Ekkehard Mössner

Increasing the efficacy of CD20 antibody therapy through the engineering of a new type II anti-CD20 antibody with enhanced direct and immune effector cell-mediated B-cell cytotoxicity

CD20 is an important target for the treatment of B-cell malignancies, including non-Hodgkin lymphoma as well as autoimmune disorders. B-cell depletion therapy using monoclonal antibodies against CD20, such as rituximab, has revolutionized the treatment of these disorders, greatly improving overall survival in patients. Here, we report the development of GA101 as the first Fc-engineered, type II humanized IgG1 antibody against CD20. Relative to rituximab, GA101 has increased direct and immune effector cell-mediated cytotoxicity and exhibits superior activity in cellular assays and whole blood B-cell depletion assays. In human lymphoma xenograft models, GA101 exhibits superior antitumor activity, resulting in the induction of complete tumor remission and increased overall survival. In nonhuman primates, GA101 demonstrates superior B cell-depleting activity in lymphoid tissue, including in lymph nodes and spleen. Taken together, these results provide compelling evidence for the development of GA101 as a promising new therapy for the treatment of B-cell disorders.

Epitope characterization and crystal structure of GA101 provide insights into the molecular basis for type I/II distinction of CD20 antibodies

CD20 is a cell-surface marker of normal and malignant B cells. Rituximab, a monoclonal antibody targeting CD20, has improved the treatment of malignant lymphomas. Therapeutic CD20 antibodies are classified as either type I or II based on different mechanisms of killing malignant B cells. To reveal the molecular basis of this distinction, we fine-mapped the epitopes recognized by both types. We also determined the first X-ray structure of a type II antibody by crystallizing the obinutuzumab (GA101) Fab fragment alone and in complex with a CD20 cyclopeptide. Despite recognizing an overlapping epitope, GA101 binds CD20 in a completely different orientation than type I antibodies. Moreover, the elbow angle of GA101 is almost 30° wider than in type I antibodies, potentially resulting in different spatial arrangements of 2 CD20 molecules bound to a single GA101 or rituximab molecule. Using protein tomography, different CD20 complexes were found to be associated with the 2 antibodies, and confocal microscopy showed different membrane compartmentalization of these subpopulations of the cellular CD20 pool. Our findings offer a possible molecular explanation for the different cellular responses elicited by type I and II antibodies.

Epitope interactions of monoclonal antibodies targeting CD20 and their relationship to functional properties

Several novel anti-CD20 monoclonal antibodies are currently in development with the aim of improving the treatment of B cell malignancies. Mutagenesis and epitope mapping studies have revealed differences between the CD20 epitopes recognized by these antibodies. Recently, X-ray crystallography studies confirmed that the Type I CD20 antibody rituximab and the Type II CD20 antibody obinutuzumab (GA101) differ fundamentally in their interaction with CD20 despite recognizing a partially overlapping epitope on CD20. The Type I CD20 antibodies rituximab and ofatumumab are known to bind to different epitopes. The differences suggest that the biological properties of these antibodies are not solely determined by their core epitope sequences, but also depend on other factors, such as the elbow hinge angle, the orientation of the bound antibody and differential effects mediated by the Fc region of the antibody. Taken together, these factors may explain differences in the preclinical properties and clinical efficacy of anti-CD20 antibodies.

RG7386, a novel tetravalent FAP-DR5 antibody, effectively triggers FAP-dependent, avidity-driven DR5 hyperclustering and tumor cell apoptosis

Dysregulated cellular apoptosis and resistance to cell death are hallmarks of neoplastic initiation and disease progression. Therefore, the development of agents that overcome apoptosis dysregulation in tumor cells is an attractive therapeutic approach. Activation of the extrinsic apoptotic pathway is strongly dependent on death receptor (DR) hyperclustering on the cell surface. However, strategies to activate DR5 or DR4 through agonistic antibodies have had only limited clinical success. To pursue an alternative approach for tumor-targeted induction of apoptosis, we engineered a bispecific antibody (BsAb), which simultaneously targets fibroblast-activation protein (FAP) on cancer-associated fibroblasts in tumor stroma and DR5 on tumor cells. We hypothesized that bivalent binding to both FAP and DR5 leads to avidity-driven hyperclustering of DR5 and subsequently strong induction of apoptosis in tumor cells but not in normal cells. Here, we show that RG7386, an optimized FAP-DR5 BsAb, triggers potent tumor cell apoptosis in vitro and in vivo in preclinical tumor models with FAP-positive stroma. RG7386 antitumor efficacy was strictly FAP dependent, was independent of FcR cross-linking, and was superior to conventional DR5 antibodies. In combination with irinotecan or doxorubicin, FAP-DR5 treatment resulted in substantial tumor regression in patient-derived xenograft models. FAP-DR5 also demonstrated single-agent activity against FAP-expressing malignant cells, due to cross-binding of FAP and DR5 across tumor cells. Taken together, these data demonstrate that RG7386, a novel and potent antitumor agent in both mono- and combination therapies, overcomes limitations of previous DR5 antibodies and represents a promising approach to conquer tumor-associated resistance to apoptosis. Mol Cancer Ther; 15(5); 946–57. ©2016 AACR.

Response to: Monoclonal antibodies targeting CD20

Dear Editor, Please allow us to briefly reply to the comments by Dr. Goldenberg and colleagues. As stated in the disclosures, some authors are employees of Roche Glycart AG or Roche Diagnostics GmbH and have contributed to the (non-) clinical development of the glycoengineered Type II CD20 antibody GA101 (obinutuzumab). In writing this review, we aimed to give a general overview of the epitopes, structural and mechanistic features of CD20 antibodies in development. Due to length restrictions, this work could not necessarily be comprehensive and did not cover, for example, bispecific antibodies. Furthermore, the discussion of clinical data was intentionally not within the scope of this review because it is covered in other expert reviews.1-3 Overall, we do not agree with the conclusions drawn by Dr. Goldenberg and colleagues. In their publication, Goldenberg et al. show indeed that veltuzumab has a slower off-rate than rituximab on B cells, however, scatchard analysis showed that the KD values were comparable, with 8.6 nM for rituximab and 10.8 nM for veltuzumab.4 More importantly, this finding did not appear to translate into major functional differentiation. Veltuzumab showed in vitro anti-proliferative activity, antibody-dependent cell-mediated cytotoxicity and B cell depletion in human peripheral blood lymphocytes of healthy volunteers comparable to rituximab.4 There was only a significant difference in complement-dependent cytotoxicity (CDC) induction reported where veltuzumab was less than 2-fold more potent than rituximab.4 While statistically significant, we think that this small difference in EC50 values is not biologically relevant, in particular, as the authors demonstrate themselves by depletion of natural killer cells and neutrophils that the in vivo efficacy of veltuzumab in xenograft models relies on antibody-dependent effector cell function, but not on CDC. In light of these in vitro data the reported superiority of veltuzumab over rituximab in nonclinical xenograft models as well as a potential for clinical superiority is difficult to rationalize. We believe that it is not justified to claim clinical superiority of the 80 mg (for the subcutaneous formulation) or 80 mg/m2 (for intravenous infusions) weekly x 4 doses of veltuzumab over the standard rituximab regimen of 375 mg/mm2 based solely on a historical comparison of complete response (CR)/unconfirmed CR rates from non-randomized Phase 1/2 clinical trials in (relapsed/refractory) follicular non-Hodgkin lymphoma (NHL) patients as it is stated in the letter of Goldenberg and colleagues. In line with this, the cited publications by Morschhauser et al.5 and Negrea et al.6 do not claim superiority over rituximab. To claim superiority, the authors would have to provide clinical data where veltuzumab and rituximab are compared head-to-head at the respective (standard) doses. The dose and schedule of obinutuzumab (GA101) cannot be extrapolated from the rituximab dose and schedule due to the differences in the antibodies and their mechanisms of action. In the case of obinutuzumab, we have shown that by increasing the dose from 10 to 30 mg/kg in the SU-DHL4 diffuse large B cell lymphoma (DLBCL) nonclinical xenograft model we could induce complete tumor remissions, whereas increasing the dose of rituximab from 10 to 30 mg/kg did not result in enhanced efficacy and only slowed down tumor progression.7 These nonclinical data and data from Phase 2 clinical trials comparing doses of 400/400 mg vs 1600/ 800 mg obinutuzumab in relapsed/refractory iNHL and DLBCL patients8,9 support the rationale that in the case of obinutuzumab a flat dose of 1000 mg is studied in clinical trials. This may be particularly true for patients with higher tumor loads. We also want to point out that there is no clinical evidence available supporting the statement that higher doses of CD20 antibodies result in more pronounced side effects. In the best interest of patients, we believe that novel CD20 antibodies such as obinutuzumab, ofatumumab or veltuzumab have to be studied in clinical trials first, at the optimal dose for that antibody as determined in clinical Phase 1/2 studies, and second, in randomized clinical trials in direct comparison to the approved/standard of care rituximab based regimen. This belief is the basis of the development paradigm for obinutuzumab, which is currently being studied in combination with chemotherapy head-to-head to rituximab in three first line Phase 3 clinical trials in CLL, indolent NHL and DLBCL. Ultimately, only these clinical data can tell whether nonclinical findings translate into superior clinical benefit in patients with B cell malignancies.

GA101 P329G LALA, a variant of obinutuzumab with abolished ADCC, ADCP and CDC function but retained cell death induction, is as efficient as rituximab in B cell depletion and antitumor activity

CD20-antibodies are believed to mediate three different mechanisms of action (MOA): 1) direct cell death (DCD)/apoptosis, 2) antibody-dependent cellular cytotoxicity (ADCC) and phagocytosis (ADCP), 3) complement-dependent cytotoxicity (CDC). The relative contribution of these MOAs to the clinical

Abstract 3629: Engineering a novel asymmetric head-to-tail 2+1 T-cell bispecific (2+1 TCB) IgG antibody platform with superior T-cell killing compared to 1+1 asymmetric TCBs

T cell bispecific antibodies that recruit and engage T cells for tumor cell killing through binding to the T cell receptor (TCR) upon binding to a tumor antigen (TA) and subsequent crosslinking have attracted broad interest. Here, we describe a novel asymmetric head-to-tail 2+1 T cell bispecific antibody (2+1 TCB) platform characterized by the fusion of a flexible Fab fragment to the N-terminus of the CD3e Fab of a heterodimeric asymmetric bispecific TA-CD3e IgG1 antibody in head-to-tail geometry via a flexible linker. The resulting TCB is monovalent for CD3e (KD 70-100 nM) and binds bivalently with avidity to the TA on the target cell. Correct heavy chain pairing is enabled by knob-into-holes technology, correct light chain pairing by CrossMAb technology or using a common light chain. This enables production with standard processes in CHO cells. To exclude FcgR-mediated unspecific TCR and FcgR co-activation resulting in unspecific cytokine release, Fc- effector functions (ADCC, ADCP, CDC) are abolished by introduction of P329G LALA mutations while FcRn binding and IgG-like pharmacokinetic properties are retained as shown in mouse and Cynomolgus. For comparative profiling, the following TCBs were generated with specificity for the tumor antigens MCSP/CSPG4, FOLR1/FRalpha, CD19 and CD20: 2+1 TCBCD3-inside, 2+1 TCBCD3-outside, one-armed 1+1 TCBCD3-inside and a classical asymmetric 1+1 IgG TCB. In vitro Jurkat-NFAT, T cell killing, activation and proliferation assays show that both 2+1 TCB formats mediate superior potency of killing (for CSPG4, FOLR1, CD19, CD20) and superior absolute killing (for CSPG4, CD19) compared to the respective classical asymmetric 1+1 IgG TCB. Surprisingly, the 2+1 TCBCD3-inside format was found to be superior in potency compared to the 2+1 TCBCD3-outside format, although its binding affinity for CD3e is reduced. These data confirm that TCBs mediate extremely potent T cell killing with fM-pM EC50 values based on CD3e antibodies with affinities of only 70-100 nM. Notably, for CD19 both, 2+1 TCBCD3-inside and one-armed 1+1 TCBCD3-inside, mediate comparable potency and overall killing, and both were superior compared to the asymmetric 1+1 IgG TCB. These data underline the importance of the head-to-tail geometry with two Fabs on one arm attached to each other via a flexible G4S-linker. Finally, using 2+1 and 1+1 FOLR1 TCBs we demonstrate that bivalent binding allows better differentiation in killing of cells with high vs. low FOLR1 expression as compared to monovalent binding. Taken together, we demonstrate that the 2+1 TCBCD3-inside is the most potent, efficacious and versatile TCB design. Due to its orientation with the CD3e Fab inside, it allows the conversion of existing antibodies into potent TCBs without format restriction. Based on this platform, CEA CD3 TCB (RG7802, Phase I/Ib) and CD20 CD3 TCB (RG6026, Phase I) have entered clinical trials. Citation Format: Christian Klein, Christiane Neumann, Tanja Fauti, Tina Weinzierl, Anne Freimoser-Grundschober, Inja Waldhauer, Linda Fahrni, Sylvia Herter, Erwin van Puijenbroek, Sara Colombetti, Johannes Sam, Sabine Lang, Sherri Dudal, Wolfgang Schafer, Jorg T. Regula, Samuel Moser, Oliver Ast, Ralf Hosse, Ekkehard Mossner, Peter Brunker, Marina Bacac, Pablo Umana. Engineering a novel asymmetric head-to-tail 2+1 T-cell bispecific (2+1 TCB) IgG antibody platform with superior T-cell killing compared to 1+1 asymmetric TCBs [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3629. doi:10.1158/1538-7445.AM2017-3629

Abstract LB-236: M4-3-ML2, a novel glycoengineered humanized IgG1 antibody, targeting a membrane-proximal epitope of MCSP/CSPG4 exhibits potent ADCC induction in vitro and in vivo anti-tumoral efficacy in disseminated melanoma models

Proceedings: AACR 103rd Annual Meeting 2012‐‐ Mar 31‐Apr 4, 2012; Chicago, IL MCSP/CSPG4 is a large transmembrane proteoglycan identified in melanomas as HMW-MAA. In the mouse it is known as neurite growth factor 2 (NG2), a marker of pericyte recruitment. MCSP has been used as a target for clinical imaging of (uveal) melanomas by immunoscintigraphy. MCSP shows uniform and abundant expression in ca. 60-80% of melanoma, and was described in lobular breast carcinoma, glioblastoma, osteo- & chondrosarcoma, and basal cell carcinoma. It is present at high levels on pericytes of tumor neovasculature, but down-regulated as vessels mature. Normal tissue expression is low and it is not detected on PBMCs. We have generated human/Cynomolgus cross-reactive antibodies against a membrane-proximal MCSP epitope by mouse immunization with a linear peptide derived from the membrane proximal D3 domain followed by boosting with melanoma cells. The mouse antibody LC007 was selected for humanization due to its potent induction of ADCC as a chimeric antibody, compared to antibodies to membrane distal epitopes of MCSP. LC007 as chimeric IgG1 and its humanized IgG1 derivative M4-3-ML2 are characterized by the following properties: i) Specific binding to the native epitope on MCSP+ melanoma cells, but no induction of internalization; ii) Specific IHC staining of MCSP+ cells in FFPET samples; iii) ca 10 nM monovalent affinity for hMCSP D3 domain. Moreover, glycoengineering of LC007 and M4-3-ML2 antibodies using GlycoMab technology resulted in increased binding affinity for hFcgRIIIa and enhanced ADCC potency and absolute killing of melanoma cell lines. As expected, neither up to 10 ug/mL wildtype, nor glycoengineered M4-3-ML2 induced relevant cytokine (IL-6, TNF-α, IFN-γ) release in human whole blood supporting that MCSP is not expressed there. Subsequently, we studied anti-tumoral efficacy of the chimeric antibody LC007 and the humanized antibody M4-3-ML2 in disseminated models of MV3 and MDA-MB435 melanoma after i.v. injection of tumor cells in hCD16 transgenic Scid mice, which express the functional human high affinity FcgRIIIa receptor on NK cells. Both, glycoengineered LC007 and M4-3-ML2 mediated efficacy in terms of enhanced median and overall survival in both disseminated xenograft models, and were superior to the respective non-glycoengineered antibodies. Taken together, our studies support MCSP/CSPG4 as an attractive target for antibody-based cancer immunotherapy. Further studies investigating the anti-angiogenic effect of MCSP antibodies via their action on pericytes/vascular smooth muscle cells are ongoing. 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 LB-236. doi:1538-7445.AM2012-LB-236