG1-iS18 may act as a possible alternative therapeutic tool for metastatic breast and oesophageal cancer treatment.

EGFRvIII is the most prevalent tumor-specific variant of the wild-type EGFR and represents an attractive tumor target in various solid tumors such as GMB, HNSCC or NSCLC. Despite the clinical successes achieved with EGFR targeting antibodies or small molecule inhibitors little therapeutic progress has been made with EGFRvIII. There is still a high medical need in such cancers and several agents are in development to address this notoriously difficult target: Celldex′ vaccine rindopepimut in combination with bevacizumab in late-stage development, or ADCs in early development by Amgen or Abbvie. However, despite the high tumor specificity of EGFRvIII expression, no potent T- or NK-cell engaging are currently in clinical development. It has been recognized that there may be more therapeutic potential with bi-specific antibodies recruiting immune effectors for targeted destruction of antigen-positive tumor cells. We developed tetravalent, bi-specific TandAbs that recognize EGFRvIII, and recruiting either T-cells or NK-cells by binding to the activating receptors CD3 and CD16A, respectively. This targeted antibody approach allows the selective destruction of EGFRvIII+ tumor cells employing highly potent and efficacious immune effector cells whilst sparing normal cells that are EGFRvIII- or cells presenting the ubiquitously expressed EGFR. The selected EGFRvIII/CD3 and EGFRvIII/CD16A TandAbs exhibited exquisite specificity towards EGFRvIII in Western Blot, SPR, ELISA, and flow cytometric assays. No binding was observed to recombinant EGFR or to EGFR+ cells. They also displayed potent cytotoxicity towards EGFRvIII+ cell lines with EC50 values in the low picomolar range. No cytotoxicity was observed on EGFRvIII- target cells or cells expressing EGFR demonstrating the high selectivity of anti-EGFRvIII TandAbs. Importantly, in the absence of EGFRvIII+ target cells our TandAbs did not elicit T- or NK-cell activation and activation-induced immune cell proliferation suggesting an excellent safety profile. In vivo pharmacodynamics for anti-EGFRvIII TandAbs was demonstrated in a mouse xenograft model. The clinical relevance of EGFRvIII as a tumor marker and the binding of our anti-EGFRvIII variable domains were evaluated by immunohistochemistry. Binding was shown in glioblastoma and other solid tumors, but no expression was detectable on healthy tissue. In summary, the presented in vitro and in vivo studies qualify EGFRvIII/CD3 and EGFRvIII/CD16A TandAbs as highly attractive therapeutic antibody candidates and provides us with the possibility of employing both T-cell and NK-cell recruiting strategies for selective immunotherapy of EGFRvIII+ tumors. Due to tumor specific expression of EGFRvIII and the absence of off-target activity our TandAbs display an excellent safety profile reducing the risks of side effects associated with other anti-EGFRvIII therapies. Citation Format: Kristina Ellwanger, Uwe Reusch, Ivica Fucek, Michael Weichel, Thorsten Gantke, Stefan Knackmuss, Martin Treder. Anti-EGFRvIII TandAbs recruiting either T or NK cells are highly specific and potent therapeutic antibody candidates for the treatment of EGFRvIII+ tumors. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 580.

e14008Background: EGFRvIII is the most prevalent tumor-specific variant of the EGFR family and represents an attractive tumor target such as in glioblastoma, head and neck or lung cancer. EGFRvIII ...

e14007Background: The epidermal growth factor receptor (EGFR) is an important and established target for the treatment of several solid tumors, including colorectal, head and neck and lung cancer. ...

The 37kDa/67kDa laminin receptor (LRP/LR) is a central receptor mediating interactions between tumour cells and the basement membrane and is thereby a key player in adhesion and invasion, essential processes in metastatic cancer. To affect continued tumour growth, tumours induce angiogenesis for the constant delivery of nutrients and oxygen. This study aims to determine the blocking effect of the anti-LRP/LR specific antibody, W3 on the angiogenic potential of HUVE (human umbilical vein endothelial) cells. Flow cytometric analysis revealed that 97% of HUVE cells display cell surface LRP/LR. An angiogenesis assay was conducted employing HUVE cells seeded on the basement membrane reconstituent Matrigel™ supplemented with the pro-angiogenic factor vascular endothelial growth factor (VEGF). Post 18h incubation at 37°C tubular structures, namely tube lengths were assessed. Treatment of established tubular structures with 100 µg/ml anti-LRP/LR specific antibody completely blocked angiogenesis. Our findings suggest a central role of the 37kDa/67kDa LRP/LR in tube formation and recommends anti-LRP/LR specific antibodies as potential therapeutic tools for treatment of tumour angiogenesis.

The 37-kDa/67-kDa laminin receptor [laminin receptor precursor/high-affinity laminin receptor (LRP/LR)] is thought to play a major role in invasion and adhesion, key components of metastatic cancer. Lung cancer, cervical cancer, colon cancer and prostate cancer are among the top 10 cancer types worldwide. Here, we report that LRP/LR levels on the surface of lung cancer cells, cervical cancer cells, colon cancer cells and prostate cancer cells are significantly increased compared to non-tumorigenic fibroblasts. Adhesion of lung cancer cells, cervical cancer cells, colon cancer cells and prostate cancer cells to laminin-1 is significantly reduced, employing the anti-LRP/LR-specific antibody IgG1-iS18. Invasion of these cell lines into the Matrigel™ matrix was significantly impeded with IgG1-iS18. The Pearson's correlation coefficient proves a correlation between LRP/LR cell-surface levels and invasion potential, as well as adhesion and invasion, respectively. Our findings suggest that IgG1-iS18 antibody might act as alternative therapeutic tool for treatment of various metastatic cancer types.

IgM antibodies exist both in a pentameric soluble form and as membrane-bound monomers mainly on the surface of naive B cells, where they are part of the antigen receptor complex. Naive B cells, constituting 75% of the peripheral blood B cell repertoire in humans (Klein et al., 1997), contain the largest diversity of an individual's rearranged immunoglobulin genes. The naturally occurring antibody repertoire contains specific antibodies against various antigens. In a primary immune response, B cells expressing antigen-specific IgM molecules are activated and differentiate into antibody-producing and -secreting plasma cells. Secreted antigen-specific IgM molecules are the first immunoglobulins occurring during a primary immune response. On the other hand, so-called natural antibodies exist independently of antigenic stimulation and are thought to contribute to the first line of defense against infections (Carsetti et al., 2004; Ochsenbein & Zinkernagel, 2000) as well as malignancy (Brandlein et al., 2003). In addition to antibody-secreting plasma cells, memory B cells are generated during a primary immune response, a process that includes somatic hypermutation in the germinal centers. Most of the memory B cells have undergone a class switch and do not express IgM. In humans, however, IgM molecules with somatic mutations have been identified (Van Es et al., 1992). These somatically mutated IgM molecules contribute to an individual's immunological memory and constitute about 10% of the total peripheral blood B cell repertoire (Klein et al., 1997). IgM-expressing memory B cells protect against infections by encapsulated bacteria, and develop during the first year of life (Kruetzmann et al., 2003).

The 37/67 kDa laminin receptor (LRP/LR) acts as a receptor for prions providing a promising target for the treatment of prion diseases. Recently, we selected anti-LRP/LR single-chain antibodies (scFvs) and proved a reduction of the peripheral PrPScpropagation by passive immunotransfer into scrapie-infected mice. Here, we report the development of anin vivogene delivery system based on adeno-associated virus (AAV) vectors expressing scFvs-S18 and -N3 directed against LRP/LR. Transduction of neuronal and non-neuronal cells with recombinant (r)AAV serotype 2 vectors encoding scFv-S18, -N3 and -C9 verified the efficient secretion of the antibodies. These vectors were administered via stereotactic intracerebral microinjection into the hippocampus of C57BL/6 mice, followed by intracerebral inoculation with 10 % RML at the same site 2 weeks post-injection of rAAV. After 90 days post-infection, scFv-S18 and -N3 expression resulted in the reduction of peripheral PrPScpropagation by approximately 60 and 32 %, respectively, without a significant prolongation of incubation times and survival. Proof of rAAV vector DNA in spleen samples by real-time PCR strongly suggests a transport or trafficking of rAAV from the brain to the spleen, resulting in rAAV-mediated expression of scFv followed by reduced PrPSclevels in the spleen most likely due to the blockage of the prion receptor LRP/LR by scFv.

To harness the cytotoxic capacity of immune effector cells for the treatment of several types of solid tumors, we developed tetravalent bifunctional antibodies that recognize EGFRvIII, the deletion variant III of EGFR. Their second functionality binds with high affinity to CD3, thereby directing T-cells to eliminate EGFRvIII+ cancer cells. The expression of EGFRvIII on various solid tumor types, and its absence from healthy tissues, provides an opportunity to develop cytotoxic antibodies that solely target cancer; these would spare normal tissues, and substantially reduce the side effects associated with EGFR therapy. Using phage display, we identified scFvs that selectively bind to EGFRvIII but not to EGFR. The binding affinities of these highly EGFRvIII-specific antibodies were substantially improved, employing affinity maturation techniques, and achieved KDs in the 100pM range and lower. We engineered a panel of bispecific EGFRvIII/CD3 TandAbs with a broad range of binding and cytotoxic properties. Mono- and bivalent binding, specificity for both EGFRvIII and CD3, T cell-mediated cytotoxic activity, and target-mediated T-cell activation were characterized in a panel of in vitro assays. EGFRvIII/CD3 TandAbs exhibit exquisite specificity towards the EGFRvIII antigen in Western Blot, SPR, ELISA, and FACS assays of EGFRvIII+ cells. No specific binding was observed to recombinant EGFR antigen or to EGFR-expressing cells. Apparent affinities of EGFRvIII/CD3 TandAbs to EGFRvIII were up to 25fold improved relative to the monovalently binding scFvs, and achieved a KD of 11pM for the best binding EGFRvIII/CD3 TandAb. Improvement of the binding affinities in the bivalent TandAb format was largely due to slower dissociation. TandAbs with high affinity for EGFRvIII were most potent in killing assays, displaying cytotoxicity towards EGFRvIII-expressing F98 glioma and CHO cells with EC50 in the range of 1pM-10pM. No cytotoxicity was observed on EGFR+ cells or EGFRvIII- cells up to the maximally-evaluated TandAb concentration of 0.5μM (which is 100000 fold higher than the EC50) demonstrating the high selectivity of EGFRvIII/CD3 TandAbs for the tumor-specific EGFRvIII. High affinity binding to CD3 was necessary for efficacious T cell recruitment as shown by the correlation of CD3-binding and cytotoxic potency of EGFRvIII/CD3 TandAbs. Importantly, in the absence of EGFRvIII+ target cells in vitro, TandAbs did not elicit T cell activation, as demonstrated by their lack of proliferation: this specificity contributed to a good preclinical safety profile. Biophysical and pharmacological characterization of several candidates is currently ongoing, whereby the first EGFRvIII/CD3 TandAbs demonstrated a robust dose-dependent growth retardation of EGFRvIII+ subcutaneous xenograft tumors. In summary, EGFRvIII/CD3 TandAbs are specific and highly potent drug candidates for the treatment of EGFRvIII+ malignancies. Citation Format: Kristina Ellwanger, Uwe Reusch, Ivica Fucek, Michael Weichel, Carmen Herbrecht, Stefan Knackmuss, Erich Rajkovic, Vera Molkenthin, Eugene A. Zhukovsky. EGFRvIII T-cell TandAbs are specific and highly potent drug candidates for the treatment of a variety of solid tumors. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 2471. doi:10.1158/1538-7445.AM2015-2471

3057 Background: Antibody-based targeting agents specific for CD33 and CD3 are evolving as promising strategies for treating AML. T550, T564, and T589 are TandAbs that are bispecific as well as tet...

7071 Background: Acute Myeloid Leukemia (AML) is a hematologic malignancy with dire prognosis; immunotherapy may offer a much needed alternative to cytotoxic chemotherapy. TandAbs are tetravalent, ...

7067 Background: CD33 had been validated as an AML target in randomized studies of the antibody-drug conjugate, gemtuzumab ozogamicin (GO) in a subset of patients, but currently explored CD33-targe...

Binding of fibrinogen to platelet integrin alphaIIbbeta3 mediates platelet aggregation, and thus inhibition of alphaIIbbeta3 represents a powerful therapeutic strategy in cardiovascular medicine. However, the currently used inhibitors of alphaIIbbeta3 demonstrate several adverse effects like thrombocytopenia and bleeding, which are associated with their property to bind to non-activated alphaIIbbeta3. To circumvent these problems, we designed blocking single-chain antibody-fragments (scFv) that bind to alphaIIbbeta3 exclusively in its activated conformation. Two naive phage libraries were created: a natural phage library, based on human lymphocyte cDNA, and a synthetic library, with randomized VHCDR3. We performed serial rounds of subtractive panning with depletion on non-activated and selection on activated alphaIIbbeta3, which were provided on resting and ADP-stimulated platelets and CHO cells, expressing wild-type or mutated and thereby activated alphaIIbbeta3. In contrast to isolated, immobilized targets, as generally used for phage display, this unique cell-based approach for panning allowed the preservation of functional integrin conformation. Thereby, we obtained several scFv-clones that demonstrated exclusive binding to activated platelets and complete inhibition of fibrinogen binding and platelet aggregation. Interestingly, all activation-specific clones contained an RXD pattern in the HCDR3. Binding studies on transiently expressed point mutants and mouse-human domain-switch mutants of alphaIIbbeta3 indicate a binding site similar to fibrinogen. In conclusion, we generated human activation-specific scFvs against alphaIIbbeta3, which bind selectively to activated alphaIIbbeta3 and thereby potently inhibit fibrinogen binding to alphaIIbbeta3 and platelet aggregation.

To target NK cells against non-Hodgkin’s lymphoma, we constructed a bispecific diabody (BsDb) with reactivity against both human CD19 and FcγRIII (CD16). Bacterially produced CD19 × CD16 BsDb specifically interacted with both CD19+ and CD16+ cells and exhibited significantly higher apparent affinity and slower dissociation from the tumor cells than from effector cells. It was able to induce specific lysis of tumor cells in the presence of isolated human NK cells or nonfractionated PBLs. The combination of the CD19 × CD16 BsDb with a previously described CD19 × CD3 BsDb and CD28 costimulation significantly increased the lytic potential of human PBLs. Treatment of SCID mice bearing an established Burkitt’s lymphoma (5 mm in diameter) with human PBLs, CD19 × CD16 BsDb, CD19 × CD3 BsDb, and anti-CD28 mAb resulted in the complete elimination of tumors in 80% of animals. In contrast, mice receiving human PBLs in combination with either diabody alone showed only partial tumor regression. These data clearly demonstrate the synergistic effect of small recombinant bispecific molecules recruiting different populations of human effector cells to the same tumor target.

To harness the cytotoxic capacity of T cells for the treatment of EGFRvIII+ cancers, we developed a humanized bispecific tetravalent antibody, with two binding sites for CD3 (or CD16A) and EGFRvIII, the EGFRvIII/CD3 (and EGFRvIII/CD16A) RECRUIT-TandAb. EGFR dysregulation has been linked to numerous cancers, and both small molecules and EGFR targeting antibodies have successfully reached the clinic. The deletion variant III (EGFRvIII), the most common mutant form of EGFR is expressed exclusively in cancer tissues contributing to oncogenic transformation. The restricted EGFRvIII expression on various solid tumor types provides an opportunity to develop cytotoxic antibodies that solely target cancer, sparing normal tissues, and substantially reduce the side effects associated with EGFR therapy. Using phage display libraries, we identified scFvs that selectively bound to the mutated and not the native form of EGFR. T cells are potent tumor-killing effector cells that cannot be recruited by native antibodies; hence we engineered a panel of bispecific EGFRvIII/CD3 TandAbs, capable of T cell-recruitment, with a broad range of binding and cytotoxic properties. The TandAbs' binding properties, T cell-mediated cytotoxic activity, and target-mediated T cell activation were characterized in a panel of in vitro assays. They exhibited exquisite specificity towards EGFRvIII in ELISA and FACS assays, with no wild-type EGFR binding observed, up to the maximally evaluated TandAb concentration of 1 μM. The most potent TandAbs displayed cytotoxicity towards EGFRvIII expressing F98 glioma and CHO cells with EC50 = 25 pM. We also assayed the cytotoxicity of these TandAbs towards EGFR+ cells as a more sensitive probe of residual binding to the native form. No cytotoxicity was observed up to the maximally evaluated TandAb concentration of 0.5 μM. The cytotoxic potency of the EGFRvIII/CD3 TandAb was higher than that of a comparator bivalent bispecific antibody, constructed from different anti-EGFRvIII/CD3 moieties or of EGFRvIII/CD16A TandAbs, which recruit NK cells instead of cytotoxic T cells. EGFRvIII/CD3 TandAbs with high affinity binding to CD3 were most efficacious for T cell recruitment and tumor cell killing and yet, in the absence of EGFRvIII+ target cells in vitro, TandAbs did not elicit T cell activation, as measured by their lack of proliferation, contributing to a good preclinical safety profile. In vivo EGFRvIII/CD3 TandAbs demonstrated a robust dose-dependent growth retardation of EGFRvIII+ subcutaneous Xenograft tumors in NOD/scid mice reconstituted with human PBMC; this tumor growth inhibition was more pronounced than that associated with cetuximab. In summary, our data demonstrate the strict specificity and high potency of the anti-tumor cytotoxicity mediated by the EGFRvIII/CD3 TandAb, a novel highly potent drug candidate for the treatment of EGFRvIII+ malignancies. Citation Format: Eugene A. Zhukovsky, Kristina Ellwanger, Uwe Reusch, Markus Eser, Fionnuala McAleese, Ivica Fucek, Carmen Burkhardt, Stefan Knackmuss. A highly cytotoxic EGFRvIII/CD3 TandAb recruits T cells to specifically and potently kill several types of solid tumor cancers. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2643. doi:10.1158/1538-7445.AM2014-2643

e19546 Background: AFM11 is a bispecific TandAb molecule that recruits T cells (via an anti-CD3 domain) to CD19+ tumor cells (via an anti-CD19 domain). Due to the lack of a relevant species, the no...

To harness the potent tumor-killing capacity of T cells for the treatment of CD19+ malignancies, we developed a humanized bispecific tetravalent antibody, with two binding sites for CD3 and CD19, the CD19/CD3 RECRUIT-TandAb AFM11. CD19 is expressed from early B cell development through differentiation into plasma cells, and is an attractive alternative to CD20 as a target for the development of therapeutic antibodies to treat B cell malignancies such as Non Hodgkin Lymphoma. Since native antibodies cannot recruit T cells, we engineered a bispecific anti-CD19/anti-CD3 TandAb. The tumor-specific CD19 antigen module targets the TandAb to cancer cells, while simultaneously, the CD3 effector module recruits and activates T cells, leading to cancer cell lysis. The advantages of the TandAb technology, relative to other bi-functional fragment antibody scaffolds, include: improved pharmacokinetics (PK) enabling intravenous dosing, more drug-like properties, and avidity-enhanced efficacy for the targeting and killing of tumor cells. We evaluated in vitro efficacy and safety using CD19+ cell lines, and in vivo efficacy in a murine NOD/scid xenograft model reconstituted with human PBMC. Further, we used standard preclinical IND enabling assays to evaluate tissue cross reactivity, PK, and toxicological profile (local tolerance, hematocompatibility, effects on hematopoesis, etc). In vitro assays demonstrated the higher potency and efficacy of target cell lysis by AFM11 relative to a bispecific tandem scFv (that is currently in clinical evaluation). CD8+ T cells dominate early AFM11-mediated cytotoxicity (4 hrs) while after 24 hrs both CD4+ and CD8+ T cells equally contribute to tumor lysis with EC50 between 0.5 – 5 pM; cytotoxicity was independent of CD19 cell-surface density. AFM11 exhibited similar cytotoxicity over effector:target ratios ranging from 5:1 to 1:5, and facilitated serial T cell-killing of its targets. The advantage of AFM11 over the bispecific tandem scFv was most pronounced at lower effector:target ratios. AFM11 activated T cells only in the presence of CD19+ cells. In PBMC cultures, AFM11 induced CD69 and CD25 expression, T cell proliferation, and production of IFN-γ, TNF-α, IL-2, IL-6, and IL-10. Depletion of CD19+ cells from PBMC abrogated these effects, demonstrating that the T cell activation is strictly CD19+ target-dependent. Thus, AFM11 should not elicit the devastating cytokine release observed when full-length antibodies bind CD3. Up to one week co-incubation with AFM11 did not inhibit T cell cytotoxicity, suggesting that the TandAb does not induce anergy. In vivo, AFM11 induced dose-dependent growth inhibition of Raji tumors; a single 0.5 mg/kg dose exhibited efficacy similar to 5 daily injections. In the tissue cross reactivity study, only tissues containing CD19+ and CD3+ cells were stained by AFM11; all other tissues, including vital organs, displayed no cross reactivity. Similarly, no local intolerance was observed in rabbits, and no effect on myeloid and erythroid progenitors was observed in a colony-forming assay. Strong accumulation of 125I-labeled AFM11 was observed in the tumors of mice engrafted with CD19+ cancer cells, and no unspecific organ accumulation was observed. Finally, evaluated on the basis of Cmax and the area under the curve (AUC), AFM11 exhibited dose linearity (20 – 500 mg AFM11 dose range) upon single i.v. bolus administration in mice; half-life (T1/2) ranged from 18.4 to 22.9 hr. In summary, AFM11 is a highly efficacious novel drug candidate for the treatment of CD19+ malignancies with an advantageous safety profile and anticipated dosing regimen. Disclosures: Zhukovsky: Affimed Therapeutics AG: Employment, Equity Ownership. Reusch:Affimed Therapeutics AG: Employment. Burkhardt:Affimed Therapeutics AG: Employment. Knackmuss:Affimed Therapeutics AG: Employment. Fucek:Affimed Therapeutics AG: Employment. Eser:Affimed Therapeutics AG: Employment. McAleese:Affimed Therapeutics AG: Employment. Ellwanger:Affimed Therapeutics AG: Employment. Little:Affimed Therapeutics AG: Consultancy, Equity Ownership.

3068 Background: CD19, due to broader expression on B cell subtypes, is an attractive alternative to CD20 as a target for treatment of B cell malignancies. T cells are potent tumor-killing effectors that cannot be recruited by full length antibodies, however TandAb technology harnesses their cytotoxic nature for oncology indications. The CD3 RECRUIT TandAb AFM11 enables T cells to potently and specifically kill CD19+ tumors and possesses advantageous PK properties enabling intravenous dosing. Methods: We constructed AFM11, a human bispecific tetravalent antibody with two binding sites for both CD3 and CD19. In vitro efficacy and safety were evaluated on CD19+ cell lines and primary tumors. In vivo efficacy was evaluated in a murine NOD/scid xenograft model reconstituted with human PBMC. Results: In vitro assays demonstrate higher potency and efficacy of target cell lysis by AFM11 relative to a bispecific tandem scFv. CD8+ T cells dominate early cytotoxicity (4 hrs) while after 24 hrs both CD4+ and CD8+ T cells equally contribute to tumor lysis with EC50 of 0.5 – 5 pM; cytotoxicity is independent of cell CD19 density. AFM11 exhibits similar cytotoxicity at Effector:Target ratios from 5:1 to 1:5 and facilitates T cell serial killing of its targets. AFM11 activates T cells only in the presence of CD19+ cells. In PBMC cultures AFM11 induces CD69 and CD25 expression, T cell proliferation, and production of IFN-γ, TNF-α, IL-2, IL-6, and IL-10. Depletion of CD19+ cells from PBMC abrogates these effects, and indicates strict CD19+ target-dependent T cell activation. Thus, AFM11 should not elicit the devastating cytokine release observed when full length antibodies bind CD3. Cell lysis by AFM11 is restricted to CD19+ targets asCD19- bystanders are not lysed in co-culture assays. Up to one week co-incubation with AFM11 does not inhibit T cell cytotoxicity and thus it does not induce anergy. In vivo AFM11 exhibits a dose-dependent growth inhibition of Raji tumors; a single dose of AFM11 exhibits similar efficacy as 5 daily injections. Conclusions: AFM11 is a highly efficacious novel drug candidate for the treatment of CD19+ malignancies with an advantageous safety profile and anticipated dosing regimen.

CD19 is expressed from early B cell development through differentiation into plasma cells, and is an attractive alternative to CD20 as a target for therapeutic antibodies that treat B cell malignancies. T cells are potent tumor-killing effector cells that cannot be recruited by native antibodies. The TandAb technology enables the development of bispecific antibody therapeutics for oncology indications that harness the cytotoxic nature of T cells. The CD3 RECRUIT-TandAb AFM11, a human bispecific tetravalent antibody with two binding sites for both CD3 and CD19, exhibits highly potent cytotoxic activity and possesses pharmacokinetic properties that are considerably improved relative to alternative bispecific formats. In vitro assays demonstrate high affinity binding of AFM11 to CD3+ and CD19+ cells leading to higher potency and efficacy of T cell recruitment and target cell lysis relative to alternative bispecific antibody formats(tandem scFv, diabody). Cytotoxicity assays show that AFM11 mediates target cell lysis by CD4+ and CD8+ T cells with EC50 values in the low to sub-picomolar range that is independent of the target cell CD19 density. AFM11 exhibits nearly identical cytotoxicity across a broad range of effector-to-target ratios, from 5:1 to 1:5, and facilitates serial target cell killing by T cells. Binding of antibodies to CD3 on T cells may elicit devastating cytokine release if it results in systemic rather than target-specific activation of T cells. Indeed, the anti-CD3 antibody OKT3 exhibits high toxicity in humans due to non-specific activation of T cells. Thus, in vitro assays were used to assess AFM11 off-target activity. AFM11 activates T cells inly in the presence of CD19+ cells, resulting in induction of CD69 and CD25 expression and the on-target release of IFN-γ, TNF-α, IL-2, and IL-6. However, depletion of CD19+ cells from PBMC cultures completely abrogates AFM11-mediated activation marker expression, T cell proliferation, and cytokine release, and supports strict CD19+ target-dependent T cell activation. Conversely, OKT3, produced cytokine release and T cell proliferation in all cultures except purified T cells. AFM11-mediated target cell lysis is specific for and restricted to CD19+ target cells since CD19− bystanders were not lysed in assays with mixed target populations. Binding of AFM11 to CD3 also did not induce anergy since its incubation with T cells for several days did not inhibit their potent cytotoxicity. AFM11 demonstrated a robust dose-dependent inhibition of growth of Raji tumor xenografts in NOD/scid mice reconstituted with human PBMC. Moreover, a single dose of AFM11 resulted in tumor growth inhibition similar to that produced by 5 daily injections. Altogether, these data suggest that AFM11 is a highly efficacious novel drug candidate for the treatment of hematological malignancies with advantages in both safety profile and anticipated dosing regimen. Citation Format: Eugene A. Zhukovsky, Uwe Reusch, Carmen Burkhardt, Stefan Knackmuss, Ivica Fucek, Markus Eser, Fionnuala McAleese, Kristina Ellwanger. Bispecific TandAbs: a safe and potent platform for T cell-mediated killing of CD19+ cells. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 1243. doi:10.1158/1538-7445.AM2013-1243

AFM13 represents a new class of bispecific antibodies, based on the RECRUIT-TandAb platform, which enables the generation of therapeutics with improved efficacy, convenient administration regimens, and excellent safety profiles. AFM13 was developed for the treatment of Hodgkin Lymphoma (HL) and other CD30+ malignancies. AFM13 targets CD30 on HL tumor cells and recruits NK cells via CD16A. Preclinical data demonstrate specific and highly efficient anti-tumor activity by the selectively recruited NK cell immune response, which addresses key deficiencies of monoclonal antibodies: (i) reduced binding to the 158F allotype of CD16A and (ii) non-selective binding to immune effector cells via both activating CD16A and non-signalling CD16B receptors. In vitro assays demonstrate that AFM13 binds with high affinity to both CD30 and CD16A antigens and rapidly induces lysis of CD30+ cells at picomolar concentrations. Further, AFM13 activates CD16A only in the presence of tumor cells: there is no systemic activation of NK-cells in the absence of target cells. A robust GMP process in mammalian cells and a lyophilized formulation with excellent stability are established. Finally, no toxicity is observed in Cynomolgus monkeys. AFM13 was investigated in an open-label single-arm phase I dose escalation trial in heavily pre-treated patients with relapsed/refractory HL. Each patient received 4 weekly doses of AFM13. Seven dose levels from 0.01 to 7.0 mg/kg were escalated in cohorts of 3 patients. In addition, one cohort of 4 patients received AFM13 twice a week at 4.5 mg/kg for 4 weeks. All dose levels of AFM13 proved to be well tolerated and safe and clear signs of anti-tumor activity were demonstrated. We have performed an analysis of the activation of NK cells and clearance of soluble CD30 in humans since these are key parameters for the mode of action of AFM13. We confirmed a correlation between anti-tumor activity parameters and biomarkers based on a statistically significant dose-dependent increase in the activation of NK cells and reduction in soluble CD30 levels. Since the activation of NK cells has been reported to facilitate low-level expression of CD30 on NK cells, which we also observed in several patients in the highest AFM13 dose groups, we investigated whether this may affect NK cell cytotoxicity. Our ex vivo evaluation of samples from patients that received AFM13 and in vitro evaluation of activated NK cells demonstrated no inhibition of NK cell-mediated cytotoxic activity. Furthermore, AFM13 exhibits a remarkable half-life of 1 day which represents a substantial increase relative to that of diabody-like formats currently being evaluated in the clinic for haematological malignancies. Thus, AFM13 has demonstrated encouraging biologic activity and represents a new targeted therapy for heavily pre-treated patients with HL. Citation Format: Achim Rothe, Max S. Topp, Anas Younes, Uwe Reusch, Stefan Knackmuss, Erich Rajkovic, Dennis A. Eichenauer, Horst Hummel, Katrin S. Reiners, Markus Dietlein, Joerg Kessler, Miroslav Ravic, Christian Hucke, Eugene Zhukovsky, Elke Pogge von Strandmann, Andreas Engert. The CD30/CD16A RECRUIT TandAb AFM13 activates patient NK cells resulting in specific tumor cell killing. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 4572. doi:10.1158/1538-7445.AM2013-4572

Abstract 3721 Background: CD19 is expressed from early B cell development through differentiation into plasma cells, and is an attractive alternative to CD20 as a target for the development of therapeutic antibodies to treat B cell malignancies. T cells are potent tumor-killing effector cells that cannot be recruited by native antibodies. The CD3 RECRUIT-TandAb AFM11, a humanized bispecific tetravalent antibody with two binding sites for both CD3 and CD19, is a novel therapeutic for the treatment of NHL that harnesses the cytotoxic nature of T cells. Methods: We engineered a bispecific anti-CD19/anti-CD3e tetravalent TandAb with humanized and affinity-matured variable domains. The TandAb's binding properties, T cell-mediated cytotoxic activity, and target-mediated T cell activation were characterized in a panel of in vitro assays. In vivo efficacy was evaluated in a murine NOD/scid xenograft model reconstituted with human PBMC. Results: AFM11 mediates highly potent CD19+ tumor cell lysis in cytotoxicity assays performed on a panel of cell lines (JOK-1, Raji, Nalm-6, MEC-1, VAL, Daudi) and primary B-CLL tumors: EC50 values are in the low- to sub-picomolar range and do not correlate with the expression density of CD19 on the target cell lines. The cytotoxic activity of tetravalent AFM11 is superior to that of alternative bivalent antibody formats possessing only a single binding site for both CD19 and CD3. High affinity binding of AFM11 to CD19 and to CD3 is essential for efficacious T cell recruitment. Both CD8+ and CD4+ T cells mediate cytotoxicity however the former exhibit much faster killing. We observe that AFM11 displays similar cytotoxic efficacy at different effector to target ratios (from 5:1 to 1:5) in cytotoxicity assays; this suggests that T cells are engaged in the serial killing of CD19+ target cells. In the absence of CD19+ target cells in vitro, AFM11 does not elicit T cell activation as manifested by cytokine release (from a panel of ten cytokines associated with T cell activation), their proliferation, or their expression of activation markers. AFM11 activates T cells exclusively in the presence of its targets and mediates lysis of CD19+ cells while sparing antigen-negative bystanders. In the absence of CD19+ target cells, AFM11 concentrations in excess of 500-fold over EC50 induce down-modulation of the CD3/TCR complex. Yet, AFM11-treated T cells can be re-engaged for target cell lysis. All of these features of AFM11-induced T cell activation may contribute additional safety without compromising its efficacy. In vivo AFM11 demonstrates a robust dose-dependent inhibition of subcutaneous Raji tumors in mice. At 5 mg/kg AFM11 demonstrates a complete suppression of tumor growth, and even at 5 ug/kg tumor growth is reduced by 60%. Moreover, we observe that a single administration of AFM11 produces inhibition of tumor growth similar to that of 5 consecutive administrations. Conclusions: In summary, our in vitro and in vivo experiments with AFM11 demonstrate the high potency and efficacy of its anti-tumor cytotoxicity. Thus, AFM11 is a novel highly efficacious drug candidate for the treatment of B cell malignancies with an advantageous safety profile. Disclosures: Zhukovsky: Affimed Therapeutics AG: Employment, Equity Ownership. Reusch:Affimed Therapeutics AG: Employment. Burkhardt:Affimed Therapeutics AG: Employment. Knackmuss:Affimed Therapeutics AG: Employment. Fucek:Affimed Therapeutics AG: Employment. Eser:Affimed Therapeutics AG: Employment. McAleese:Affimed Therapeutics AG: Employment. Ellwanger:Affimed Therapeutics AG: Employment.

8059 Background: CD19 is expressed from early B cell development to the differentiation into plasma cells and is an attractive target for B cell malignancies either lacking CD20 expression or refractory to anti-CD20 antibody therapies. T cells are potent tumor killing effector cells that are not recruited by native antibodies. The CD3 RECRUIT-TandAb AFM11, a human bispecific tetravalent antibody with two binding sites for both CD3 and CD19, is a novel therapeutic for the treatment of NHL that harnesses the cytotoxic nature of T cells. Methods: A bispecific anti-CD19/anti-CD3 tetravalent TandAb with humanized and affinity matured variable domains was constructed. The TandAb’s binding, T-cell mediated cytotoxic activity, and cytokine release were characterized in a panel of in vitro assays. In vivo efficacy was evaluated in a murine NOD/scid xenograft model reconstituted with human PBMC. Results: AFM11 mediates highly potent target tumor cell lysis in cytotoxicity assays: EC50 values are low to sub-picomolar range in a panel of CD19+ cell lines and primary B-CLL tumor cells. The cytotoxic activity of tetravalent AFM11 is superior to that of alternative bivalent antibody formats possessing only a single binding site for both CD19 and CD3. High affinity binding of AFM11 to CD19, and more so to CD3 (low to sub-nanomolar Kd), is essential for efficacious T cell recruitment. The high affinity bivalent binding of AFM11 to CD3 does not trigger T cell activation in the absence of CD19+ target cells in functional in vitro assays. AFM11 activates T cells only in the presence of its targets and mediates lysis while sparing antigen-negative bystanders. AFM11 induces down-modulation of the CD3/TCR complex in the absence of target cells and at high concentrations. Also, AFM11-treated T cells can be re-engaged for target cell lysis. These features of AFM11-induced T cell activation may contribute additional safety with no compromise of efficacy. Finally, AFM11 demonstrates a robust dose-dependent inhibition of subcutaneous Raji tumors in mice. Conclusions: AFM11 is a novel highly efficacious drug candidate for the treatment of B cell malignancies with an advantageous safety profile.

e18532 Background: CD30 is a validated tumor target that is expressed on both Hodgkin’s Lymphoma cells and some activated immune cells. To improve antibody-based therapies targeting CD30, Affimed developed an NK cell-recruiting bispecific antibody. Methods: To enhance the recruitment and activation of NK cells for tumor cell lysis, we isolated a human antibody that is highly specific for the A isoform of FcgRIII (CD16A). Its variable domain (Fv) was used to construct a tetravalent bispecific TandAb, consisting of four Fvs linked in tandem, with two binding sites for both the CD16A on NK cell and the CD30 target cell antigens. The resulting anti-CD30/anti-CD16A bispecific tetravalent TandAb AFM13 was compared to both an anti-CD30/anti-CD16A bispecific diabody and an Fc-enhanced anti-CD30 IgG in binding and cytotoxicity assays in vitro. Pharmacokinetics and safety of the TandAb were investigated in Cynomolgus monkeys. Results: AFM13 is retained longer on NK cells than either the anti-CD30 IgG or the bispecific diabody due to its bivalency for CD16A; it binds both CD16A alleles (158F, 158V) equally well indicating the potential for broader patient response than that achieved with current therapeutic mAbs. AFM13 did not induce systemic activation of NK cells. It displays higher potency (EC50) and higher efficacy (maximal lysis) than either the anti-VD30 IgGs or the bispecific diabody in a cytotoxicity assay. The cytotoxic activity of AFM13 is unaffected by physiological concentrations of polyclonal IgG, whereas that of anti-CD30 IgGs is dramatically reduced. Cynomolgus monkeys were identified as the only relevant species for toxicity testing. Pharmacokinetics studies reveal AFM13 half-lives of up to 22.9 h after repeated dosing. MTD was not reached in a single dose escalation study ranging to 100 mg/kg maximum. A 28 days toxicity study, with dosing up to 10 mg/kg every other day, revealed an excellent safety profile. No signs of immunotoxicity or systemic cytokine release were observed. Conclusions: AFM13 may be a more effective treatment for Hodgkin’s Lymphoma that is refractory to treatment with anti-CD30 IgGs and is under evaluation in a Phase I clinical trial.

NK cells are professional cytotoxic immune effector cells and, unlike cytotoxic T cells, exist in a constitutively activated state not requiring pre- or co-stimulation. Thus, NK cells are attractive effectors to employ for cancer immunotherapy with bispecific antibodies. To improve the recruitment and activation of NK cells for lysis of tumor cells, we isolated a human antibody that was highly specific for the A isoform of FcαRIII (CD16A) and displayed substantially lower dissociation constant than native or Fc-optimized mAbs. The variable domains of this antibody were used to construct a tetravalent bispecific antibody, termed TandAb, consisting solely of antibody variable domains, with two valencies for CD16A (on effector cells) and two for CD30 (on target cells). CD30 represents a validated tumor target which is expressed on Hodgkin's Reed-Sternberg lymphoma cells, and on some activated immune cells. The binding properties of the TandAb and its efficacy for inducing tumor cell lysis were compared to those of a native anti-CD30 mAb, a mAb with enhanced Fc receptor binding properties, and an anti-CD30 bispecific diabody, all of which comprised the same anti-CD30 variable domains as the TandAb. Due to its bivalency for CD16A, the TandAb was retained longer on NK cells than the IgGs or the bispecific diabody, and relatively high concentrations of IgG were necessary to displace it from NK cells. The TandAb bound equally well to both CD16A alleles (158F, 158V) indicating a potentially broader patient response than that achieved with current therapeutic mAbs. The TandAb did not induce any systemic activation of NK cells even though it's binding was bivalent to CD16A. In a cytotoxicity assay, the TandAb displayed higher potency (EC50) and higher efficacy (maximal lysis) than the IgGs or the bispecific diabody. Cynomolgus monkeys were identified as the only relevant species for toxicity testing. Pharmacokinetics studies revealed half-lives for the TandAb up to 22.9 h after repeated dosing. No MTD was reached in a single dose escalation study to a maximum dose of 100 mg/kg. A pivotal 28 days toxicity study in which animals were dosed repeatedly up to 10 mg/kg every other day revealed an excellent safety profile. Several immunological parameters were assessed indicating no signs of immunotoxicity or systemic cytokine release. In conclusion, TandAbs may provide a more effective treatment for tumors that are refractory to treatment with IgGs and is currently being evaluated in the Phase I clinical trial. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3522. doi:1538-7445.AM2012-3522