ErbB-2 and ErbB3 binding bispecific antibodies for use in treating cells carrying an NRG1 fusion gene - Patent Application 20070122997

Abstract

The present invention relates to the field of antibodies. In particular, the present invention relates to the field of therapeutic (human) antibodies for the treatment of ErbB-2 / ErbB-3 positive cells. More particularly, the present invention relates to the treatment of cells containing an NRG1 fusion gene that includes at least a portion of the NRG1 gene fused to sequences from a different chromosomal location.

Description

[Technical Field]

[0001] This application claims priority to European Patent Application No. 17164292.9, filed March 31, 2017, the contents of which are incorporated herein by reference. [Background technology]

[0002] The present invention relates to the field of antibodies. In particular, the present invention relates to the field of therapeutic (human) antibodies for the treatment of ErbB-2 / ErbB-3 positive cells. More particularly, the present invention relates to the treatment of tumors containing an NRG1 fusion gene, which contains at least a portion of the NRG1 gene fused to sequences from a different chromosomal location.

[0003] Neuregulin-1 (NRG1) has been proposed as a candidate oncogene and tumor suppressor gene. It encodes a ligand capable of binding to ErbB family receptors, making it highly likely to be involved in epithelial cancer. To date, there are over 16 soluble and transmembrane proteins derived from the NRG1 gene. Proteolytic processing of the extracellular domain of transmembrane NRG1 isoforms releases soluble factors. HRG1-β1 is one of the proteins encoded by this gene. It contains an Ig domain and an EGF-like domain required for direct binding to the receptor tyrosine kinases ErbB-3 and ErbB-4. The NRG1 gene and isoforms are known under several different aliases, including: neuregulin 1; pro-NRG1; HRGA; SMDF; HGL; GGF; NDF; NRG1 intronic transcript 2 (non-protein coding); heregulin, alpha (45 kD, ERBB2 P185-activator); acetylcholine receptor-inducing activity; pro-neuregulin-1, membrane-bound isoform; sensory and motor neuron-derived factor; Neu differentiation factor; glial growth factor 2; NRG1-IT2; MSTP131; MST131; ARIA; GGF2; HRG1; and HRG. The external identifiers for the NRG1 gene are HGNC:7997; Entrez Gene:3084; Ensembl:ENSG00000157168; OMIM:142445, and UniProtKB:Q02297.

[0004] NRG1 isoforms are formed by alternative splicing and include transmembrane, outer membrane-bound, shedding, secreted, or intracellular forms (Falls, 2003; Hayes and Gullick, 2008). They bind to ErbB-3 or ErbB-4, which likely signal as heterodimers with ErbB-2 (HER2). Although NRG1-encoded proteins are typically considered mitogens, they can also be potently proapoptotic: notably, expressing NRG1 in cells can trigger apoptosis of the expressing cells (Weinstein et al., 1998).

[0005] The NRG1 gene has been identified as a potential cancer-critical gene in two apparently contradictory contexts. First, it is a leading candidate for a potential major tumor suppressor gene located on chromosome 8p, the short arm of chromosome 8. Loss of chromosome 8p is one of the most frequent genomic events in epithelial cancers, including those of the breast, colon, bladder, and prostate. This has been successively demonstrated by loss of heterozygosity, comparative genomic hybridization (CGH), and array-CGH studies (see Birnbaum et al., 2003; Pole et al., 2006 for references). The classical interpretation of this loss of chromosome 8p is that it represents a tumor suppressor gene. Loss of chromosome 8p in carcinoma cell lines has been mapped using fluorescence in situ hybridization and array-comparative genomic hybridization (array-CGH). Most breaks were found to be proximal to or actually within NRG1, making NRG1 and genes immediately telomeric to it prime candidates for such tumor suppressors (Pole et al., 2006; Cooke et al., 2008). Second, NRG1 appears to be a target of chromosomal translocations in breast cancer and is therefore a potential oncogene (for review, see Chua et al., 2009).

[0006] In the present invention, it has been found that tumors carrying chromosome 8p alterations exhibit growth inhibition in response to exposure to a bispecific antibody comprising a first antigen-binding site capable of binding to the extracellular portion of ErbB-2 and a second antigen-binding site capable of binding to the extracellular portion of ErbB-3. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] WO2015 / 130173 [Patent Document 2] U.S. Patent Application Publication No. 20030078385 (Arathoon et al. - Genentech) [Patent Document 3] PCT application number PCT / NL2013 / 050294 (WO2013 / 157954A1)

Non-licensed literature

[0008]

Non-patent document 1

Non-patent document 2

Non-patent document 3

Non-patent document 4

Non-patented document 5

Non-patent document 6

Non-patent document 7

Non-patent document 8

Non-patented document 9

[0009] In one aspect, there is provided a method for treating an individual having ErbB-2 and ErbB-3 positive cells, the method comprising administering to the individual in need thereof a bispecific antibody comprising a first antigen-binding site capable of binding to the extracellular portion of ErbB-2 and a second antigen-binding site capable of binding to the extracellular portion of ErbB-3, wherein the cells comprise an NRG1 fusion gene comprising at least a portion of the NRG1 gene fused to sequences from a different chromosomal location. Typically, the cells comprise an NRG1 fusion gene comprising at least the 3' end of the NRG1 gene fused to 5' sequences from a different chromosomal location.

[0010] The cell can be a cancer cell. The cancer cell can be a cancer cell associated with an NRG1 fusion gene, for example, a cancer cell driven by an NRG1 fusion.

[0011] In another aspect, there is provided a method for treating an individual having or at risk of having an ErbB-2 and ErbB-3 positive tumor, comprising administering to an individual in need thereof a bispecific antibody comprising a first antigen-binding site capable of binding to the extracellular portion of ErbB-2 and a second antigen-binding site capable of binding to the extracellular portion of ErbB-3, wherein cells of the tumor comprise an NRG1 fusion gene comprising a portion of an NRG1 fusion fused to sequences from a different chromosomal location, e.g., a 5' sequence, e.g., the 3' end of the NRG1 gene.

[0012] An individual at risk of having an ErbB-2 and ErbB-3 positive tumor can be an individual who is in remission.

[0013] Preferably, the NRG1 fusion gene expresses a protein comprising the NRG1 EGF-like domain. Preferably, the NRG fusion is a fusion of NRG1 with a gene on human chromosome 8. Preferably, the gene on human chromosome 8 encodes a efflux protein or a plasma membrane-associated protein. Preferably, the NRG1 fusion gene is a fusion of the 3' end of the NRG1 gene with the 5' sequence of one of genes selected from the group consisting of CD74; DOC4; TNFRSF10B; CLU; VAMP2; SLC3A2; RBPMS; WRN; SDC4; KIF13B; SLECA2; PDE7A; ATP1B1; CDK1; BMPR1B; MCPH1 and RAB2IL1.

[0014] Preferably, the cells are epithelial cells. Preferably, the cells are breast cancer cells, ovarian cancer cells, lung cancer cells, such as non-small cell lung cancer cells, or metastases thereof.

[0015] Preferably, the tumor is of epithelial origin. Preferably, the tumor is a breast cancer, ovarian cancer, lung cancer, or a metastasis thereof.

[0016] The cell can be, for example, a cancer cell, such as an ovarian cancer cell, that contains a CLU-NRG1 fusion or a RAB2IL1-NRG1 fusion.

[0017] The cell can be, for example, a cancer cell, such as a breast cancer cell, that contains a DOC4-NRG1 fusion.

[0018] The cells may be cancer cells, such as NSCLC (lung) cancer cells, such as a subtype called invasive mucinous adenocarcinoma, which contain, for example, VAMP2-NRG1, RBPMS-NRG1, WRN-NRG1, SDC4-NRG1, SLEC3A2-NRG1, KIF13B-NRG1 or CD74-NRG1.

[0019] Preferably, the individual has undergone treatment that is targeted at EGFR inhibition, preferably with an EGFR binding antibody, preferably cetuximab.

[0020] Preferably, the method further comprises determining ErbB-1 cell surface receptor density, ErbB-2 cell surface receptor density, ErbB-3 cell surface receptor density, ErbB-4 cell surface receptor density, or a combination thereof, on cells of the tumor. Preferably, the cells or tumor have less than 400,000 ErbB-1 cell surface receptors per cell, preferably less than 200,000 ErbB-1 cell surface receptors per cell.

[0021] Preferably, the method further comprises the step of administering to the individual an ErbB-1 inhibitor, preferably cetuximab.

[0022] Preferably, in the methods disclosed herein, the ErbB-2 / ErbB-3 positive cells or tumors have fewer than 50,000 ErbB-3 cell surface receptors per cell.

[0023] Preferably, in the methods disclosed herein, the cell(s) of the tumor have a heregulin expression level that is higher than that of MCF7 cells.

[0024] As will be apparent to one of skill in the art, the bispecific antibodies disclosed herein are also for use in the preparation of medicaments and for use in therapy, as disclosed herein.

[0025] In particular, a bispecific antibody comprising a first antigen-binding site capable of binding to the extracellular portion of ErbB-2 and a second antigen-binding site capable of binding to the extracellular portion of ErbB-3, for use in treating an individual having ErbB-2 and ErbB-3 positive cells, the cells comprising an NRG1 fusion gene comprising at least the 3' end of the NRG1 gene fused to 5' sequences derived from a different chromosomal location.

[0026] The cell can be a cancer cell. The cancer cell can be a cancer cell associated with an NRG1 fusion, for example, a cancer cell driven by an NRG1 fusion.

[0027] The bispecific antibody is also for use in treating ErbB-2 / ErbB-3 positive tumors, the cells of which contain an NRG1 fusion gene comprising the 3' end of the NRG1 gene fused to 5' sequences from a different chromosomal location.

[0028] Preferably, in the methods and uses disclosed herein, the first antigen-binding site binds to domain I of ErbB-2 and the second antigen-binding site binds to domain III of ErbB-3, and preferably the affinity of the first antigen-binding site for ErbB-2 is lower than the affinity of the second antigen-binding site for ErbB-3. i) the antibody comprises at least the CDR1, CDR2 and CDR3 sequences of an ErbB-2-specific heavy chain variable region selected from the group consisting of MF2926, MF2930, MF1849; MF2973, MF3004, MF3958, MF2971, MF3025, MF2916, MF3991, MF3031, MF2889, MF2913, MF1847, MF3001, MF3003 and MF1898, or MF2930, MF1849; MF2973, MF3004, MF3958, MF2971, MF3025, MF2916, MF3991, MF3031, MF2889, MF2913, MF1847, MF3001, MF3003 or MF1898; and / or ii) the antibody comprises at least the CDR1, CDR2 and CDR3 sequences of an ErbB-3 specific heavy chain variable region selected from the group consisting of MF3178; MF3176; MF3163; MF3099; MF3307; MF6055; MF6056; MF6057; MF6058; MF6059; MF6060; MF6061; MF6062; MF6063; MF6064; MF6065; MF6066; MF6067; MF6068; MF6069; MF6070; MF6071; MF6072; MF6073 and MF6074; or MF6073 or MF6074. Preferably, the antibody comprises CDR sequences which differ by at most 3 amino acids, preferably at most 2 amino acids, preferably at most 1 amino acid from the CDR1, CDR2 and CDR3 sequences of i) An ErbB-2-specific heavy chain variable region sequence selected from the group consisting of the heavy chain variable region sequences of MF2926, MF2930, MF1849; MF2973, MF3004, MF3958, MF2971, MF3025, MF2916, MF3991, MF3031, MF2889, MF2913, MF1847, MF3001, MF3003, and MF1898, or the antibody comprises a heavy chain variable region sequence that differs by up to 15 amino acids from the heavy chain variable region sequence of MF2926, MF2930, MF1849; MF2973, MF3004, MF3958, MF2971, MF3025, MF2916, MF3991, MF3031, MF2889, MF2913, MF1847, MF3001, MF3003, or MF1898; and / or ii) comprises an ErbB-3-specific heavy chain variable region sequence selected from the group consisting of the heavy chain variable region sequences of MF3178; MF3176; MF3163; MF3099; MF3307; MF6055; MF6056; MF6057; MF6058; MF6059; MF6060; MF6061; MF6062; MF6063; MF6064; MF6065; MF6066; MF6067; MF6068; MF6069; MF6070; MF6071; MF6072; MF6073 and MF6074; Alternatively, the antibody comprises a heavy chain variable region sequence that differs by up to 15 amino acids from the heavy chain variable region sequence of MF3178; MF3176; MF3163; MF3099; MF3307; MF6055; MF6056; MF6057; MF6058; MF6059; MF6060; MF6061; MF6062; MF6063; MF6064; MF6065; MF6066; MF6067; MF6068; MF6069; MF6070; MF6071; MF6072; MF6073 or MF6074. Preferably, the antibody comprises at least the CDR1, CDR2 and CDR3 sequences of the ErbB-2-specific heavy chain variable region MF3958, and the antibody comprises at least the CDR1, CDR2 and CDR3 sequences of the ErbB-3-specific heavy chain variable region MF3178. Preferably, the bispecific antibody comprises a "heavy chain for erbB-2 binding" shown in Sequence Listing Part 1D and a "heavy chain for erbB-3 binding" shown in Sequence Listing Part 1D.

[0029] Preferably, the first antigen-binding site and the second antigen-binding site comprise a light chain variable region comprising an IgVKl-39 gene segment, most preferably a rearranged germline human kappa light chain IgVKl-39*01 / IGJKl*01 or IgVκ1-39*01 / IGJκ5*01. Preferably, the light chain variable region comprises a CDR1 having the sequence (RASQSISSYLN), a CDR2 having the sequence (AASSLQS), and a CDR3 having the sequence (QQSYSTPPT). DETAILED DESCRIPTION OF THE INVENTION

[0030] An NRG1 fusion gene comprises at least a portion of the NRG1 gene fused to sequence from a different chromosomal location. "At least a portion" indicates that the entire NRG-1 gene or a portion thereof can be present in the fusion. The fusion preferably comprises at least the coding sequence of exons 6, 7, and 8. Another way to define the NRG1 portion in an NRG1 fusion gene is that it comprises the EGF-like domain of NRG1. At least a portion of the NRG1 gene can be fused to sequence from a different chromosomal location such that the sequence is located 5' or 3' to at least a portion of the NRG1 gene.

[0031] Preferably, the 3' end of the NRG1 gene can be fused to 5' sequences from a different chromosomal location. The NRG1 gene encodes various isoforms of NRG1. The various isoforms and their predicted functions are described in Adelaide et al. (2003). The GGF and GGF2 isoforms contain a kringle-like sequence plus Ig and EGF-like domains; the SMDF isoform shares only the EGF-like domain with other isoforms. The EGF-like domain is encoded by the 3' end of the gene. The EGF-like domain is present in all NRG1 fusion genes of the present invention. Fusions have been found in which the 5' sequences from different chromosomal locations contain at least one extracellular domain, as well as an export signal and / or transmembrane domain of a plasma membrane protein. One example is the CD74-NRG1 fusion. The 5' sequences from different chromosomal locations may contain sequences that activate transcription of NRG1, such as promoters or enhancers. The 5' sequences are typically sequences from genes other than NRG1. This sequence may include a coding region, an expression control sequence, such as a promoter or enhancer, or a combination thereof. The NRG fusion includes a 5' sequence from a different location, which may be from a different chromosome or from another part of chromosome 8. In a preferred embodiment, the 5' sequence is from a gene on human chromosome 8.

[0032] The 3' end of the NRG1 gene, e.g., the NRG-1 gene, in the fusion preferably contains at least the coding sequence of exons 6, 7, and 8. Another way to define the NRG1 portion of the NRG1 fusion gene is that it contains the EGF-like domain of NRG1. This domain is encoded by the 3' end of the NRG1 gene (exons 6-8) and is required for binding to ErbB-3. The NRG1 fusion contains the coding region for this EGF-like domain in-frame at the 3' end of the fusion. The EGF-like domain is typically a sequence of approximately 30-40 amino acid residues in length, and its prototype is found in the sequence of epidermal growth factor (EGF) [PMID: 2288911, PMID: 6334307, PMID: 1522591, PMID: 6607417, PMID: 3282918, PMID: 11498013]. It is known to exist in more or less conserved forms in numerous other, mostly animal, proteins. A common feature of EGF-like domains is that they are found in the extracellular domains of membrane-bound proteins or in proteins known to be secreted (exception: prostaglandin G / H synthase). EGF domains typically contain six cysteine ​​residues (in EGF) that have been shown to participate in disulfide bonds. Its primary structure is a two-stranded beta-sheet followed by a loop to a short C-terminal two-stranded sheet. The subdomains between the conserved cysteines vary in length.

[0033] The NRG1 fusion gene is preferably a fusion of the 3' end of the NRG1 gene with the 5' sequence of one of the genes selected from the group consisting of CD74; DOC4; TNFRSF10B; CLU; VAMP2; SLC3A2; RBPMS; WRN; SDC4; KIF13B; SLECA2; PDE7A; ATP1B1; CDK1; BMPR1B; MCPH1 and RAB2IL1.

[0034] An NRG1 fusion gene can be a fusion of at least a portion of the NRG1 gene with a sequence from a different chromosomal location located 3' to the NRG1 gene. Such an NRG1 fusion gene can be a fusion of at least a portion of the NRG1 gene with a sequence from a different chromosomal location located 3' to the NRG1 gene, such as CD74, STMN2, PMEPA1, PROSC, or PSAP. The receptors for all NRG1 isoforms are tyrosine kinase transmembrane receptors of the ErbB family. This family is also called the human epidermal growth factor (EGF) receptor family (HER). This family has four members: ErbB (erythroblastoma)-1, ErbB-2, ErbB-3, and ErbB-4. The epidermal growth factor (EGF) receptor (EGFR, ErbB1, or HER1). These receptors (reviewed in Yarden and Pines, 2012) are widely expressed on epithelial cells. Upregulation of HER receptors or their ligands, such as heregulin (HRG) or epidermal growth factor (EGF), is a frequent event in human cancers (Wilson, Fridlyand, et al., 2012). In particular, overexpression of ErbB-1 and ErbB-2 occurs in epithelial tumors and is associated with tumor invasion, metastasis, chemotherapy resistance, and poor prognosis (Zhang, Berezov, et al., 2007). In normal breast tissue, ErbB-3 has been shown to be important in the growth and differentiation of luminal epithelium. For example, loss / inhibition of ErbB-3 results in selective expansion of basal over luminal epithelium (Balko, Miller, et al., 2012). Ligand binding to the extracellular domain of RTKs induces receptor dimerization, both between the same (homodimerization) and different (heterodimerization) receptor subtypes. Dimerization can activate the intracellular tyrosine kinase domain, which undergoes autophosphorylation and can then activate several downstream pro-growth signaling pathways, including those mediated by mitogen-activated protein kinases (MAPKs) and the pro-survival pathway Akt (reviewed in Yarden and Pines, 2012).No specific endogenous ligand has been identified for ErbB-2, and therefore it is generally assumed to signal via heterodimerization (Sergina, Rausch, et al., 2007). ErbB-3 can be activated by the association of its ligands. These ligands include, but are not limited to, neuregulin (NRG) and heregulin (HRG).

[0035] ErbB-1 is known by various names, the most common of which is EGFR. EGFR has an extracellular domain (ECD) composed of four subdomains, two of which are involved in ligand binding and two of which are involved in homodimerization and heterodimerization. EGFR integrates extracellular signals from various ligands to generate diverse intracellular responses. The major signaling pathway activated by EGFR consists of the Ras-mitogen-activated protein kinase (MAPK) mitogenic signaling cascade. Activation of this pathway is initiated by recruitment of Grb2 to tyrosine-phosphorylated EGFR. This leads to activation of Ras via the Grb2-bound Ras-guanine nucleotide exchange factor, son of sevenless (SOS). Furthermore, the PI3-kinase-Akt signaling pathway is also activated by EGFR, but this activation is significantly stronger in the presence of coexpression of ErbB-3 (HER3). EGFR is associated with several human epithelial malignancies, particularly cancers of the breast, bladder, non-small cell lung, colon, ovary, head and neck, and brain. Activating mutations in this gene, resulting in an autocrine activation loop, and overexpression of the receptor and its ligand have been found. Therefore, this RTK has been widely used as a target for cancer therapy. Small molecule inhibitors targeting the RTK and monoclonal antibodies (mAbs) against the extracellular ligand-binding domain have both been developed and have shown some clinical success to date, mostly in select groups of patients. The database accession number for the human EGFR protein and its encoding gene is (GenBank NM_005228.3). This accession number is provided primarily to provide further methods of identification of the EGFR protein as a target; the actual sequence of the EGFR protein to which the antibody binds may vary due to mutations in the encoding gene, such as those present in some cancers.

[0036] The words cancer and tumor are typically used herein to refer to both cancers unless specifically stated otherwise.

[0037] When reference is made herein to EGFR, unless otherwise specified, this reference refers to human EGFR. The antigen-binding site that binds to EGFR binds to EGFR and its various variants, such as those expressed on some EGFR-positive tumors.

[0038] The term "ErbB-3," as used herein, refers to the protein encoded by the ERBB3 gene in humans. Alternative names for this gene or protein are HER3; LCCS2; MDA-BF-1; c-ErbB-3; c-ErbB3; ErbB3-S; p180-ErbB3; p45-sErbB3; and p85-sErbB3. When reference is made herein to ErbB-3, this reference refers to human ErbB-3. Antibodies comprising an antigen-binding site that binds to ErbB-3 bind to human ErbB-3. Due to the similarity of sequence and tertiary structure between human and other mammalian orthologs, ErbB-3 antigen-binding sites may, but do not necessarily, also bind to such orthologs. The database accession numbers for the human ErbB-3 protein and the gene encoding it are (NP_001005915.1, NP_001973.2, NC_000012.11, NC_018923.2, NT_029419.12). These accession numbers are provided primarily to provide further methods of identifying ErbB-3 as a target; the actual sequence of the ErbB-3 protein to which the antibody binds may vary due to mutations in the encoding gene, such as those present in some cancers. The ErbB-3 antigen-binding site binds to ErbB-3 and its various variants, such as those expressed by some ErbB-3-positive tumor cells. The antigen-binding site that binds to ErbB-3 preferably binds to domain III of ErbB-3.

[0039] The term "ErbB-2," as used herein, refers to the protein encoded by the ERBB-2 gene in humans. Alternative names for this gene or protein include CD340; HER-2; HER-2 / neu; MLN 19; NEU; NGL; and TKR1. The ERBB-2 gene is often referred to as HER2 (for human epidermal growth factor receptor 2). When reference is made to ErbB-2 herein, this reference refers to human ErbB-2. Antibodies comprising an antigen-binding site that binds to ErbB-2 bind to human ErbB-2. Due to sequence and tertiary structure similarities between human and other mammalian orthologs, ErbB-2 antigen-binding sites may, but do not necessarily, bind to such orthologs. The database accession numbers for the human ErbB-2 protein and the gene encoding it are (NP_001005862.1, NP_004439.2, NC_000017.10, NT_010783.15, NC_018928.2). These accession numbers are provided primarily to provide further methods of identifying ErbB-2 as a target; the actual sequence of the ErbB-2 protein to which the antibody binds may vary due to mutations in the encoding gene, such as those present in some cancers. The ErbB-2 antigen-binding site binds to ErbB-2 and its various variants, such as those expressed by some ErbB-2-positive tumor cells. The antigen-binding site that binds to ErbB-2 preferably binds to domain I of ErbB-2.

[0040] CD74 is known by several aliases, including CD74 molecule; CD74 antigen (invariant polypeptide of the major histocompatibility complex, class II antigen-associated); CD74 molecule, major histocompatibility complex, class II invariant chain; HLA-DR antigen-associated invariant chain; class II antigen gamma chain; Ia-associated invariant chain; MHC HLA-DR gamma chain; HLA-DR-gamma; DHLAG; P33; HLA class II histocompatibility antigen gamma chain; Ia antigen-associated invariant chain; Ia-GAMMA; and HLADG. The external IDs for CD74 are HGNC: 1697; Entrez Gene: 972; Ensembl: ENSG00000019582; OMIM: 142790; and UniProtKB: P04233.

[0041] DOC4 or Teneurin transmembrane protein 4 (TENM4) is known under several different names, such as protein Odd Oz / Ten-M homolog 4; tenascin-M4; Ten-M4; Ten-4; ODZ4; TNM4; Odz, Odd Oz / Ten-M homolog 4 (Drosophila); Odz, Odd Oz / Ten-M homolog 4; teneurin-4; KIAA1302; Doc4; and ETM5. The external IDs for DOC4 are HGNC:29945; Entrez Gene:26011; Ensembl:ENSG00000149256; OMIM:610084 and UniProtKB:Q6N022.

[0042] TNFRSF10B or TNF receptor superfamily member 10b is known under several different names: tumor necrosis factor receptor superfamily, member 10b; TNF-related apoptosis-inducing ligand receptor 2; death receptor 5; TRAIL-R2; TRAILR2; KILLER; TRICK2; ZTNFR9; DR5; p53-regulated DNA damage-inducible death receptor (Killer); tumor necrosis factor receptor superfamily member 10B; tumor necrosis factor receptor-like protein ZTNFR9; death domain-containing receptor for TRAIL / Apo-2L; apoptosis-inducing protein TRICK2A / 2B; apoptosis-inducing receptor TRAIL-R2; cytotoxic TRAIL receptor-2; Fas-like protein; TRAIL receptor 2; CD262 antigen; KILLER / DR5; TRICK2A; TRICK2B; TRICKB; and CD262. The external Ids of TNFRSF10B are HGNC:11905; Entrez Gene:8795; Ensembl:ENSG00000120889; OMIM:603612; and UniProtKB:O14763.

[0043] The CLU gene or clusterin is known under several different names, for example, testosterone-suppressed prostate message 2; apolipoprotein J; complement-related protein SP-40,40; complement cytolysis inhibitor; complement lysis inhibitor; sulfated glycoprotein 2; Ku70-binding protein 1; NA1 / NA2; TRPM-2; APO-J; APOJ; KUB1; CLI; clusterin (complement lysis inhibitor, SP-40,40, sulfated glycoprotein 2, testosterone-suppressed prostate message 2, apolipoprotein J); aging-related gene 4 protein; aging-related protein 4; SGP-2; SP-40; TRPM2; AAG4; CLU1; CLU2; and SGP2. The external IDs of CLU are HGNC:2095; Entrez Gene:1191; Ensembl:ENSG00000120885; OMIM:185430; and UniProtKB:P10909.

[0044] VAMP2, or vesicle-associated membrane protein 2, is known under several different names, such as synaptobrevin 2; SYB2; vesicle-associated membrane protein 2; and synaptobrevin-2. The external IDs for VAMP2 are HGNC: 12643; Entrez Gene: 6844; Ensembl: ENSG00000220205; OMIM: 185881; and UniProtKB: P63027.

[0045] SLCA3A2 or solute carrier family 3 member 2 is known under several different names, such as lymphocyte activation antigen 4F2 large subunit; solute carrier family 3 (activator of dibasic and neutral amino acid transport), member 2; antigen identified by monoclonal antibodies 4F2, TRA1.10, TROP4, and T43; solute carrier family 3 (amino acid transport heavy chain), member 2; 4F2 cell surface antigen heavy chain; CD98 heavy chain; 4F2HC; MDU1; antigen defined by monoclonal antibody 4F2, heavy chain; antigen defined by monoclonal antibody 4F2; 4F2 heavy chain antigen; 4F2 heavy chain; CD98 antigen; CD98HC; 4T2HC; NACAE; CD98, and 4F2. The external IDs for SLC3A2 are HGNC:11026; Entrez Gene:6520; Ensembl:ENSG00000168003; OMIM:158070; and UniProtKB:P08195.

[0046] RBPMS, or RNA-binding proteins with multiple splicing, is known under several different names, such as RNA-binding proteins with multiple splicing; heart and RRM expressed sequence; HERMES; RNA-binding proteins with multiple splicing; and RBP-MS. The external IDs for RBPMS are HGNC:19097; Entrez Gene:11030; Ensembl:ENSG00000157110; OMIM:601558; and UniProtKB:Q93062.

[0047] WRN, or Werner syndrome RecQ-like helicase, is known under several different names, such as Werner syndrome RecQ-like helicase; DNA helicase, RecQ-like type 3; RecQ protein-like 2; exonuclease WRN; RECQL2; RECQ3; Werner syndrome ATP-dependent helicase; Werner syndrome, RecQ helicase-like; Werner syndrome; EC 3.6.4.12; EC 3.1.-.-; EC 3.6.1; and RECQL3. The external IDs for WRN are HGNC:12791; Entrez Gene:7486; Ensembl:ENSG00000165392; OMIM:604611 and UniProtKB:Q14191.

[0048] SDC4 or syndecan-4 is known under several different names, such as syndecan-4 (amphiglycan, ryudocan); syndecan proteoglycan 4; ryudocan core protein; amphiglycan; SYND4; ryudocan amphiglycan; and syndecan-4. The external IDs for SDC4 are HGNC:10661; Entrez Gene:6385; Ensembl:ENSG00000124145; OMIM:600017; and UniProtKB:P31431.

[0049] Various NRG1 fusion genes are described in Dhanasekaran et al. (2014).

[0050] The present invention provides a method of treating an individual having ErbB-2 and ErbB-3 positive cells or tumors. Alternatively, the individual may be at risk of having the tumor. The method comprises administering to an individual in need thereof a bispecific antibody comprising a first antigen-binding site capable of binding to the extracellular portion of ErbB-2 and a second antigen-binding site capable of binding to the extracellular portion of ErbB-3. The method is characterized in that a cell(s) of the tumor contain an NRG1 fusion gene comprising the 3' end of the NRG1 gene fused to a 5' sequence derived from a different chromosomal location.

[0051] The cell can be a cancer cell, which can be a cancer cell associated with an NRG1 fusion, e.g., a cancer cell driven by an NRG1 fusion.

[0052] The antigen-binding site in an antibody is typically located in the variable domain, which comprises a heavy chain variable region and a light chain variable region.

[0053] Preferably, the individual has undergone treatment aimed at inhibiting EGFR, preferably with an EGFR binding antibody, preferably cetuximab.

[0054] The methods of treatment of the present invention preferably further comprise determining the number of ErbB-1 cell surface receptors; ErbB-2 cell surface receptors; ErbB-3 cell surface receptors; ErbB-4 cell surface receptors or combinations thereof on the cells(s) of the tumor.

[0055] The treatment methods of the present invention preferably further include a step of determining whether cells contain NRG1 fusions or whether tumors contain cells with NRG1 fusions. This can be performed, for example, on biopsied cells. Various methods are available, many of which are known in the art. In the case of NRG1 fusions, the region where chromosomal breakage occurs is known, so determining whether a tumor contains such an NRG1 fusion is routine for those skilled in the art. One method is PCR amplification using primers that span the junction in the NRG1 fusion. This can be easily performed for NRG1 fusions that are known to exist. Novel fusions can also be easily detected. For example, a suitable method is junction cloning technology, such as that used to find integration sites of retroviral genomes. A suitable method is LAM-PCR. See Schmidt et al., Nature Methods 4, 1051-1057 (2007) doi:10.1038 / nmeth1103, and the specific reference to LAM-technology therein.

[0056] The methods of treatment of the present invention are preferably characterized in that the cells or tumors have fewer than 400,000 ErbB-1 cell surface receptors per cell, preferably fewer than 200,000 ErbB-1 cell surface receptors per cell.

[0057] In a preferred embodiment, the methods of treatment of the present invention further comprise the step of administering to the individual an ErbB-1 inhibitor, preferably cetuximab.

[0058] The method of treatment defined herein may also be defined as a compound or combination of compounds for use in the treatment of. A suitable combination of compounds is a bispecific antibody as defined herein and an ErbB-1 inhibitor.

[0059] To determine whether a cell or tumor is positive for ErbB-2 and ErbB-3, a skilled artisan can, for example, determine the amplification and / or immunohistochemical staining of ErbB-2 and ErbB-3. At least 10% of tumor cells in a biopsy should be positive for both ErbB-2 and ErbB-3. A biopsy may also contain 20%, 30%, 40%, 50%, 60%, 70% or more positive cells. ErbB-1 positive tumors can be identified similarly.

[0060] Preferably, the positive cancer is breast cancer, e.g., early-stage breast cancer. However, the present invention is applicable to a wide range of ErbB-2, ErbB-3, or ErbB-2 / ErbB-3 positive cancers, such as gastric cancer, colorectal cancer, colon cancer, gastroesophageal cancer, esophageal cancer, endometrial cancer, ovarian cancer, breast cancer, liver cancer, lung cancer including non-small cell lung cancer, clear cell sarcoma, salivary gland cancer, head and neck cancer, brain cancer, bladder cancer, pancreatic cancer, prostate cancer, kidney cancer, skin cancer, melanoma, etc. The cells are preferably epithelial cells. Alternatively, the cells or tumors are preferably cells or tumors of epithelial origin. In preferred embodiments, the cells or tumors are breast cancer, ovarian cancer, lung cancer, or metastases thereof. Preferably, the tumors are of epithelial origin. Preferably, the tumors are breast cancer, ovarian cancer, lung cancer, or metastases thereof.

[0061] Patients with ErbB2-positive cells or tumor cells can be classified based on the number of ErbB-2 receptors on the tumor cell surface. Tumors with more than 1,000,000 ErbB-2 receptors on their cell surface are typically classified as ErbB-2[+++], tumors with between 150,000 and 1,000,000 ErbB-2 receptors are classified as ErbB-2[++], and tumors with less than 150,000 ErbB-2 receptors are classified as ErbB-2[+]. Preferably, patients are classified as ErbB-2[++] or ErbB-2[+++]. Preferably, ErbB-2 / ErbB-3-positive tumors have at least 1,000,000 ErbB-2 cell surface receptors per cell.

[0062] Preferably, methods are provided wherein the ErbB-2 / ErbB-3 positive cells or tumors have at least 150,000 ErbB-2 cell surface receptors per cell and fewer than 50,000 ErbB-3 cell surface receptors per cell. Preferably, methods are provided wherein the ErbB-2 / ErbB-3 positive cells or tumors have at least 1,00,000 ErbB-2 cell surface receptors per cell and fewer than 50,000 ErbB-3 cell surface receptors per cell.

[0063] In some embodiments, the methods disclosed herein are advantageous in that a particular patient population is first determined based on, for example, ErbB-1, ErbB-2, and / or ErbB-3 cell surface receptor density. Thus, the methods disclosed herein preferably include determining the ErbB-1 cell surface receptor density, ErbB-2 cell surface receptor density, ErbB-3 cell surface receptor density, and / or ErbB-4 cell surface receptor density for the cells or tumor. As used herein, the term cell surface receptor density refers to the number of receptors present on the cell surface per cell.

[0064] Preferably, the methods disclosed herein further comprise determining ErbB-2 cell surface receptor density for the cells or tumor. Patients can be classified using immunohistochemistry or fluorescent in situ hybridization. The HercepTest™ and / or HER2 FISH (pharm Dx™), both marketed by Dako Denmark A / S, and / or the HERmark® assay, marketed by Monogram Biosciences, are examples of suitable assays for determining ErbB-2 or ErbB-3 cell surface receptor density. Other methods for determining ErbB-2 receptor cell density are well known to those skilled in the art. In vivo methods for determining ErbB-2 are also known. See, for example, Chernomoridik et al., Mol Imaging. 2010 Aug;9(4):192-200 and Ardeshirpour et al., Technol Cancer Res Treat. 2014 Oct;13(5):427-434. Preferably, the methods disclosed herein further comprise determining ErbB-2 cell surface receptor density for the cells or tumor. Such methods are known to those skilled in the art (see, e.g., van der Woning and van Zoelen Biochem Biophys Res Commun. 2009 Jan. 9;378(2):285-9). Preferably, the methods disclosed herein further comprise determining ErbB-1 cell surface receptor density for the cells or tumor. Such methods are known to those skilled in the art (see, e.g., EGFR pharmDx™ Kit (Dako) and McDonagh et al. Mol Cancer Ther 2012;11:582). Similar methods can be used to determine ErbB-4 cell surface receptor density.

[0065] In some embodiments, ErbB-1, ErbB-2, ErbB-3 and ErbB-4 cell surface receptor densities are determined by FACS analysis on biopsied tumor cells.

[0066] Preferably, ErbB-2 / ErbB-3 positive cells or tumor cells have a relatively high level of heregulin expression. Heregulin is a growth factor involved in the growth of ErbB3 positive cells or tumor cells. Typically, when cells or tumor cells express high levels of heregulin (called heregulin stress), currently known treatments such as trastuzumab, pertuzumab, and lapatinib are no longer able to inhibit cell or tumor growth. This phenomenon is called heregulin resistance. In particular, the heregulin expression level is higher than that of MCF7 cells. Heregulin expression levels are measured, for example, using qPCR using cell or tumor RNA (see, for example, Shames et al., PLOS ONE, February 2013, Vol. 8, No. 2, pp. 1-10 and Yonesaka et al., Sci. Transl. Med., Vol. 3, No. 99 (2011); pp. 1-11), or using protein detection methods such as ELISA, preferably using blood, plasma, or serum samples (see, for example, Yonesaka et al., Sci. Transl. Med., Vol. 3, No. 99 (2011); pp. 1-11).

[0067] High heregulin levels are typically present during the formation of metastases (i.e., migration, invasion, growth, and / or differentiation of cells or tumor cells or tumor-initiating cells). Typically, tumor-initiating cells are identified based on stem cell markers, such as CD44, CD24, CD133, and / or ALDH1. Therefore, these processes can barely be countered by currently known treatments, such as trastuzumab and pertuzumab. The bispecific antibodies disclosed herein are capable of countering the formation of metastases in subjects with cell tumors containing an NRG1 fusion gene that includes the 3' end of the NRG1 gene fused to 5' sequences from a different chromosomal location.

[0068] Thus, there is further provided a method for combating the formation of metastases in a subject having ErbB-2, ErbB-3 or ErbB-2 / ErbB-3 positive cells or tumors, comprising the step of administering to the subject a bispecific antibody comprising a first antigen-binding site that binds ErbB-2 and a second antigen-binding site that binds ErbB-3, wherein the ErbB-2, ErbB-3 or ErbB-2 / ErbB-3 positive cells or tumor cells have a heregulin expression level that is at least 60%, preferably at least 70%, more preferably at least 80%, more preferably at least 85%, more preferably at least 90% or 95% of the heregulin expression level of BXPC3 or MCF7 cells. Also provided is a bispecific antibody comprising a first antigen-binding site that binds ErbB-2 and a second antigen-binding site that binds ErbB-3 for use in treating or preventing the formation of metastases, wherein ErbB-2, ErbB-3 or ErbB-2 / ErbB-3 positive cells or tumor cells have a heregulin expression level that is at least 60%, preferably at least 70%, more preferably at least 80%, more preferably at least 85%, more preferably at least 90% or 95% of the heregulin expression level of BXPC3 or MCF7 cells. Further provided is the use of a bispecific antibody according to the invention for the preparation of a medicament for the treatment or prevention of formation of metastases, wherein ErbB-2, ErbB-3 or ErbB-2 / ErbB-3 positive cells or tumor cells have a heregulin expression level that is at least 60%, preferably at least 70%, more preferably at least 80%, more preferably at least 85%, more preferably at least 90% or 95% of the heregulin expression level of BXPC3 or MCF7 cells.

[0069] The subject is preferably a human subject. The subject is preferably eligible for monoclonal antibody therapy using an ErbB-2 specific antibody, such as trastuzumab.

[0070] The amount of bispecific antibody administered to a patient typically falls within the therapeutic window, meaning that a sufficient amount is used to achieve a therapeutic effect, but does not exceed a threshold that causes unacceptable side effects. The lower the amount of antibody required to achieve the desired therapeutic effect, the larger the therapeutic window typically is. The selected dosage level depends on various factors, including the route of administration, timing of administration, rate of elimination of the specific compound used, duration of treatment, other drugs, compounds and / or materials used in combination, the age, sex, weight, condition, overall health and previous medical history of the patient being treated, and similar factors well known in the medical field. The dosage can be within the range of the dosing regime for trastuzumab, or can be lower.

[0071] Bispecific antibodies can be formulated as pharmaceutical compositions comprising a pharmaceutically acceptable carrier, diluent, or excipient, and further optional active agents. Antibodies and compositions comprising the antibodies can be administered by any route, including parenteral, enteral, and topical administration. Parenteral administration is typically by injection, including, for example, intravenous, intramuscular, intraarterial, intrathecal, intraventricular, intracapsular, intraorbital, intracardiac, intradermal, intraperitoneal, transtracheal, subcutaneous, subcuticular, intraarticular, subcapsular, subarachnoid, intraspinal, intracerebrospinal, intratumoral, and intrasternal injection and infusion.

[0072] In a preferred embodiment, an ErbB-1 inhibitor can be combined with treatment with a bispecific antibody disclosed herein. The ErbB-1 inhibitor can be administered simultaneously with the bispecific antibody or sequentially. Treatment with the ErbB-1 inhibitor can be separated from treatment with the bispecific antibody by minutes, hours, or days. Preferably, the ErbB-2 / ErbB3 cells or tumors are also positive for ErbB1. Preferably, the combined treatment (combined therapy) is appropriate for ErbB-2 / ErbB3 cells or tumors having more than 5,000 surface receptors per cell, preferably at least 20,000 surface receptors per cell, and more preferably more than 50,000 surface receptors per cell.

[0073] Suitable ErbB-1 inhibitors are known in the art and refer to compounds that inhibit at least one biological activity of ErbB-1 (EGFR), particularly compounds that reduce the expression or signaling activity of ErbB-1. Preferred ErbB-1 inhibitors bind to the extracellular binding site of this tyrosine kinase receptor molecule and block the binding of natural ligands such as EGF. Such inhibitors include antibodies, antibody portions, and epitope-containing peptides that target this extracellular EGF receptor-binding domain. Preferably, the ErbB-1 inhibitor is an anti-ErbB-1 antibody, preferably selected from cetuximab, matuzumab, necitumumab, nimotuzumab, panitumumab, or zalutumumab. The present invention further relates to ErbB-1 inhibitors that can bind to or interact with the intracellular phosphorylation site or domain of the tyrosine kinase receptor molecule and prevent or reduce phosphorylation by the tyrosine kinase. This can be achieved by small (chemical) molecule drugs. Preferred inhibitors include afatinib, erlotinib, gefitinib, lapatinib, osimertinib and neratinib.

[0074] The present disclosure provides bispecific antibodies for use in the methods and treatments described herein. Suitable bispecific antibodies comprise a first antigen-binding site that binds to ErbB-2 and a second antigen-binding site that binds to ErbB-3, and the bispecific antibodies can reduce or decrease ligand-induced receptor function of ErbB-2 and ErbB-3 on ErbB-3-positive cells. Preferred antibodies and their preparation are disclosed in WO2015 / 130173, which is incorporated herein by reference. The examples in WO2015 / 130173 further describe some properties of the antibodies, such as ligand binding and epitope mapping.

[0075] As used herein, the term "antigen-binding site" refers to a site derived from, and preferably present on, a bispecific antibody capable of binding to an antigen. An unmodified antigen-binding site is typically formed by and present in the variable domains of an antibody. The variable domain comprises the antigen-binding site. The variable domain that binds to an antigen is the variable domain that comprises the antigen-binding site that binds to the antigen.

[0076] In one embodiment, an antibody variable domain comprises a heavy chain variable region (VH) and a light chain variable region (VL). The antigen-binding site may be present in the combined VH / VL variable domains, or in the VH region alone, or in the VL region alone. When the antigen-binding site is present in only one of the two regions of the variable domain, the counterpart variable region may contribute to folding and / or stability of the binding variable region, but does not contribute significantly to antigen binding itself.

[0077] As used herein, antigen binding refers to the typical binding ability of an antibody to its antigen. An antibody containing an antigen-binding site that binds to ErbB-2 binds to ErbB-2 and, under otherwise identical conditions, binds to the homologous receptors ErbB-1 and ErbB-4 of the same species at a lower binding capacity by at least 100 times. An antibody containing an antigen-binding site that binds to ErbB-3 binds to ErbB-3 and, under otherwise identical conditions, does not bind to the homologous receptors ErbB-1 and ErbB-4 of the same species. Considering that the ErbB family is a family of cell surface receptors, binding is typically assessed on cells expressing the receptor. Antibody binding to an antigen can be assessed in various ways. One method is to incubate the antibody with the antigen (preferably cells expressing the antigen), remove unbound antibody (preferably by a washing step), and detect the bound antibody with a labeled antibody that binds to the bound antibody.

[0078] Antigen binding by an antibody is typically mediated through the specific three-dimensional structure of both the antibody's complementary region and the antigen and variable domain, which precisely binds these two structures together (a lock-and-key interaction), as opposed to random, nonspecific attachment of the antibody. Because antibodies typically recognize an epitope of an antigen, and such epitopes may also exist in other compounds, an antibody according to the present invention that binds to ErbB-2 and / or ErbB-3 may also recognize other proteins if such other compounds contain the same epitope. Thus, the term "binding" does not exclude the binding of the antibody to another protein or proteins containing the same epitope. Such other proteins are preferably not human proteins. The ErbB-2 antigen-binding site and ErbB-3 antigen-binding site defined herein typically do not bind to other proteins on the membranes of cells in postnatal, preferably adult, humans. The bispecific antibodies disclosed herein are typically capable of binding to ErbB-2 and ErbB-3 with a binding affinity of at least 1×10e-6 M, as outlined in more detail below.

[0079] The term "interferes with binding," as used herein, means that the antibody is directed to an epitope on ErbB-3 and that the antibody competes with the ligand for binding to ErbB-3. The antibody may reduce ligand binding, displace the ligand if it is already bound to ErbB-3, or at least partially prevent the ligand from binding to ErbB-3, for example, through steric hindrance.

[0080] The term "antibody," as used herein, refers to a proteinaceous molecule, preferably one belonging to the immunoglobulin class of proteins, that contains one or more variable domains that bind to an epitope on an antigen, where such domains are derived from or share sequence homology with the variable domains of an antibody. Antibodies for therapeutic use preferably resemble as closely as possible the natural antibody of the subject to be treated (e.g., a human antibody for a human subject). Antibody binding can be expressed in terms of specificity and affinity. Specificity determines which antigen or its epitope is specifically bound by a binding domain. Affinity is a measure of the strength of binding to a particular antigen or epitope. Specific binding is defined as binding with an affinity (KD) of at least 1×10e-6 M, more preferably 1×10e-7 M, and more preferably greater than 1×10e-9 M. Typically, antibodies for therapeutic applications have an affinity of up to 1×10e-10 M or higher. Antibodies, such as bispecific antibodies of the present invention, contain the constant domain (Fc portion) of a natural antibody. The antibodies of the present invention are typically bispecific full-length antibodies, preferably of the human IgG subclass. Preferably, the antibodies disclosed herein are of the human IgG1 subclass. Such antibodies have good ADCC properties, favorable half-lives upon in vivo administration to humans, and CH3 engineering technologies exist that can provide engineered heavy chains that preferentially form heterodimers over homodimers upon co-expression in clonal cells.

[0081] The antibodies of the invention disclosed herein are preferably "full-length" antibodies. The term "full-length" is defined to include essentially complete antibodies, although such antibodies need not possess all of the functions of an intact antibody. For the avoidance of doubt, a full-length antibody comprises two heavy chains and two light chains. Each chain comprises a constant (C) region and a variable (V) region, which can be broken down into domains designated CH1, CH2, CH3, VH, and CL and VL. Antibodies bind to antigens via the variable domains contained in the Fab portion, and after binding, can interact with molecules and cells of the immune system via the constant domains, primarily via the Fc portion. The terms "variable domain," "VH / VL pair," and "VH / VL" are used interchangeably herein. Full-length antibodies according to the present invention encompass antibodies in which mutations may be present that provide desired characteristics. Such mutations should not result in the deletion of a substantial portion of any of these regions. However, antibodies in which one or several amino acid residues have been deleted without substantially altering the binding characteristics of the resulting antibody are encompassed within the term "full-length antibody." For example, an IgG antibody may have 1 to 20 amino acid residues inserted, deleted, or a combination thereof in the constant region. For example, if an antibody itself has low ADCC activity, the ADCC activity of the antibody may be improved by slightly modifying the antibody's constant region (Junttila, T.T., K. Parsons et al. (2010), "Superior In vivo Efficacy of Afucosylated Trastuzumab in the Treatment of HER2-Amplified Breast Cancer," Cancer Research 70(11):4481-4489).

[0082] Full-length IgG antibodies are preferred due to their favorable half-life and the need to be close to the fully autologous (human) molecule for immunogenicity reasons. The antibodies disclosed herein are preferably bispecific IgG antibodies, preferably bispecific full-length IgG1 antibodies. IgG1 is preferred due to its long circulating half-life in humans. To prevent any immunogenicity in humans, it is preferred that the bispecific IgG antibody is human IgG1.

[0083] The term "bispecific" (bs) means that one portion of an antibody (defined above) binds to one epitope on an antigen, while the second portion binds to a different epitope. The different epitopes are typically present on different antigens. The first and second antigens are, in effect, two different proteins. A preferred bispecific antibody is one that comprises portions of two different monoclonal antibodies and, as a result, binds to two different types of antigens. One arm of the bispecific antibody typically comprises the variable domain of one antibody, and the other arm comprises the variable domain of another antibody. The heavy chain variable regions of a bispecific antibody are typically different from each other, while the light chain variable regions are preferably the same. A bispecific antibody in which different heavy chain variable regions are associated with the same or a common light chain is also called a bispecific antibody with a common light chain.

[0084] Preferred bispecific antibodies can be obtained by co-expression of two different heavy chains and one common light chain in a single cell. If wild-type CH3 domains are used, co-expression of two different heavy chains and one common light chain results in three different species: AA, AB, and BB. To increase the percentage of the desired bispecific product (AB), CH3 engineering can be used, or in other words, heavy chains with compatible heterodimerization domains, as defined herein below, can be used.

[0085] The term "compatible heterodimerization domain," as used herein, refers to a protein domain that has been engineered such that engineered domain A' preferentially forms heterodimers with engineered domain B', and vice versa, while A'-A' and B'-B' homodimerization is diminished.

[0086] The term "common light chain" refers to a light chain that may be identical or may have some amino acid sequence differences, but the binding specificity of the full-length antibody is not affected. For example, it is possible to prepare and find light chains that are not identical but are still functionally equivalent, such as by introducing and testing conservative amino acid changes, changes in amino acids in regions that do not contribute, or only partially contribute, to binding specificity when paired with a heavy chain. The terms "common light chain," "common VL," "single light chain," and "single VL" are all used interchangeably herein, with or without the addition of the term "rearranged."

[0087] The common light chain (variable region) preferably has a germline sequence. Preferred germline sequences are light chain variable regions that are frequently used in the human repertoire and have good thermodynamic stability, yield, and solubility. In a preferred embodiment, the light chain comprises a light chain region comprising the amino acid sequence of the O12 / IgVκ1-39*01 gene segment shown in SEQ ID NO: 1C "Common Light Chain IGKV1-39 / jk1," with 0 to 10, preferably 0 to 5, amino acid insertions, deletions, substitutions, additions, or a combination thereof. IgVκ1-39 is an abbreviation for the immunoglobulin variable kappa 1-39 gene. This gene is also known as immunoglobulin kappa variable 1-39; IGKV139; IGKV1-39; O12a, or O12. The external identifiers for this gene are HGNC: 5740; Entrez Gene: 28930; Ensembl: ENSG00000242371. The variable regions of IGKV1-39 are listed in Sequence 1C. The V regions can be combined with one of five J regions. Sequence 1C describes two preferred sequences of IgVκ1-39 combined with a J region. The combined sequences are designated IGKV1-39 / jk1 and IGKV1-39 / jk5; alternative names are IgVκ1-39*01 / IGJκ1*01 or IgVκ1-39*01 / IGJκ5*01 (nomenclature according to the IMGT database world wide web at imgt.org).

[0088] The O12 / IgVκ1-39*01 constituting the light chain variable region is preferably a germline sequence. It is even more preferred that the IGJκ1*01 or / IGJκ5*01 constituting the light chain variable region is a germline sequence. In a preferred embodiment, the IGKV1-39 / jk1 or IGKV1-39 / jk5 light chain variable region is a germline sequence.

[0089] In a preferred embodiment, the light chain variable region comprises germline O12 / IgVκ1-39*01. In a preferred embodiment, the light chain variable region comprises kappa light chain IgVκ1-39*01 / IGJκ1*01 or IgVκ1-39*01 / IGJκ5*01. In a preferred embodiment, IgVκ1-39*01 / IGJκ1*01. The light chain variable region preferably comprises germline kappa light chain IgVκ1-39*01 / IGJκ1*01 or germline kappa light chain IgVκ1-39*01 / IGJκ5*01, preferably germline IgVκ1-39*01 / IGJκ1*01.

[0090] Obviously, those skilled in the art will recognize that "common" also refers to functional equivalents of light chains that are not identical in amino acid sequence. Many variants of the above light chains exist in which mutations (deletions, substitutions, additions) exist that do not substantially affect the formation of a functional binding region. The light chain may also be a light chain identified herein above with 1 to 5 amino acid insertions, deletions, substitutions, or combinations thereof.

[0091] Preferably, the first antigen-binding site and the second antigen-binding site both comprise a light chain variable region comprising a CDR1 having the sequence (RASQSISSYLN), a CDR2 having the sequence (AASSLQS), and a CDR3 having the sequence (QQSYSTPPT).

[0092] The term "ErbB-1," as used herein, refers to the protein encoded by the ERBB-1 gene in humans. Alternative names for this gene or protein include EGFR, ERBB, HER1, Erb-B2 receptor tyrosine kinase 1. When reference is made to ErbB-1 herein, the reference refers to human ErbB-1.

[0093] The term "ErbB-2," as used herein, refers to the protein encoded by the ERBB-2 gene in humans. Alternative names for this gene or protein include CD340; HER-2; HER-2 / neu; MLN 19; NEU; NGL; and TKR1. The ERBB-2 gene is often referred to as HER2 (for human epidermal growth factor receptor 2). When reference is made herein to ErbB-2, this reference refers to human ErbB-2. Antibodies comprising an antigen-binding site that binds to ErbB-2 bind to human ErbB-2. Due to sequence and tertiary structure similarities between human and other mammalian orthologs, ErbB-2 antigen-binding sites may, but do not necessarily, bind to such orthologs. The database accession numbers for the human ErbB-2 protein and the gene encoding it are (NP_001005862.1, NP_004439.2, NC_000017.10, NT_010783.15, NC_018928.2). These accession numbers are provided primarily to provide further methods of identification of ErbB-2 as a target; the actual sequence of the ErbB-2 protein bound by an antibody may vary due to, for example, mutations in the encoding gene, such as those that occur in some cancers. The ErbB-2 antigen-binding site binds to ErbB-2 and its various variants, such as those expressed by some ErbB-2-positive cells or tumor cells.

[0094] The term "ErbB-3," as used herein, refers to the protein encoded by the ERBB-3 gene in humans. Alternative names for this gene or protein are HER3; LCCS2; MDA-BF-1; c-ErbB-3; c-erbb-3; erbb-3-S; p180-Erbb-3; p45-sErbb-3; and p85-sErbb-3. When reference is made herein to ErbB-3, this reference refers to human ErbB-3. Antibodies comprising an antigen-binding site that binds to ErbB-3 bind to human ErbB-3. Due to sequence and tertiary structure similarities between human and other mammalian orthologs, ErbB-3 antigen-binding sites may, but do not necessarily, also bind to such orthologs. The database accession numbers for the human ErbB-3 protein and the gene encoding it are (NP_001005915.1, NP_001973.2, NC_000012.11, NC_018923.2, NT_029419.12). These accession numbers are provided primarily to provide further methods of identification of ErbB-3 as a target; the actual sequence of the ErbB-3 protein bound by the antibody may vary due to, for example, mutations in the encoding gene, such as those that occur in some cancers. The ErbB-3 antigen-binding site binds to ErbB-3 and its various variants, such as those expressed by some ErbB-2-positive cells or tumor cells.

[0095] The term "ErbB-4," as used herein, refers to the protein encoded by the ERBB-4 gene in humans. Alternative names for this gene or protein include HER4, Erb-B2 receptor tyrosine kinase 4, and human epidermal growth factor receptor 4. When reference is made herein to ErbB-1, this reference refers to human ErbB-4.

[0096] The antibodies disclosed herein can reduce ligand-induced receptor function of ErbB-3 on ErbB-2 and ErbB-3-positive cells. In the presence of excess ErbB-2, the ErbB-2 / ErbB-3 heterodimer can provide growth signals to expressing cells in the absence of detectable ligands for the ErbB-3 chains in the heterodimer. This ErbB-3 receptor function is referred to herein as the ligand-independent receptor function of ErbB-3. This ErbB-2 / ErbB-3 heterodimer also provides growth signals to expressing cells in the presence of ErbB-3 ligands. This ErbB-3 receptor function is referred to herein as the ligand-induced receptor function of ErbB-3.

[0097] The term "ErbB-3 ligand," as used herein, refers to a polypeptide that binds to and activates ErbB-3. Examples of ErbB-3 ligands include, but are not limited to, neuregulin 1 (NRG) and neuregulin 2, betacellulin, heparin-binding epidermal growth factor, and epiregulin. The term includes biologically active fragments and / or variants of naturally occurring polypeptides.

[0098] Preferably, the ligand-induced receptor function of ErbB-3 is ErbB-3 ligand-induced growth of ErbB-2 and ErbB-3 positive cells. In a preferred embodiment, the cells are MCF-7 cells (ATCC® HTB-22™), SKBR3 (ATCC® HTB-30™), NCI-87 (ATCC® CRL-5822™), BxPC-3-luc2 cells (Perkin Elmer 125058), BT-474 cells (ATCC® HTB-20™), or JIMT-1 cells (DSMZ number: ACC 589).

[0099] As used herein, ligand-induced receptor function is reduced by at least 20%, preferably at least 30, 40, 50, 60, or at least 70%, and in one particularly preferred embodiment, the ligand-induced receptor function is reduced by 80%, more preferably 90%. This reduction is preferably determined by determining ligand-induced receptor function in the presence of a bispecific antibody disclosed herein and comparing it to the same function in the absence of the antibody under otherwise identical conditions. These conditions include at least the presence of an ErbB-3 ligand. The amount of ligand present is preferably that amount that induces half-maximal growth of ErbB-2 and ErbB-3 positive cell lines. The ErbB-2 and ErbB-3 positive cell lines for this test are preferably the MCF-7 cell line (ATCC® HTB-22™), the SKBR3 cell line (ATCC® HTB-30™), the JIMT-1 cell line (DSMZ ACC 589) or the NCI-87 cell line (ATCC® CRL-5822™). The test for determining ErbB-3 ligand-induced receptor function and / or ligand is preferably a test for ErbB-3 ligand-induced growth reduction, as specified in the Examples.

[0100] The ErbB-2 protein contains several domains (see Figure 1 in Landgraf, R Breast Cancer Res. 2007; 9(1): 202 for reference). These extracellular domains are referred to as domains I to IV. The location of binding to each domain in the antigen-binding sites of the antibodies described herein has been mapped. Bispecific antibodies having an antigen-binding site (first antigen-binding site) that binds to domain I or domain IV of ErbB-2 contain heavy chain variable regions that maintain their distinct binding specificity and affinity for ErbB-2 when combined with various light chains. Bispecific antibodies having an antigen-binding site (first antigen-binding site) binding to domain I or domain IV of ErbB-2 (first antigen-binding site) and an antigen-binding site (second antigen-binding site) for ErbB-3 have been found to be more effective in reducing ligand-induced receptor function of ErbB-3 compared to bispecific antibodies containing an antigen-binding site (first antigen-binding site) binding to another extracellular domain of ErbB-2. Bispecific antibodies containing an antigen-binding site (first antigen-binding site) binding to ErbB-2, where the antigen-binding site binds to domain I or domain IV of ErbB-2, are preferred. Preferably, the antigen-binding site binds to domain IV of ErbB-2. Preferred antibodies contain a first antigen-binding site binding to domain I of ErbB-2 and a second antigen-binding site binding to domain III of ErbB-3.

[0101] In a preferred embodiment, the antibody comprises an antigen-binding site that binds to at least one amino acid in Domain I of ErbB-2 selected from the group consisting of T144, T164, R166, P172, G179, S180, and R181, and a surface-exposed amino acid residue located within about 5 amino acid positions of T144, T164, R166, P172, G179, S180, or R181.

[0102] In a preferred embodiment, the antibody comprises an antigen-binding site that binds to at least one amino acid in domain III of ErbB-3, preferably selected from the group consisting of R426 and surface-exposed amino acid residues located within 11.2 Å of R426 in the native ErbB-3 protein.

[0103] Bispecific antibodies that bind to ErbB-2 and have an antigen-binding site (first antigen-binding site) that further comprises ADCC have been found to be more effective than other ErbB-2-binding antibodies that did not have significant ADCC activity, particularly in vivo. Therefore, bispecific antibodies that exhibit ADCC are preferred. Antibodies whose first antigen-binding site binds to domain IV of ErbB-2 have been found to have intrinsic ADCC activity. Domain I-binding ErbB-2-binding antibodies with low intrinsic ADCC activity can be engineered to enhance ADCC activity. The Fc region mediates antibody function by binding to different receptors on immune effector cells, such as macrophages, natural killer cells, B cells, and neutrophils. Some of these receptors, such as CD16A (FcγRIIIA) and CD32A (FcγRIIA), activate cells to mount responses to antigens. Other receptors, such as CD32B, inhibit immune cell activation. By engineering the Fc region (by introducing amino acid substitutions) to bind to activating receptors with greater selectivity, antibodies with a greater ability to mediate the desired cytotoxic activity by anti-cancer Mabs can be created.

[0104] One technique for enhancing ADCC of antibodies is non-fucosylation (see, e.g., Junttila, T.T., K. Parsons et al. (2010), "Superior In vivo Efficacy of Afucosylated Trastuzumab in the Treatment of HER2-Amplified Breast Cancer," Cancer Research 70(11):4481-4489). Accordingly, further provided are bispecific antibodies disclosed herein that are non-fucosylated. Alternatively, or in addition, several other strategies can be used to achieve ADCC enhancement, including, for example, glycoengineering (Kyowa Hakko / Biowa, GlycArt (Roche), and Eureka Therapeutics) and mutagenesis (Xencor and Macrogenics), all of which aim to improve Fc binding to the low-affinity activating FcγRIIIa and / or reduce binding to the low-affinity inhibitory FcγRIIb.

[0105] Several in vitro methods exist for determining the efficacy of antibodies or effector cells in eliciting ADCC. These include the chromium-51 [Cr51] release assay, europium [Eu] release assay, and sulfur-35 [S35] release assay. Typically, a labeled target cell line expressing a particular surface-exposed antigen is incubated with an antibody specific for that antigen. After washing, effector cells expressing the Fc receptor CD16 are typically co-incubated with the antibody-labeled target cells. Target cell lysis is then typically measured by the release of intracellular label, e.g., by scintillation counting or spectrophotometry.

[0106] In preferred bispecific antibodies, the affinity of the second antigen-binding site for ErbB-3-positive cells is equal to or preferably higher than the affinity of the first antigen-binding site for ErbB-2-positive cells. The affinity (KD) of the second antigen-binding site for ErbB-3-positive cells is preferably 2.0 nM or less, more preferably 1.5 nM or less, more preferably 1.39 nM or less, and more preferably 0.99 nM or less. In a preferred embodiment, the affinity of the second antigen-binding site for ErbB-3 on SK-BR-3 cells is 2.0 nM or less, more preferably 1.5 nM or less, more preferably 1.39 nM or less, and preferably 0.99 nM or less. In one embodiment, the affinity is within the range of 1.39 to 0.59 nM. In a preferred embodiment, the affinity of this second antigen-binding site for ErbB-3 on BT-474 cells is 2.0 nM or less, more preferably 1.5 nM or less, more preferably 1.0 nM or less, more preferably less than 0.5 nM, more preferably 0.31 nM or less, and more preferably 0.23 nM or less. In one embodiment, the affinity is in the range of 0.31 to 0.15 nM. The above-mentioned affinity is preferably measured using steady-state cell affinity assays, in which cells are incubated at 4°C with radioactively labeled antibody, as described in the Examples of WO2015 / 130173, and then cell-bound radioactivity is measured.

[0107] The affinity (KD) of the first antigen-binding site for ErbB-2-positive cells is preferably 5.0 nM or less, more preferably 4.5 nM or less, and more preferably 3.9 nM or less. In a preferred embodiment, the affinity of the first antigen-binding site for ErbB-2 on SK-BR-3 cells is 5.0 nM or less, preferably 4.5 nM or less, more preferably 4.0 nM or less, more preferably 3.5 nM or less, more preferably 3.0 nM or less, and more preferably 2.3 nM or less. In one embodiment, the affinity is within the range of 3.0 to 1.6 nM. In a preferred embodiment, the affinity of the first antigen-binding site for ErbB-2 on BT-474 cells is 5.0 nM or less, preferably 4.5 nM or less, and more preferably 3.9 nM or less. In one embodiment, the affinity is within the range of 4.5 to 3.3 nM. The above-mentioned affinities are preferably measured using steady-state cell affinity assays, in which cells are incubated at 4°C with radioactively labeled antibodies, and then cell-bound radioactivity is measured, as described in the examples of WO2015 / 130173.

[0108] Preferably, the bispecific antibodies used in the disclosed methods do not significantly affect cardiomyocyte survival. Cardiotoxicity is a known risk factor in ErbB-2 targeted therapy, and the frequency of complications increases when trastuzumab is used in conjunction with anthracyclines, thereby inducing cardiac stress.

[0109] The bispecific antibodies disclosed herein are preferably used in humans. Therefore, preferred antibodies are human or humanized. Human tolerance to a polypeptide is governed by many different aspects. T cell-mediated, B cell-mediated, etc., immunity is one of the variables involved in human tolerance to a polypeptide. The constant region of the bispecific antibody is preferably a human constant region. This constant region may contain one or more, preferably no more than 10, preferably no more than 5 amino acid differences, from the constant region of a naturally occurring human antibody. Preferably, the constant portion is derived entirely from a naturally occurring human antibody. The various antibodies produced herein are derived from a human antibody variable domain library. These variable domains are human. The unique CDR regions can be human, synthetic, or derived from another organism. A variable region is considered human if it has an amino acid sequence identical to that of a naturally occurring human antibody, apart from the CDR regions. The variable region of the ErbB-2-binding VH, ErbB-3-binding VH, or light chain of the antibody may contain one or more, preferably no more than 10, preferably no more than 5 amino acid differences from the variable region of a naturally occurring human antibody, not counting potential differences in the amino acid sequence of the CDR regions. Such variations also occur naturally with the aid of somatic hypermutation.

[0110] Antibodies can be derived from various animal species, at least with regard to the heavy chain variable region. For example, it is common practice to humanize mouse heavy chain variable regions, etc. There are various ways in which this can be achieved, including CDR-grafting into a human heavy chain variable region having a 3D structure that matches that of the mouse heavy chain variable region; deimmunization of the mouse heavy chain variable region, which is preferably performed by removing known or suspected T-cell or B-cell epitopes from the mouse heavy chain variable region. Removal is typically by substituting one or more amino acids in the epitope with another (typically conservative) amino acid, resulting in a modified sequence of the epitope such that it is no longer a T-cell or B-cell epitope.

[0111] Such deimmunized mouse heavy chain variable regions are less immunogenic in humans than the original mouse heavy chain variable region. Preferably, the variable region or domain is further humanized, e.g., veneered. By using veneering techniques, exterior residues that are easily encountered by the immune system are selectively replaced with human residues to provide hybrid molecules comprising either a weakly immunogenic veneered surface or a substantially non-immunogenic veneered surface. The animal used in the present invention is preferably a mammal, more preferably a primate, and most preferably a human.

[0112] The bispecific antibodies disclosed herein preferably comprise a human antibody constant region. According to differences in their heavy chain constant domains, antibodies are grouped into five classes or isotypes: IgG, IgA, IgM, IgD, and IgE. These classes or isotypes comprise at least one of the above heavy chains, designated by the corresponding Greek letter. Preferably, the constant region comprises an IgG constant region, more preferably an IgG1 constant region, preferably a mutated IgG1 constant region. Some variations in the IgG1 constant region, such as the allotypes G1m1, 17, and G1m3, are naturally occurring and / or tolerated without altering the immunological properties of the resulting antibody. Typically, insertions, deletions, substitutions, or combinations thereof of approximately 1 to 10 amino acids are tolerated in the constant region.

[0113] Preferred bispecific antibodies disclosed herein include: - at least the CDR3 sequence, preferably at least the CDR1, CDR2 and CDR3 sequences, or at least the heavy chain variable region sequence of an ErbB-2 specific heavy chain variable region selected from the group consisting of: MF2926, MF2930, MF1849; MF2973, MF3004, MF3958, MF2971, MF3025, MF2916, MF3991, MF3031, MF2889, MF2913, MF1847, MF3001, MF3003 and MF1898, or a heavy chain variable region sequence which differs from the listed heavy chain variable region sequence by at most 15 amino acids, preferably by at most 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids, more preferably by at most 1, 2, 3, 4 or 5 amino acids; and / or - from the group consisting of MF3178;MF3176;MF3163;MF3099;MF3307;MF6055;MF6056;MF6057;MF6058;MF6059;MF6060;MF6061;MF6062;MF6063;MF6064;MF6065;MF6066;MF6067;MF6068;MF6069;MF6070;MF6071;MF6072;MF6073 and MF6074 At least the CDR3 sequence of a selected ErbB-3-specific heavy chain variable region, preferably at least the CDR1, CDR2 and CDR3 sequences, or at least the heavy chain variable region sequence, or a heavy chain variable region sequence that differs from the listed heavy chain variable region sequences by at most 15 amino acids, preferably by at most 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids, more preferably by at most 1, 2, 3, 4 or 5 amino acids.

[0114] The CDR sequences are preferably altered to improve the binding potency or stability of the antibody, for example, for optimization purposes. Optimization is performed, for example, by mutagenesis procedures, after which the stability and / or binding affinity of the resulting antibody is preferably tested and improved ErbB-2 or ErbB-3-specific CDR sequences are preferably selected. Those skilled in the art are well able to create antibody variants containing at least one altered CDR sequence. For example, conservative amino acid substitutions are applied. Examples of conservative amino acid substitutions include the substitution of one hydrophobic residue, such as isoleucine, valine, leucine, or methionine, for another hydrophobic residue, and the substitution of one polar residue for another, for example, arginine for lysine, glutamic acid for aspartic acid, or glutamine for asparagine.

[0115] Preferred antibodies comprise a variable domain that binds to ErbB-2, wherein the VH chain of this variable domain comprises the amino acid sequence of VH chain MF2926; MF2930; MF1849; MF2973; MF3004; MF3958 (which is humanized MF2971); MF2971; MF3025; MF2916; MF3991 (which is humanized MF3004); MF3031; MF2889; MF2913; MF1847; MF3001, MF3003 or MF1898; or up to 15, preferably 1, 2 or 3, of the above VH chain sequences. MF3031; MF2889; MF2913; MF1847; MF3001, MF3003 or MF1898, with insertions, deletions, substitutions or combinations thereof of up to 1, 3, 4, 5, 6, 7, 8, 9 or 10, more preferably up to 1, 2, 3, 4 or 5 amino acids. The VH chain of the variable domain that binds ErbB-2 preferably comprises the amino acid sequence: - MF1849; or - MF2971 or a humanized version thereof, wherein this humanized version preferably comprises the amino acid sequence of MF3958; or MF3004 or a humanized version thereof, wherein said humanized version preferably comprises the amino acid sequence of MF3991. In one embodiment, the VH chain of the variable domain that binds to ErbB-2 comprises the following amino acid sequence: VH chain MF1849; or MF2971 or a humanized version thereof, wherein the humanized version preferably comprises the amino acid sequence of MF3958; or MF3004 or a humanized version thereof, wherein the humanized version preferably comprises the amino acid sequence of MF3991, wherein the recited VH sequences have, with respect to each sequence, up to 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10, more preferably up to 1, 2, 3, 4 or 5 amino acid insertions, deletions, substitutions or combinations thereof. In a preferred embodiment, the VH chain of the variable domain that binds ErbB-2 comprises the amino acid sequence of MF3958; or the amino acid sequence of MF3958 with up to 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10, more preferably up to 1, 2, 3, 4 or 5 amino acid insertions, deletions, substitutions or a combination thereof, relative to this VH chain sequence.

[0116] The VH chain of the variable domain that binds to Erb-B3 preferably comprises the amino acid sequence of VH chain MF3178; MF3176; MF3163; MF3099; MF3307; MF6055; MF6056; MF6057; MF6058; MF6059; MF6060; MF6061; MF6062; MF6063; MF6064; MF6065; MF6066; MF6067; MF6068; MF6069; MF6070; MF6071; MF6072; MF6073 or MF6074; or up to 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 7 MF6065; MF6066; MF6067; MF6068; MF6069; MF6070; MF6071; MF6072; MF6073 or MF6074, having an insertion, deletion, substitution or combination thereof of at most 1, 7, 8, 9 or 10, more preferably at most 1, 2, 3, 4 or 5 amino acids. The VH chain of the variable domain that binds to Erb-B3 preferably comprises the amino acid sequence of MF3178, MF3176, MF3163, MF6058, MF6061 or MF6065; or the amino acid sequence of MF3178, MF3176, MF3163, MF6058, MF6061 or MF6065 with up to 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10, more preferably up to 1, 2, 3, 4 or 5 amino acid insertions, deletions, substitutions or combinations thereof with respect to each VH chain sequence. In a preferred embodiment, the VH chain of the variable domain that binds to ErbB-3 comprises the amino acid sequence of MF3178; or the VH chain sequence has up to 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10, more preferably up to 1, 2, 3, 4, or 5 amino acid insertions, deletions, substitutions, or a combination thereof. Preferably, the above-mentioned amino acid insertions, deletions, and substitutions are not present in the CDR3 region.The above-mentioned amino acid insertions, deletions and substitutions are preferably not present in the CDR1 and CDR2 regions, and the above-mentioned amino acid insertions, deletions and substitutions are preferably not present in the FR4 region.

[0117] Preferably, the antibody comprises at least the CDR1, CDR2, and CDR3 sequences of MF1849, MF2971, MF3958, MF3004, or MF3991, most preferably at least the CDR1, CDR2, and CDR3 sequences of MF3958. Preferably, the antibody comprises at least the CDR1, CDR2, and CDR3 sequences of MF3178, MF3176, MF3163, MF6058, MF6061, or MF6065, most preferably at least the CDR1, CDR2, and CDR3 sequences of MF3178.

[0118] Preferably, the ErbB-2-specific heavy chain variable region comprises the amino acid sequence of VH chain MF3958 with insertions, deletions, substitutions or combinations thereof of up to 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10, more preferably up to 1, 2, 3, 4, or 5 amino acids for this VH (preferably, the insertions, deletions, or substitutions are not present in CDR1, CDR2, or CDR3). Preferably, the ErbB-3-specific heavy chain variable region comprises the amino acid sequence of VH chain MF3178 with insertions, deletions, substitutions or combinations thereof of up to 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10, more preferably up to 1, 2, 3, 4, or 5 amino acids for this VH. The insertions, deletions, substitutions or combinations thereof of one or more amino acids are preferably not present in the CDR1, CDR2, or CDR3 regions of the VH chain. They are also preferably not present in the FR4 region.The amino acid substitutions are preferably conservative amino acid substitutions.

[0119] Preferably, the ErbB-2-specific heavy chain variable region comprises the amino acid sequence of VH chain MF3991 with a maximum of 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10, more preferably 1, 2, 3, 4, or 5, amino acid insertions, deletions, substitutions, or a combination thereof for the VH (preferably, the insertions, deletions, or substitutions are not present in CDR1, CDR2, or CDR3). Preferably, the ErbB-3-specific heavy chain variable region comprises the amino acid sequence of VH chain MF3178 with a maximum of 15, preferably 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10, more preferably 1, 2, 3, 4, or 5, amino acid insertions, deletions, substitutions, or a combination thereof for the VH. The one or more amino acid insertions, deletions, substitutions, or a combination thereof are preferably not present in the CDR1, CDR2, or CDR3 regions of the VH chain. They are also preferably not present in the FR4 region. The amino acid substitutions are preferably conservative amino acid substitutions.

[0120] Preferably, the first antigen-binding site of the antibody comprises at least the CDR1, CDR2 and CDR3 sequences of MF3958, or CDR1, CDR2 and CDR3 sequences that differ from the CDR1, CDR2 and CDR3 sequences of MF3958 by at most three, preferably at most two, and preferably at most one amino acid, and the second antigen-binding site comprises at least the CDR1, CDR2 and CDR3 sequences of MF3178, or CDR1, CDR2 and CDR3 sequences that differ from the CDR1, CDR2 and CDR3 sequences of MF3178 by at most three, preferably at most two, and preferably at most one amino acid.

[0121] Preferably, the bispecific antibody comprises i) a first antigen-binding site comprising an ErbB-2-specific heavy and light chain variable region comprising the CDR1, CDR2 and CDR3 sequences of MF3958, and ii) a second antigen-binding site comprising an ErbB-3-specific heavy and light chain variable region comprising the CDR1, CDR2 and CDR3 sequences of MF3178.

[0122] Preferably, the ErbB-2-specific heavy chain variable region has the MF3958 sequence, and the ErbB-3-specific heavy chain variable region has the MF3178 sequence. This combination is also referred to as the PB4188 antibody. Preferably, the PB4188 antibody is non-fucosylated.

[0123] Preferably, the bispecific antibody comprises a "heavy chain for erbB-2 binding" as shown in Sequence Listing Part 1D and a "heavy chain for erbB-3 binding" as shown in Sequence Listing Part 1D.

[0124] Preferably, the antigen-binding site of the bispecific antibody comprises the germline light chain O12, preferably the rearranged germline human kappa light chain IgVκ1-39*01 / IGJκ1*01 or a fragment or functional derivative thereof (nomenclature according to the IMGT database World Wide Web at imgt.org). * 01 / IGJκ1 * In the case of IGKV1-39 / IGKJ1, a huVκ1-39 light chain, or huVκ1-39, is used. This light chain may have 1, 2, 3, 4, or 5 amino acid insertions, deletions, substitutions, or a combination thereof. The 1, 2, 3, 4, or 5 amino acid substitutions are preferably conservative amino acid substitutions, and the insertions, deletions, substitutions, or a combination thereof are preferably not present in the CDR3 region of the VL chain, and preferably not present in the CDR1, CDR2, or CDR3 region or the FR4 region of the VL chain. Preferably, the first antigen-binding site and the second antigen-binding site comprise the same light chain variable region, or more precisely, a common light chain. Preferably, the light chain variable region comprises a CDR1 having the sequence (RASQSISSYLN), a CDR2 having the sequence (AASSLQS), and a CDR3 having the sequence (QQSYSTPPT). Preferably, the light chain variable region comprises the consensus light chain sequence shown in Sequence Listing Part 1C.

[0125] Various methods are available for producing bispecific antibodies and are disclosed in WO2015 / 130173. One method involves expressing two different heavy chains and two different light chains in cells and collecting the antibodies produced by the cells. The antibodies produced in this manner typically include a population of antibodies with different combinations of heavy and light chains, some of which are the desired bispecific antibodies. The bispecific antibodies can then be purified from the population of antibodies.

[0126] The ratio of bispecific antibodies to other antibodies produced by a cell can be increased in various ways. Preferably, this ratio is increased by expressing two essentially identical light chains in the cell, rather than expressing two different light chains. This concept is also referred to in the art as the "common light chain" approach. When an essentially identical light chain functions with two different heavy chains to form variable domains with different antigen-binding sites and associated different binding properties, the ratio of bispecific antibodies to other antibodies produced by the cell is significantly improved over the expression of two different light chains. The ratio of bispecific antibodies produced by the cell can be further improved by encouraging the pairing of two different heavy chains with each other rather than the pairing of two identical heavy chains. The art describes various methods by which such heterodimerization of heavy chains can be achieved. One method is to create "knob-into-hole" bispecific antibodies. See U.S. Patent Application Publication No. 20030078385 (Arathoon et al. - Genentech). Another and preferred method is described in PCT application PCT / NL2013 / 050294 (WO2013 / 157954A1), which is incorporated herein by reference. Methods and means for producing bispecific antibodies from a single cell are disclosed, thereby providing a means by which the formation of bispecific antibodies is prioritized over the formation of monospecific antibodies.

[0127] The sequences referred to in this disclosure are shown below and in FIG.

[0128] Sequence 1A (erbB-2 specific) MF2926: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0129] [ka]

[0130] Amino acid sequence: QVQLQQSGPELVKPGASVMISCKASGYSFTGYHMNWVKQSPEKSLEWIGDINPSIGTTAHNQIFRAKATMTVDKSSNTAYMQLKSLTSEDSGVFYCVRRGDWSFDVWGTGTTVTVSS CDR1: GYHMNWVKQSPEKSLE CDR2: NQIFRA CDR3: RGDWSFDV MF2930: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0131] [ka]

[0132] Amino acid sequence: EVQLQQSGAELVKPGASVMMSCKVSGYTFTSYPIAWMKQVHGKSLEWIGNFHPYSDDTKYNENFKGKATLTVEKSSSTVYLELSRLTSDDSAVYYCARSNPLYYFAMDYWGQGTSVTVSS CDR1: SYPIAWMKQVHGKSLE CDR2: NENFKG CDR3: SNPLYYFAMDY MF1849: heavy chain variable region sequence of erbB-2 binding antibody Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0133] [ka]

[0134] Amino acid sequence: QVQLVESGGGVVQPGRSLRLSCAASGFTFSSYGMHWVRQAPGKGLEWVAVISYDGSNKYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAKGDYGSYSSYAFDYWGQGTLVTVSS CDR1: SYGMH CDR2: VISYDGSNKYYADSVKG CDR3: GDYGSYSSYAFDY MF2973: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0135] [ka]

[0136] Amino acid sequence: QVQLKQSGAELVRPGASVKLSCKASGYIFTGYYINWLRQRPGQGLEWIAKIYPGSGNTYYNEKFRGKATLTAEESSSTAYMQLSSLTSEDSAVYFCARGPHYDYDGPWFVYWGQGTLVTVSS CDR1: GYYINWLRQRPGQGLE CDR2: NEKFRG CDR3: GPHYDYDGPWFVY MF3004: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0137] [ka]

[0138] Amino acid sequence: QVQLKQSGAELVRPGASVKLSCKASGYTFTGYYINWVKQRPGQGLEWIARIYPGSGYTYYNEKFKGKATLTAEESSSTAYMHLSSLTSEDSAVYFCARPHYGYDDWYFGVWGTGTTVTVSS CDR1: GYYINWVKQRPGQGLE CDR2: NEKFKG CDR3: PHYGYDDWYFGV MF2971: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0139] [ka]

[0140] Amino acid sequence: QVQLKQSGAELVRPGASVKLSCKASGYTFTAYYINWVKQRPGQGLEWIARIYPGSGYTYYNEIFKGRATLTADESSSTAYMQLSSLTSEDSAVYFCARPPVYYDSAWFAYWGQGTLVTVSS CDR1: AYYINWVKQRPGQGLE CDR2: NEIFKG CDR3: PPVYYDSAWFAY MF3025: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0141] [ka]

[0142] Amino acid sequence: QVQLKQSGAELVRPGTSVKLSCKASGYTFTGYYINWVKQRPGQGLEWIARIYPGSGYTYYNEKFKGKATLTAEESSNTAYMHLSSLTSEDSAVYFCARPHYGYDDWYFAVWGTGTTVTVSS CDR1: GYYINWVKQRPGQGLE CDR2: NEKFKG CDR3: PHYGYDDWYFAV MF2916: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0143] [ka]

[0144] Amino acid sequence: QVQLQQSGAELVRPGASVKLSCKASGYTFTGYYINWVKQRPGQGLEWIARIYPGSGHTSYNEKFKGKATLTTEKSSSTAYMQLSSLTSEDSAVYFCARPIYFDYAGGYFDVWGTRTSVTVSS CDR1: GYYINWVKQRPGQGLE CDR2: NEKFKG CDR3: PIYFDYAGGYFDV MF3958: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0145] [ka]

[0146] Amino acid sequence: QVQLVQSGAEVKKPGASVKLSCKASGYTFTAYYINWVRQAPGQGLEWIGRIYPGSGYTSYAQKFQGRATLTADESTSTAYMELSSLRSEDTAVYFCARPPVYYDSAWFAYWGQGTLVTVSS CDR1: AYYIN CDR2: RIYPGSGYTSYAQKFQG CDR3: PPVYYDSAWFAY MF3031: heavy chain variable region sequence of erbB-2 binding antibody Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0147] [ka]

[0148] Amino acid sequence: QVQLQQSGAELVRPGASVKLSCKASGYTFTAYYINWVKQRPGQGLEWIAKIYPGSGYTYYNENFRGKATLTAEESSSTAYIQLSSLTSEDSAVYFCARGVYDYDGAWFAYWGQGTLVTVSS CDR1: AYYINWVKQRPGQGLE CDR2: NENFRG CDR3: GVYDYDGAWFAY MF3991: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0149] [ka]

[0150] Amino acid sequence: QVQLVQSGAEVKKPGASVKLSCKASGYTFTAYYINWVRQAPGQGLEWIGRIYPGSGYTSYAQKFQGRATLTADESTSTAYMELSSLRSEDTAVYFCARPHYGYDDWYFGVWGQGTLVTVSS CDR1: AYYIN CDR2: RIYPGSGYTSYAQKFQG CDR3: PHYGYDDWYFGV

[0151] Sequence 1B (erbB-3 specific) MF3178: erbB-3 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0152] [ka]

[0153] Amino acid sequence: QVQLVQSGAEVKKPGASVKVSCKASGYTFTGYYMHWVRQAPGQGLEWMGWINPNSGGTNYAQKFQGRVTMTRDTSISTAYMELSRLRSDDTAVYYCARDHGSRHFWSYWGFDYWGQGTLVTVSS CDR1: GYYMH CDR2: WINPNSGGTNYAQKFQG CDR3: DHGSRHFWSYWGFDY MF3176: erbB-3 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0154] [ka]

[0155] Amino acid sequence: EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSAISGSGGSTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARDWWYPPYYWGFDYWGQGTLVTVSS CDR1: SYAMS CDR2: AISGSGGSTYYADSVKG CDR3: DWWYPPYYWGFDY MF3163: erbB-3 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0156] [ka]

[0157] Amino acid sequence: QVQLVQSGAEVKKPGASVKVSCKASGYTFTGYYMHWVRQAPGQGLEWMGWINPNSGGTNYAQKFQGRVTMTRDTSISTAYMELSRLRSDDTAVYYCAKDSYSRHFYSWWAFDYWGQGTLVTVSS CDR1: GYYMH CDR2: WINPNSGGTNYAQKFQG CDR3: DSYSRHFYSWWAFDY MF3099: erbB-3 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0158] [ka]

[0159] Amino acid sequence: EVQLQQPGAELVRPGTSVKLSCKASGYTFTSYWMHWVKQRPGQGLEWIGILDPSDSYTTYNQKFKGKATLTVDTSSSIAYMQLSSLTSEDSALYYCARGGDYDEGGAMDYWGQGTSVTVSS CDR1: SYWMH CDR2: ILDPSDSYTTYNQKFKG CDR3: GGDYDEGGAMDY MF3307: erbB-3 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0160] [ka]

[0161] Amino acid sequence: QVQLVQSGAEVKKPGASVKVSCKASGYTFTGYYMHWVRQAPGQGLEWMGWINPNSGGTNYAQKFQGRVTMTRDTSISTAYMELSRLRSDDTAVYYCARGSRKRLSNYFNAFDYWGQGTLVTVSS CDR1: GYYMH CDR2: WINPNSGGTNYAQKFQG CDR3: GSRKRLSNYFNAFDY

[0162] Array 1C Common light chain Variable region of IGKV1-39A DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSRFSGSGSGTDFTLTISSLQPEDFATYYCQQSYSTP CDR 1: RASQSISSYLN CDR 2: AASSLQS CDR 3: QQSYSTPPT IGKV1-39 / jk1 DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSRFSGSGSGTDFTLTISSLQPEDFATYYCQQSYSTPPTFGQGTKVEIK Common light chain IGKV1-39 / jk1 (constant region underlined) DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSRFSGSGSGTDFTLTISSLQPEDFATYYCQQSYSTPPTFGQGTKVEIK RTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC IGKV1-39 / jk5 common light chain variable domain DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSRFSGSGSGTDFTLTISSLQPEDFATYYCQQSYSTPPITFGQGTRLEIK

[0163] Sequence 1D (erbB-2 specific) Heavy chain for erbB-2 binding QVQLVQSGAEVKKPGASVKLSCKASGYTFTAYYINWVRQAPGQGLEWIGRIYPGSGYTSYAQKFQGRATLTADESTSTAYMELSSLRSEDTAVYFCARPPVYYDSAWFAYWG QGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCD KTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALAPIE KTISKAKGQPREPQVYTDPPSREEMTKNQVSLTCEVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG Heavy chain for erbB-3 binding QVQLVQSGAEVKKPGASVKVSCKASGYTFTGYYMHWVRQAPGQGLEWMGWINPNSGGTNYAQKFQGRVTMTRDTSISTAYMELSRLRSDDTAVYYCARDHGSRHFWSYWGFDY WGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKS CDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPI EKTISKAKGQPREPQVYTKPPSREEMTKNQVSLKCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG

[0164] Array 1E HER2-specific Ab sequence MF2889: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0165] [ka]

[0166] Amino acid sequence: EVQLQQSGAELVRPGTSVKVSCKASGYAFTNYLIEWVKQRPGQGLEWIGVIYPEGGGTIYNEKFKGKATLTADKSSSTAYMQLSGLTSEDSAVYFCARGDYDYKYAMDYWGQGTSVTVSS CDR1: NYLIE CDR2: VIYPEGGGTIYNEKFKG CDR3: GDYDYKYAMDY MF2913: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0167] [ka]

[0168] Amino acid sequence: EVKLQQSGPELVKPGASVKISCKASGYSFTDYKMDWVKQSHGKSLEWIGNINNPNSGGVIYNQKFRGKVTLTVDRSSSAAYMELRSLTSEDTAVYYCSRGLWDAMDSWGQGTSVTVSS CDR1: DYKMDWVKQSHGKSLE CDR2: NQKFRG CDR3: GLWDAMDS MF1847: heavy chain variable region sequence of erbB-2 binding antibody Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0169] [ka]

[0170] Amino acid sequence: QVQLVESGGGVVQPGRSLRLSCAASGFTFSSYGMHWVRQAPGKGLEWVAVISYDGSNKYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAKGWWHPLLSGFDYWGQGTLVTVSS CDR1: SYGMH CDR2: VISYDGSNKYYADSVKG CDR3: GWWHPLLSGFDY MF3001: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0171] [ka]

[0172] Amino acid sequence: EVQLQQSGAELAKPGASVKLSCKTSGYNFPIYWMHWVKQRPGRGLEWIGYINPSTGYIKNNQKFKDKATLTADKSSNTAYMQLNSLTYEDSAVYYCTREGITGFTYWGQGTLVTVSS CDR1: IYWMHWVKQRPGRGLE CDR2: NQKFKD CDR3: EGITGFTY MF1898: heavy chain variable region sequence of erbB-2 binding antibody Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0173] [ka]

[0174] Amino acid sequence: QVQLVESGGGVVQPGRSLRLSCAASGFTFSSYGMHWVRQAPGKGLEWVAVISYDGSNKYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCAKDGFRRTTLSGFDYWGQGTLVTVSS CDR1: SYGMH CDR2: VISYDGSNKYYADSVKG CDR3: DGFRRTTLSGFDY MF3003: erbB-2 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0175] [ka]

[0176] Amino acid sequence: QVQLKQSGPELVKPGASVKISCKASGDAFSYSWMNWVKQRPGKGLEWIGRIYPGDGDINYNGKFKGKATLTADKSSSTAHLQLNSLTSEDSAVYFCARGQLGLEAWFAYWGQGTLVTVSS CDR1: YSWMNWVKQRPGKGLE CDR2: NGKFKG CDR3: GQLGLEAWFAY HER3-specific Ab sequence MF6058: erbB-3 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0177] [ka]

[0178] Amino acid sequence: QVQLVQSGADVKKPGASVKVTCKASGYTFTGYYMHWVRQAPGQALEWMGWINPQSGGTNYAKKFQGRVSMTRETSTSTAYMQLSRLRSDDTATYYCARDHGSRHFWSYWGFDYWGQGTLVTVSS CDR1: GYYMH CDR2: WINPQSGGTNYAKKFQG CDR3: DHGSRHFWSYWGFDY MF6061: erbB-3 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0179] [ka]

[0180] Amino acid sequence: QVQLVQSGAEVKKPGASVKVSCKASGYTFTGYYMHWVRQAPGQGLEWMGWINPQSGGTNYAQKFKGRVTMTRDTSTSTAYMELSRLRSDDTAVYYCARDHGSRHFWSYWGFDYWGQGTLVTVSS CDR1: GYYMH CDR2: WINPQSGGTNYAQKFKG CDR3: DHGSRHFWSYWGFDY MF6065: erbB-3 binding antibody heavy chain variable region sequence Nucleic acid sequence (underlined sequence encodes the end of the leader peptide):

[0181] [ka]

[0182] Amino acid sequence: QVQLVQSGAEVKKPGASVKVSCKASGYTFTSYYMHWVRQAPGQGLEWMGWINPQGGSTNYAQKFQGRVTMTRDTSTSTVYMELSRLRSEDTAVYYCARDHGSRHFWSYWGFDYWGQGTLVTVSS CDR1: SYYMH CDR2: WINPQGGSTNYAQKFQG CDR3: DHGSRHFWSYWGFDY

[0183] For purposes of clarity and concise description, features are described herein as part of the same or separate embodiments, but it will be understood that the scope of the invention may include embodiments having all or any combination of the described features. [Brief explanation of the drawings]

[0184] [Figure 1] Figure 1 shows an amino acid alignment of MF3178 variants. Dots indicate the same amino acid as MF3178 at that position. The CDR1, CDR2, and CDR3 sequences of MF3178 are in bold and underlined. [Figure 2A] 1 shows increased in vivo tumor targeting of bispecific antibodies over monoclonal antibodies. Micro-PET imaging demonstrates that the PB4188 variant accumulates more efficiently in tumors compared to the HER3 monoclonal. [Figure 2B] Figure 1 shows increased in vivo tumor targeting of bispecific antibodies over monoclonal antibodies. Gamma-counter quantification of radioactivity present in tumors confirmed that levels of the PB4188 variant in tumors were 2.5-fold higher than those of the parent anti-HER3 antibody. [Figure 2C] Figure 1 shows increased in vivo tumor targeting of bispecific antibodies over monoclonal antibodies. Quantitative biodistribution of tumor uptake in the four mAb groups at 48 hours. Results are presented as percentage of injected dose per gram of tissue (%ID / g), with error bars indicating ±SD. [Figure 3A] Figure 1 shows dose-response curves for antibody antagonists in EGFR:HER2, HER2:HER3, and HER2:HER4 assays. Reporter cells were seeded at 2.5K / well for EGFR:HER2 or 5K / well for HER2:HER3 and HER2:HER4 at 37°C for 4 hours. Antibodies were serially diluted and incubated at 37°C for 3 hours, followed by stimulation with 10 ng / ml EGF or 30 ng / ml HRG-β2 for EGFR:HER2 or HER2:HER3 and HER2:HER4, respectively, for 16 hours. Agonist reference stimulation curves were obtained by incubating titrations of ligand alone for 24 hours. Each data point represents the mean and standard deviation of four replicates for one dose. Data were plotted in GraphPad Prism, and curve fitting was performed using a log(inhibitor) vs. response variable (four-parameter) fit to calculate IC50s. [Figure 3B] This is a continuation of Figure 3A. [Figure 4]Figure 1 shows the weight changes of mice in different groups. Body weight changes after administration of MCLA-128, PG2863, and PG2869 antibodies to female BALB / c nude mice bearing established OV-10-0050 tumors. Data points represent group mean body weights. Error bars represent standard error of the mean (SEM). [Figure 5] Figure 1 shows the relative change (%) in body weight. BW change was calculated based on the animal's weight on the first day of dosing. Data points represent the percent group mean change in BW. Error bars represent the standard error of the mean (SEM). [Figure 6] Figure 1 shows tumor growth curves. Tumor volume traces after administration of MCLA-128, PG2863, and PG2869 antibodies to female BALB / c nude mice bearing established OV-10-0050 tumors. Data points represent group means, and error bars represent standard error of the mean (SEM). [Figure 7] Figure 1 shows growth inhibition of tumor lines MDA-MB-175 and OV-10-0050 in vitro and in vivo. DOC4-NRG1 and CLU-NRG1 gene fusions are expressed in MDA-MB-175 cell line (breast) and OV-10-0050 PDX (ovarian), respectively. Left panel: In vitro MCLA-128 treatment inhibits MDA-MB-175 cell proliferation. Right panel: In vivo MCLA-128 treatment (25 mg / kg weekly until day 28) reduced tumor growth and eliminated tumors in 6 / 8 animals. [Example]

[0185] Example 1 ErbB-2-guided targeting Imaging experiments were conducted to compare the HER2 x HER3 bispecific antibody (PB4188) with a HER3 bivalent monoclonal antibody. Variants of the bAb PB4188 and the anti-HER3 MF3178 (parent antibody) were labeled with 64Cu and intravenously injected into mice xenografted with HER2 gene-amplified JIMT-1 tumors. Micro-PET imaging demonstrated that the PB4188 variant accumulated more efficiently in tumors compared with the HER3 monoclonal antibody (Figure 2A). Gamma-counter quantification of radioactivity present in the tumor confirmed that the levels of the PB4188 variant in tumors were 2.5-fold higher than those of the parent anti-HER3 antibody (Figure 2B). Overall, the in vitro and in vivo data demonstrate that HER2 targeting is responsible for the enhanced binding of PB4188 to tumor cells. Further studies were performed using the anti-HER2 (MF3958) antibody. FIG. 2C summarizes the results for each antibody injected into mice xenografted with 64Cu-labeled HER2 gene-amplified JIMT-1 tumors (n=4 mice for each antibody treatment).

[0186] method Biodistribution study. Variants of bAb PB4188, anti-HER2 MF3958 and anti-HER3 MF3178, were conjugated to a bifunctional chelator [Paterson 2014 Dalton Transactions]. The target-binding characteristics of the conjugated products were confirmed using a flow cytometry-based assay. The proteins were then labeled with 64Cu, and mice bearing JIMT-1 breast xenografts were administered the radiolabeled antibodies via the tail vein ("iv" for Figures 2A-2B and 2C) or intraperitoneally ("ip" for Figure 2C). MicroPET / CT images were acquired 48 hours after injection, after which the tumors were excised and radioactivity was measured in a gamma counter. Results were expressed as a percentage of the injected dose per gram of tissue.

[0187] Example 2 Inhibition of heterodimer formation We used a heterodimerization assay based on enzyme fragment complementation technology. The β-galactosidase enzyme can be artificially split into two inactive fragments, the enzyme donor and the enzyme acceptor, which combine to form the active enzyme only when brought into close proximity. The sequences encoding either the enzyme donor or the enzyme acceptor are linked to the extracellular and transmembrane domains of their respective heterodimerization partners. Both genes are then cotransfected into U2OS cells to express the extracellular domain of an RTK receptor linked to one of the β-galactosidase domains (ED or EA). Upon agonist stimulation of one RTK receptor, both RTK receptors dimerize, inducing the formation of a fully reconstituted, active β-galactosidase enzyme. Finally, β-galactosidase activity is measured by adding a substrate that generates light upon hybridization.

[0188] The antibodies were tested in EGFR:HER2, HER2:HER3, and HER3:HER4 heterodimerization reporter cell lines. RTK heterodimerization assays were performed using the bispecific antibody MCLA-128 (MF3178 arm and MF3958 arm); anti-HER3 antibodies MF3178 / PG3178 and PG3793 / AMG-888 / patritumab; and anti-HER2 antibodies MF3958 / PG3958, PG2867 / trastuzumab, PG2869 / pertuzumab, and Perjeta (a clinical batch of pertuzumab). EGF and HRG titration in the EGFR:HER2, HER2:HER3, and HER2:HER4 assays demonstrated dose-dependent agonist responses (Figure 3). MCLA-128 specifically demonstrated complete inhibition of HER2:HER3 dimer formation but had no effect on EGFR:HER2 or HER2:HER4 heterodimerization. In contrast, trastuzumab (PG2867) behaved as a partial antagonist in both the EGFR:HER2 and HER2:HER3 assays.

[0189] MCLA-128 and PG3178 were the most potent and completely inhibited HRG-induced HER2:HER3 dimerization (Table 1).

[0190] [Table 1]

[0191] In the HER2:HER3 assay, trastuzumab was approximately four times less potent than MCLA-128 or PG3178. Perjeta (clinical pertuzumab) behaved as a full antagonist in all three assays, producing a profile similar to that of PG2867 (pertuzumab). In the HER2:HER4 assay, both anti-HER2 PG3958 and PG2867 (pertuzumab) showed a modest decrease in dimerization that appeared dose-dependent. In the EGFR:HER2 assay, small nonspecific responses were observed at high concentrations of PG1337, MCLA-128, PG3178, and PG3958.

[0192] MCLA-128 showed specific inhibition of only the HER2:HER3 heterodimer, indicating that upon binding to HER2, MCLA-128 should neither sterically impair the interaction of HER2 with EGFR upon EGF stimulation nor impair the heterodimerization of HER2 with HER4 upon HRG stimulation.

[0193] The latter is consistent with observations in HRG-induced cell cycle-based proliferation assays of T47D cells. Assays using these cells failed to demonstrate the inhibitory activity of either MCLA-128 or PG3178, likely due to the higher expression of HER4 compared to HER3. HRG appears to signal preferentially through HER2:HER4 instead of HER2:HER3 in T47D cells, explaining the lack of efficacy of MCLA-128 and demonstrating the specificity of MCLA-128 for HRG-induced HER2:HER3 dimers but not for HRG-induced HER2:HER4 dimers.

[0194] In this study, trastuzumab blocked EGF-induced and HRG-induced heterodimerization of EGFR:HER2 and HER2:HER3, respectively. Trastuzumab and pertuzumab behaved as partial and full antagonists, respectively, consistent with the generally accepted assertion that trastuzumab blocks ligand-independent activation of HER2, while pertuzumab blocks ligand-dependent signaling. The fact that trastuzumab inhibitory responses are observed in these assays may be due to overexpression of both targets, which may allow for a more sensitive readout than traditional immunoprecipitation experiments.

[0195] Finally, PG3793 exhibited lower binding affinity to MCF-7 than PG3178, but its lower potency in the HER2:HER3 heterodimerization assay was less dramatic (a 30-fold difference in binding assay affinity versus a 2.5-fold difference in dimerization assay potency). This discrepancy between binding affinity and antagonist potency has been previously observed in the cases of MCLA-128 and PG3178. PG3178 binds to MCF-7 with slightly better affinity than MCLA-128, but MCLA-128 outperforms PG3178 in cell cycle-based proliferation assays.

[0196] Example 3 Testing objectives and regulatory compliance The objective of this study was to evaluate the in vivo anti-tumor efficacy of MCLA-128, PG2863, and PG2869 antibodies in treating the OV-10-0050 subcutaneous human ovarian cancer PDX model in BALB / c nude mice.

[0197] Experimental design The experimental design is shown in Table 2. In all groups, blood was sampled on day 2 (24 hours after the first dose) in four animals and on day 6 (5 days after the first dose) in the remaining four animals. At the indicated time points, 50–100 μl of blood was collected into sterile collection tubes (Microvette CB300Z coagulation activator / serum, Sarstedt BV, catalog number 16.440.100). The samples were allowed to clot for 45 minutes at room temperature, centrifuged at 3000 rpm for 10 minutes, and the aqueous layer (approximately 20 μl of serum) was removed and immediately stored at −80°C in another 1.5 mL sterile Eppendorf tube. Samples were shipped on dry ice.

[0198] [Table 2]

[0199] All animals were treated on days 1, 8, 15, 22, 29 (weekly treatment for 5 weeks) and the route was IP for all groups.

[0200] Tumor samples were collected 48 hours after the last dose (day 31). Tumors were fixed in neutral buffered formalin (tissue:fixative ratio of at least 1:20) for 24 hours and then converted to FFPE blocks.

[0201] Preparation of neutral buffered formalin: Place one bag of PBS powder in a clear 5L volumetric flask, add 4.5L deionized water, and stir to disperse the powder to obtain a clear solution. Then add 500ml formaldehyde and stir until a uniform solution is achieved.

[0202] material Animals: Species: Mouse (Mus musculus); Strain: BALB / c nude; Age: 6-8 weeks; Sex: Female; Weight: 18-22 g; Number of animals: 32 mice + spare

[0203] Animal supplier: Shanghai Sino-British SIPPR / BK Laboratory Animal Co., LTD.

[0204] Diet: Animals had free access to radiation-sterilized dry granular diet for the entire study period; Water: Animals had free access to sterile drinking water.

[0205] Antibody packaging and storage conditions: MCLA-128; cryovials, 2.5 mg / ml, 10 x 1.5 ml / vial, store at 4°C PG2863; cryovials, 10 x 1.5 ml / vial at 2.5 mg / ml, store at 4°C PG2869; cryovials, 2.5mg / ml, 10 x 1.5ml / vial, store at 4°C

[0206] Generation of PDX models The OV-10-0050 human ovarian cancer PDX model was originally established from a 48-year-old female patient with grade 3 adenocarcinoma of the ovary. Surgically resected clinical samples were transplanted into nude mice (defined as passage 0, P0), and subsequent serial transplants were defined as P1, P2, etc. P6 tumor tissue was used in this study.

[0207] Tumor transplantation Each mouse was subcutaneously implanted with scissors-cut OV-10-0050 P6 tumor slices (approximately 30 mm3) into the right flank for tumor development. Treatment began 30 days after tumor implantation, when the average tumor size reached approximately 152 mm3. 32 tumor-bearing mice were randomized into four groups using a stratified randomization method, with each group consisting of eight tumor-bearing mice. The day of randomization was designated day 1, which was the day treatment began. Test articles were administered to mice according to a predetermined regimen, as shown in the experimental design table (Table 2).

[0208] observation All procedures related to the handling, care, and treatment of animals in the study were conducted in accordance with the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) guidelines and guidelines approved by WuXi AppTec's Institutional Animal Care and Use Committee (IACUC). At routine monitoring time points, animals were checked daily for normal behavior, e.g., mobility, food and water consumption (by visual inspection only), weight gain / loss (weight was measured twice weekly), tumor growth and any effects of treatment on eye / hair matting, and any other abnormal effects noted in the protocol. Mortality and observed clinical signs were recorded based on the number of animals in each subset.

[0209] Tumor measurements The tumor growth was delayed or the mice were cured. Tumor size was measured twice a week in two dimensions using calipers, and the volume was expressed in mm using the formula: V = 0.5a × b2, where a and b are the long and short diameters of the tumor, respectively. The tumor size was then used to calculate TC, T / C, and TGI values. TC was calculated using T as the median time (days) required for tumors in the treatment group to reach a given size (e.g., 500 mm3), and C as the median time (days) required for tumors in the control group to reach the same size. The T / C value (percent) is an indicator of antitumor efficacy; T and C were the mean volumes of the treatment and control groups, respectively, on a given day. TGI was calculated for each group using the formula: TGI (%) = [1 - (Ti - T0) / (Vi - V0)] x 100; Ti was the mean tumor volume of the treatment group on a given day, T0 was the mean tumor volume of the treatment group on the first day of treatment, Vi was the mean tumor volume of the vehicle control group on the same day as Ti, and V0 was the mean tumor volume of the vehicle group on the first day of treatment.

[0210] statistical analysis Summary statistics, including the mean and standard error of the mean (SEM), are provided for tumor volumes in each group at each time point (detailed in Table 3). Statistical analysis of differences in tumor volumes between groups and analysis of drug interactions was performed on data obtained at the best therapeutic time point after the final dose (29 days after grouping).

[0211] One-way ANOVA was performed to compare tumor volumes between groups, and if a significant F-statistic (P<0.001, ratio of treatment variance to error variance) was obtained, comparisons between groups were performed using Gaims-Howell. All data were analyzed using SPSS 17.0. p<0.05 was considered statistically significant.

[0212] result Mortality, morbidity and weight gain or loss Animal body weight was monitored periodically as an indirect measure of toxicity. No groups were observed to lose weight as a result of test article administration (Figure 4), and no mortality or morbidity was observed. Therefore, there appears to be no overt toxicity associated with administration of MCLA-128, PG2863, and PG2869 antibodies to tumor-bearing BALB / c nude mice.

[0213] The weight changes in female BALB / c nude mice bearing OV-10-0050 xenografts dosed with MCLA-128, PG2863, and PG2869 antibodies are shown in Figures 4 and 5. The mean tumor volumes over time in female BALB / c nude mice bearing OV-10-0050 xenografts dosed with MCLA-128, PG2863, and PG2869 antibodies are shown in Table 3. Figure 6 shows tumor growth.

[0214] [Table 3]

[0215] Results and Discussion This study evaluated the therapeutic efficacy of MCLA-128, PG2863, and PG2869 antibodies as single agents in treating the OV-10-0050 human ovarian cancer xenograft model. The results of tumor size in different groups at different time points after tumor inoculation are shown in Table 3, Table 4, and Figure 4.

[0216] The mean tumor size in vehicle-treated control mice was 1,161 mm 29 days after group allocation. 3 Treatment with test articles MCLA-128, PG2863, and PG2869 antibodies at 25 mg / kg (QW x 5 weeks) resulted in significant antitumor activity: their mean tumor sizes were 23, 108, and 1 mm3, respectively, at the same time point (T / C values ​​= 1.95%, 9.28%, and 0.06%, respectively; TGI values ​​= 112.78%, 104.37%, and 114.96%; p values ​​= 0.002, 0.003, and 0.002, respectively), and their tumor growth was all greater than 500 mm3. 3 A tumor is scored as follows: a delay of more than 14 days in tumor size compared to the vehicle group; treatment causes partial or complete regression of tumors; tumor volume is reduced by 50% or more of its day 1 volume for three consecutive measurements during the course of the study, or is ≥ 13.5 mm for one or more of those three measurements. 3 Mice were considered to have partial regression (PR) if the mean stenosis was <13.5 mm for three consecutive measurements during the course of the study. 3 Mice were considered to have complete regression (CR) if no palpable tumor was detected at the end of the study. Mice were considered to be tumor-free survivors.

[0217] Treatment with MCLA-128, PG2863, and PG-2869 resulted in different rates of PR, CR, and TFS. The number of mice in each group showing PR, CR, and TFS is shown in Table 5. All test articles were well tolerated by the tumor-bearing animals. No weight loss was observed in any of the treatment groups.

[0218] In summary, all three test antibodies as single agents in this study produced significant antitumor activity against the OV-10-0050 human ovarian cancer xenograft model, which was well tolerated by the tumor-bearing animals. These results indicated that these antibodies are safe and effective anticancer agents.

[0219] [Table 4]

[0220] [Table 5]

[0221] Example 4 MCLA-128 is a bispecific antibody that targets the HER2 and HER3 receptor tyrosine kinases (RTKs), which are involved in cancer cell proliferation and survival. MCLA-128 has been extensively tested for heregulin (HRG)-induced HER3 signaling and proliferation. MCLA-128 has demonstrated greater in vitro potency than the combination of the anti-HER2 antibodies pertuzumab (PG2869) and trastuzumab (PG2867), which can block ligand-dependent and ligand-independent HER2:HER3 signaling, respectively [Agus 2002; Juntilla 2009]; and the anti-HER3 antibody MM-121 (PG2863), which blocks HRG-induced HER3 activation [Schoeberl 2010].

[0222] MCLA-128 also exhibits antitumor activity in cells expressing a gene fusion involving the HRG gene. The MDA-MB-175 cell line contains the DOC4-NRG1 gene fusion, which creates an autocrine growth loop due to NRG1 expression. This gene fusion has not been found in cancer patients to date [Sanchez-Valdivieso 2002].

[0223] From a panel of breast cancer cell lines, MDA-MB-175 cells were sensitive to single-agent MCLA-128, demonstrating the importance of the HER3 / HRG signaling axis in this cell line (Figure 7, left panel). HER2 activation in this cell line was also demonstrated in vivo, with a single dose of pertuzumab, but not trastuzumab, inhibiting orthotopic MDA-MB-175 tumor growth. While the relevance of the DOC4-NRG1 gene fusion in breast cancer patients has been debated [Sanchez-Valdivieso 2002], other gene fusions have recently attracted attention. In particular, the CD74-NRG1 fusion has been reported by independent groups in invasive mucinous adenocarcinoma, a subgroup of non-small cell lung cancer [Fernandez-Cuesta 2014, Duruisseaux 2016]. Several other NRG1 gene fusions have also been detected, namely VAMP2-NRG1, RBPMS-NRG1, and WRN-NRG1 in lung cancer and RAB2IL1-NRG1 in ovarian cancer [Jung 2015; Dhanasekaran, 2014]. This diversity of gene fusions may be related to the location of the NRG1 gene on chromosome 8, which is susceptible to translocations [Adelaide 2003].

[0224] OV-10-0050 was found to be HER-dependent. Treatment with afatinib (an irreversible inhibitor of EGFR and HER2 that also inhibits transphosphorylation of HER3) resulted in tumor growth inhibition. The antitumor efficacy of MCLA-128 was compared to PBS (Figure 7, right panel).

[0225] Mice: NOD-SCID, Crl:NU(NCr)-Foxn1nu and BALB / c nude mice. Antibodies are administered at 25 mg / kg for 4 weeks. Tumor volumes are measured twice a week by caliper.

[0226] Example 5 A Phase I / II Study of MCLA-128, a Full-Length IgG1 Bispecific Antibody Targeting HER2 and HER3, in Patients with Solid Tumors Test duration: Enrollment in the dose-escalation part of the study (Part 1, first patient dosed on February 3, 2015) is complete after recruiting 28 patients. The first patient in Part 2, the dose-escalation phase of the study, was dosed on January 15, 2016 in Europe. The total duration of Part 2 is approximately 25 to 32 months; however, the actual duration is affected by several variables, including overall subject recruitment rates.

[0227] Number of locations: A maximum of 13 sites are estimated to be involved during the study. Additional sites may be added to ensure there is an acceptable enrollment rate or to replace non-enrolling / out-of-drug sites.

[0228] Number of patients: Twenty-eight patients were enrolled in Part 1. For Part 2, at least 20 and up to approximately 40 evaluable patients with invasive mucinous adenocarcinoma or grouped advanced / metastatic non-small cell lung cancer (NSCLC) with a reported NRG1 fusion can be enrolled.

[0229] Patients who do not complete at least two cycles of study treatment due to reasons other than disease progression are not evaluable for efficacy and will be substituted within their respective arms.

[0230] This example describes part 2. Although this example describes the administration of the Erb-2, Erb-3 binding bispecific antibody MCLA-128, this example is not intended to be limited to the use of this particular embodiment and applies to the other bispecific antibodies disclosed herein.

[0231] [Table 6]

[0232] [Table 7]

[0233] Study Design: This is a Phase I / II open-label, multicenter, multinational, dose-escalation, single-arm, assignment study to evaluate the safety, tolerability, PK, PD, immunogenicity, and antitumor activity of MCLA-128.

[0234] The test is designed in two parts: Part 1 Enrollment for Part 1 of the study was achieved on November 24, 2015, and as of January 24, 2017, all patients in Part 1 had completed the study. Nine dose levels were investigated: 40 mg, 80 mg, and 160 mg in one patient cohort, and 240 mg, 360 mg, 480 mg, 600 mg, 750 mg, and 900 mg in three patient cohorts. MCLA-128 was initially given over approximately 60 minutes on Day 1 of a 3-week treatment cycle. During Part 1, the infusion duration was extended to 2 hours, optionally increasing it to 4 hours, to mitigate infusion-related reactions (IRR).

[0235] No dose-limiting toxicities (DLTs) were experienced at any of these dose levels. Three additional patients in each of the 600 mg and 750 mg cohorts were dosed to obtain sufficient PK information.

[0236] Because the MTD was not reached at the 900 mg dose level, the Data Review Committee (DRC) for MCLA-128-CL01 decided to assign the 750 mg dose level as the RP2D for the study based on cumulative safety, available PK data, and PK simulations.

[0237] Part 2 Part 2 will include further characterization of the safety and tolerability of selected dose levels of MCLA-128, as well as an assessment of CBR, defined as the proportion of patients with CR, PR, or sustained SD (SD of at least 12 weeks duration), in an expanded group of selected patient populations.

[0238] A weekly dosing regimen using 4-week cycles consisting of a flat weekly dose of 400 mg for the first two cycles, with an 800 mg loading dose for the initial administration, will be evaluated in newly recruited patients. Beginning with cycle 3, MCLA-128 will be given at a dose of 400 mg weekly for 3 weeks, followed by a 1-week rest period. Mandatory pre-medication will be administered to mitigate IRR. However, corticosteroids are only mandatory prior to the loading dose on day 1 of cycle 1, and should only be used at the investigator's discretion for subsequent infusions to manage IRR.

[0239] The safety of the weekly schedule will be investigated during a run-in period after the first five treated patients have completed at least two treatment cycles. The DRC will review all safety data, focusing on the incidence of grade 3-4 toxicities, the incidence and severity of IRRs, and compliance. If the DRC review concludes that toxicity is unacceptable, the sponsor will continue enrolling patients on the 3-week cycle dosing regimen until a sufficient number of patients per cohort have been enrolled.

[0240] Intra-patient dose escalation will not be permitted in Part 2.

[0241] The target patient populations to be evaluated in Part 2 of the study are: NSCLC with reported NRG1 fusions - open for recruitment in Asia only

[0242] At least 20 and up to approximately 40 patients may be enrolled in each group (C-F) with a minimum of 10 patients per cohort treated weekly at the recommended dose. Previously closed cohorts may be reopened.

[0243] Duration of the procedure Patients in both Part 1 and Part 2 of the study can remain on treatment until disease progression, death, unacceptable toxicity, or discontinuation for any other reason.

[0244] Data Review Committee (DRC): All dose escalation decisions in Part 1 were made by a DRC that convened to review all available safety and PK data. DRC participants included the principal investigators (or their representatives), the sponsor's medical director, the study medical monitor, the study pharmacovigilance physician, the study project manager, the study statistician, and invited experts (e.g., clinical pharmacology experts) as needed.

[0245] In Part 2, the DRC will consider data after completion of the safety run-in period for the weekly dose before escalating the weekly dose regimen in all subsequent patients.

[0246] Test evaluation: The study consists of a screening period of up to 4 weeks (28 days) for molecular pre-screening assessments, followed by sequential treatment cycles until treatment is discontinued or terminated for any reason. Treatment cycle duration is 3 weeks (21 days) for patients treated at the initial recommended dose in Part 2 and 4 weeks (28 days) for patients treated at the weekly recommended dose in Part 2. All patients must attend the end of treatment visit within 1 week of treatment cessation and the final study visit 30 days after the end of treatment or study discontinuation.

[0247] Patients who do not progress or withdraw consent to complete the final study visit will be followed every 3 months for up to 2 years (approximately) to check their disease progression and / or survival status until the start of their next anti-cancer treatment.

[0248] If the ongoing evaluation of safety data and available PK, PD, and antitumor activity data during the study suggests that alternative dosing frequencies should be evaluated or that other patient populations should be evaluated in Part 2, these modifications will be specified in a protocol amendment prior to initiating these evaluations.

[0249] Molecular pre-screening and screening: Molecular pre-screening will be performed in local laboratories qualified to perform molecular screening for NRG1 fusions. To initiate pre-screening, patients must meet one of the following criteria: Histological diagnosis of IMA and reported absence of EGFR / ALK alterations. Note: IMA patients who did not undergo pre-screening testing for NRG1 fusions may enter the study. or Pathological examination does not provide a diagnosis of IMA, but IMA is suspected by the investigator based on symptoms, imaging features (e.g., focal consolidation, multiple bilateral nodules or consolidations), non-smoker status, and reported absence of EGFR / ALK alterations.

[0250] Before submitting fresh or archival tumor tissue for analysis to determine NRG1 fusion status, NSCLC patients identified for potential study participation must sign a molecular prescreening informed consent form (ICF). Testing can be performed at any time during the natural history of the disease (e.g., at diagnosis, during first-line treatment, at progression, etc.) up to one year prior to Day 1 of Cycle 1. Fresh tumor samples (formalin-fixed, paraffin-embedded; FFPE) or archival tumor samples no older than one year are required for assessment of the presence of NRG1 fusions. Samples should be submitted to a local laboratory qualified for testing by molecular profiling (PCR, next-generation sequencing [DNA or RNA], or FISH) of NRG1 fusion status. Patients with a positive local NRG1 fusion result are then eligible to sign the study ICF if they are willing and able to enter the study.

[0251] Primary Informed Consent Form All patients must sign the study ICF before any screening procedures or assessments are performed. Screening assessments are performed within 4 weeks prior to Cycle 1 Day 1, with the exception of the serum pregnancy test, which should be performed within 7 days of Cycle 1 Day 1. A baseline mandatory tumor sample, preferably a block from fresh or archival tissue, is required for screening consideration. The sponsor indicates a preference for fresh tissue. Archived tissue is acceptable and must be obtained within 2 years of screening, except for NSCLC, which must be within 1 year. It should be noted that for NSCLC patients, a baseline biopsy for screening is still required, even if the prescreening biopsy sample is provided to the NRG1 prescreening regional study. Following completion of all required screening assessments and confirmation of all eligibility criteria, patients may begin dosing on Cycle 1 Day 1.

[0252] Safety evaluation Concurrent illnesses will be captured at baseline; AEs and concomitant treatments will be monitored throughout study participation. Safety assessments will include reviewing the Eastern Cooperative Oncology Group (ECOG) activity index, physical examination (including height and weight), vital signs, and electrocardiogram (ECG). Cardiac function testing of left ventricular ejection fraction (LVEF) will also be performed at screening, the end of cycle 4 (or day 1 of cycle 5), the end of study visits, and any time during the study if clinically indicated. Laboratory evaluations will include clinical chemistry, hematology, coagulation studies, urinalysis, and pregnancy testing. Note that cytokine panel analysis was performed through August 1, 2017.

[0253] On all MCLA-128 dosing days, patients must remain in the clinic for at least 60 minutes from the end of the infusion (longer if PK samples are required) for observation and repeat vital signs before being discharged from the clinic. Further safety assessments should be performed as clinically indicated, and if necessary, the duration of clinic admission should be increased based on the investigator's discretion.

[0254] Immunogenicity assessment Serum titers of anti-MCLA-128 antibodies will be measured on pre-medication day 1 for each of cycles 1, 2, 3, 4, and then every four cycles thereafter (cycles 8, 12, 16, etc.), as well as at the end of treatment visit and the final study visit, in a -3 day window before MCLA-128 administration.

[0255] Pharmacokinetic evaluation Initial recommended dosing schedule for Parts 1 and 2: In Cycle 1, blood samples will be collected for PK analysis at predose on Day 1, at the end of infusion (EOI), and 1, 2, 4, 8, 24 hours after EOI, then on Days 4 (or 3), 8, and 15. In Cycles 2-4, only predose and EOI blood samples will be collected.

[0256] Part 2 recommended weekly dosing schedule: In Cycle 1, blood samples will be collected for PK analysis at predose on Day 1, EOI, 2, 4, and 24 hours after EOI, then at predose on Days 8 and 15, and at predose and EOI on Day 22. In Cycles 2 and 3, predose and EOI blood samples will be collected on Day 15. In Cycle 4, blood samples will be collected at predose on Day 1, and at predose and EOI on Day 15. Every two cycles thereafter (Cycles 6, 8, 10, etc.), a predose blood sample will be collected on Day 15.

[0257] Tumor evaluation Tumor assessment will be assessed according to RECIST version 1.1 by the local investigator. Images will be obtained at screening and at the end of each of two cycles of treatment for patients receiving the 3-week cycle regimen and every 6 weeks for patients receiving the 4-week cycle regimen.

[0258] Biomarker and Pharmacokinetic Assessment Various biomarker and pharmacokinetic studies will be performed on archived and / or fresh tumor sample material and / or blood (liquid biopsy), depending on the availability of archived or existing tumor tissue, consent for additional tumor samples, and consent for specific biomarker testing.

[0259] If sufficient sample is available, evaluate the following candidate biomarkers: · HER2, HER3, HER2:HER3 dimerization, phosphorylated HER2 (pHER2) and HER3 (pHER3), and heregulin; Circulating plasma tumor DNA (ctDNA) and tumor sample DNA (depending on availability) will be used to test for mutations in cancer genes, including those associated with HER2 and HER3 signaling. · Phosphorylated molecules in the MAPK and AKT signaling pathways; ·Fc gamma receptor polymorphisms; · Circulating tumor cells for HER2; Heregulin gene fusions

[0260] Germline DNA assessment is not included (except for Fc gamma receptor polymorphisms).

[0261] At baseline, patients will be asked to provide a mandatory tumor sample tissue, preferably a block, which may be derived from fresh or archival tissue. The sponsor has indicated a preference for fresh tissue. Archival tissue is acceptable and must be obtained within two years of screening, except for NSCLC, which must be within one year. Additionally, patients will be asked to optionally submit a tumor sample / biopsy at the end of cycle 4 and, optionally, at the end of treatment visit.

[0262] Blood samples will also be taken at these time points for liquid biopsy testing.

[0263] Eligibility Criteria: The trial will enroll patients with NSCLC.

[0264] General Inclusion Criteria for Part 2 1. 18 years of age or older; 2. At least one measurable lesion according to RECIST v1.1; 3. ECOG activity index of 0 or 1; 4. An estimated life expectancy of at least 12 weeks; 5. Toxicity suffered as a result of prior anticancer therapy (as defined by NCI CTCAE v4.03) that has resolved to ≤ Grade 1, except for alopecia, lymphopenia assessed as not clinically significant, and grade 2 neurosensory toxicity; 6. At least 4 weeks between the last radiation therapy and the first scheduled date of MCLA-128 dosing (with the exception of a maximum of 1 × 8 Gy for pain relief); 7. Complete recovery from major surgery (stable and <grade 2 toxicity tolerated); 8. Laboratory values ​​at screening: Absolute neutrophil count ≥ 1.5 x 10 without colony-stimulating factor support 9 / L; b. Platelets ≧100×10 9 / L; c. Hemoglobin ≥ 9 g / dL or ≥ 2.2 mmol / L (not transfusion dependent); d. Total bilirubin < 1.5 times the upper limit of normal (ULN) (unless due to Gilbert syndrome); e. For patients with advanced solid tumors with liver metastases, patients with confirmed bone metastases who have isolated elevations in AST (SGOT) ≤ 2.5 x ULN; ALT (SGPT) ≤ 2.5 x ULN; ≤ 5 x ULN; ALP > 5 x ULN are allowed on the study; f. Serum creatinine ≤ 1.5 × ULN or estimated glomerular filtration rate (GFR) > 50 mL / min based on the Cockcroft-Gault formula; g. Coagulation function (INR and aPTT ≤ 1.5 x ULN, unless using therapeutic anticoagulants) h. Urinary protein ≤ 2+ (measured by urine dipstick) or ≤ 100 mg / 24-hour urine; 9. At baseline, be able to provide a mandatory tumor biopsy sample (FFPE), preferably a block from fresh (preferred) or archival tissue. Archival tissue must be collected within 2 years prior to screening, except for NSCLC, which must be within 1 year. 10. An available negative pregnancy test result, as defined by a urine or blood human chorionic gonadotropin (hCG) test, during screening and within 7 days of cycle 1, day 1 in women of childbearing potential (defined as women aged 50 years or younger or with a history of amenorrhea for 12 months or less prior to study entry); 11. Sexually active male and female patients of childbearing potential must agree to use effective methods of birth control (e.g., barrier methods using spermicides, oral or non-oral contraceptives, and / or intrauterine devices) for the entire duration of the study and for 6 months after the last dose of MCLA-128. Note that infertility in female patients should be confirmed in the patient's medical record and defined as any of the following: surgical hysterectomy with bilateral oophorectomy, bilateral tubal ligation, natural menopause with last menstrual period >1 year ago; radiation-induced oophorectomy with last menstrual period >1 year ago; chemotherapy-induced menopause with a 1-year interval since last menstrual period; 12. Ability to give written informed consent prior to any study-specific screening procedures, with the understanding that the patient may withdraw that consent at any time without prejudice; 13. Able to understand mandatory and optional protocol requirements, willing and able to comply with study protocol procedures, and has signed the master informed consent document. Further consent is required for any optional biopsy sampling (tissue and / or blood) and long-term sample storage; 14. Patients with metastatic cancer who have disease progression after treatment with all available therapies known to have clinical benefit. 15. Unresectable or metastatic NSCLC meeting one of the following criteria: Biopsy-proven invasive mucinous adenocarcinoma (IMA). Note: IMA patients who did not undergo pre-screening testing for NRG1 fusions may enter the study. or NSCLC with a reported NRG1 fusion, as determined by molecular profiling using methods such as PCR, next generation sequencing [DNA or RNA] or FISH, in patients with no known driver mutations or fusions in the EGFR / ALK genes in eligible local laboratories. 16. Reported disease progression as assessed by the investigator on at least one line of standard therapy in the locally advanced or metastatic setting.

[0265] Statistical analysis: Part 1 and Part 2 Antitumor and clinical benefit variables will be summarized descriptively for each group in Part 2. Where appropriate, variables will be presented in terms of absolute and relative change from baseline. Categorical data will be presented as percentages and frequency tabulations.

[0266] If appropriate, data from patients receiving what is identified as the MTD or MRD during Part 1 and patients receiving the same dose in Part 2 can be pooled together as well as pooled independently.

[0267] The frequency and nature of serious and non-serious AEs will be assessed in absolute and relative frequency and coded according to the MedDRA medical terminology.

[0268] Part 1 Data evaluation will be descriptive in nature. Patient demographics, disease characteristics, and pharmacokinetic and pharmacodynamic variables will be summarized at each dose level. The frequency and nature of DLTs will also be summarized at each dose level.

[0269] Part 2 With N=20 per cohort in Part 2, an observed clinically meaningful correlation coefficient of at least 0.38 is distinguishable from zero with 95% confidence; smaller observed clinically insignificant correlations are not distinguishable from zero. Therefore, 20 subjects per cohort in Part 2 is considered sufficient to investigate the association between the antitumor activity of MCLA-128 and disease-related biomarkers.

[0270] If there are signs of clinical activity, additional patients may be recruited, up to a total of approximately 40. Using 40 patients, for example, a true clinical response rate of 10% to 50% can be estimated with reasonable precision of approximately ±5% to ±8%.

[0271] PK parameters will be summarized for each cohort in Part 1 and each tumor group in Part 2. Arithmetic and geometric means will be provided, along with median, range, SD, and %CV. AUC will be calculated according to the trapezoidal rule. Serum concentration profiles versus time will be plotted for each group.

[0272] (References) Yarden Y, Pines G.2012. The ERBB network: at last, cancer therapy meets systems biology. Nat Rev CancerJul 12;12(8):553-63. Wilson TR, Fridlyand J, Yan Y, Penuel E, Burton L, Chan E, Peng J, Lin E, Wang Y, Sosman J, Ribas A, Li J, Moffat J, Sutherlin DP, Koeppen H, Merchant M, Neve R, Settleman J. 2012. Widespread potential for growth-factor-driven resistance to anticancer kinase inhibitors. Nature. Jul 26;487(7408):505-9. Balko JM, Miller TW, Morrison MM, Hutchinson K, Young C, Rinehart C, Sanchez V, Jee D, Polyak K, Prat A, Perou CM, Arteaga CL, Cook RS. 2012. The receptor tyrosine kinase ErbB3 maintains the balance between luminal and basal breast epithelium. Proc Natl Acad Sci U S A. Jan 3;109(1):221-6. Zhang H, Berezov A, Wang Q, Zhang G, Drebin J, Murali R, Greene MI. 2007. ErbB receptors: from oncogenes to targeted cancer therapies. J Clin Invest. Aug;117(8):2051-8. Sergina NV, Rausch M, Wang D, Blair J, Hann B, Shokat KM, Moasser MM. 2007. Escape from HER-family tyrosine kinase inhibitor therapy by the kinase-inactive HER3. Nature. Jan 25;445(7126):437-41. Junttila TT, Akita RW, Parsons K, Fields C, Lewis Phillips GD, Friedman LS, Sampath D, Sliwkowski MX. 2009. Ligand-independent HER2 / HER3 / PI3K complex is disrupted by trastuzumab and is effectively inhibited by the PI3K inhibitor GDC-0941. Cancer Cell. May 5;15(5):429-40. Ocana A, Vera-Badillo F, Seruga B, Templeton A, Pandiella A, Amir E. 2013. HER3 overexpression and survival in solid tumors: a meta-analysis. J Natl Cancer Inst. Feb 20;105(4):266-73. Junttila, T. T., K. Parsons, et al. (2010). "Superior In vivo Efficacy of Afucosylated Trastuzumab in the Treatment of HER2-Amplified Breast Cancer." Cancer Research 70(11): 4481-4489 Merchant et al. Nature Biotechnology, Vol. 16 July 1998 pp 677-681 Adelaide et al. (2003) Genes Chromosome Cancer, 37(4), 333. Agus et al. (2002) Cancer Cell 2(2), 127. Birnbaum et al (2003). Lancet Oncol 4: 639-642. Chua et al (2009). Oncogene 28, 4041-4052 Cooke et al (2008). BMC Cancer 8: 288. Duruisseaux et al. (2016) NRG1 fusion in a French cohort of invasive mucinous lung adenocarcinoma. Canc Med. Falls DL. (2003). Exp Cell Res 284: 14-30. Fernandez-Cuesta et al. (2014) Canc Disc. 4(4), 415. Fernandez-Cuesta and Thomas (2015). Clinical Cancer Research 21(9): 1989-1994. Hayes and Gullick (2008). J Mammary Gland Biol Neoplasia 13:205-214. Jung et al. (2015) J Thor Oncol 10(7), 1107. Juntilla et al. (2009) Cancer Cell 15(5), 429. Pole et al (2006). Oncogene 25:5693-5706. Sanchez-Valdivieso et al. (2002) Br J Canc, 86(8), 1362. Schoeberl et al. (2010) Canc Res 70(6), 2485. Weinstein et al. (1998). Oncogene 17:2107-2113.

Claims

1. A composition comprising a bispecific antibody comprising a first antigen-binding site capable of binding to the extracellular portion of ErbB-2 and a second antigen-binding site capable of binding to the extracellular portion of ErbB-3 for the treatment of an individual having ErbB-2 and ErbB-3 positive cells, wherein the cells contain an NRG1 fusion gene comprising at least a portion of the NRG1 gene fused to a sequence derived from a different chromosomal location, the first antigen-binding site capable of binding to Domain I of ErbB-2, and the first antigen-binding site is capable of binding to Domain I of ErbB-2, and comprises CDR1 having the amino acid sequence AYYIN, CDR2 having the amino acid sequence RIYPGSGYTSYAQKFQG, and CDR3 of the ErbB-2-specific heavy chain variable region MF3958. a CDR3 sequence comprising the amino acid sequence PPVYYDSAWFAY, and the second antigen-binding site is capable of binding to domain III of ErbB-3 and comprises an ErbB-3-specific heavy chain variable region MF3178, a CDR1 having the amino acid sequence GYYMH, a CDR2 having the amino acid sequence WINPNSGGTNYAQKFQG, and a CDR3 sequence having the amino acid sequence DHGSRHFWSYWGFDY, and both variable domains comprise a light chain variable region comprising a CDR1 having the amino acid sequence RASQSISSYLN, a CDR2 having the amino acid sequence AASSLQS, and a CDR3 having the amino acid sequence QQSYSTPPT.

2. The composition described in claim 1, wherein the bispecific antibody is a full-length antibody of the human IgG subclass.

3. The composition described in claim 1 or 2, wherein the bispecific antibody exhibits ADCC activity.

4. The composition of claim 1, wherein the bispecific antibody is non-fucosylated.

5. 5. The composition of any one of claims 1 to 4, wherein the NRG1 fusion gene comprises at least the 3' end of the NRG1 gene fused to 5' sequences derived from a different chromosomal location.

6. The composition of any one of claims 1 to 5, wherein the cells are cancer cells.

7. The composition of claim 6, wherein the cancer cells are driven by an NRG1 fusion.

8. 8. The composition of claim 1, wherein the cells are breast cancer cells, ovarian cancer cells, lung cancer cells such as non-small cell lung cancer cells, or metastases thereof.

9. The method of claim 8, wherein the first antigen-binding site binds to ErbB-2 and comprises the heavy chain variable region sequence QVQLVQSGAEVKKPGASVKLSCKASGYTFTAYYINWVRQAPGQGLEWIGRIYPGSGYTSYAQKFQGRATLTADESTSTAYMELSSLRSEDTAVYFCARPPVYYDSAWFAYWGQGTLVTVSS; and 9. The composition of any one of claims 1 to 8, wherein the second antigen-binding site binds to ErbB-3 and comprises the heavy chain variable region sequence QVQLVQSGAEVKKPGASVKVSCKASGYTFTGYYMHWVRQAPGQGLEWMGWINPNSGGTNYAQKFQGRVTMTRDTSISTAYMELSRLRSDDTAVYYCARDHGSRHFWSYWGFDYWGQGTLVTVSS.

10. 1. A composition comprising a bispecific antibody comprising a first antigen-binding site capable of binding to the extracellular portion of ErbB-2 and a second antigen-binding site capable of binding to the extracellular portion of ErbB-3 for the treatment of an individual having or at risk of having an ErbB-2 and ErbB-3 positive tumor, wherein the treatment is characterized in that cells of the tumor express an NRG1 fusion gene comprising at least the 3' end of the NRG1 gene fused to a 5' sequence derived from a different chromosomal location; The composition comprises: the first antigen-binding site capable of binding to domain I of ErbB-2 and comprising a CDR1 having the amino acid sequence AYYIN, a CDR2 having the amino acid sequence RIYPGSGYTSYAQKFQG, and a CDR3 sequence having the amino acid sequence PPVYYDSAWFAY of the ErbB-2-specific heavy chain variable region MF3958; and the second antigen-binding site capable of binding to domain III of ErbB-3 and comprising a CDR1 having the amino acid sequence GYYMH, a CDR2 having the amino acid sequence WINPNSGGTNYAQKFQG, and a CDR3 sequence having the amino acid sequence DHGSRHFWSYWGFDY of the ErbB-3-specific heavy chain variable region MF3178; and both variable domains comprising a light chain variable region comprising a CDR1 having the amino acid sequence RASQSISSYLN, a CDR2 having the amino acid sequence AASSLQS, and a CDR3 having the amino acid sequence QQSYSTPPT.

11. The composition of claim 10, wherein the bispecific antibody is a full-length antibody of the human IgG subclass.

12. The composition of claim 10 or 11, wherein the bispecific antibody exhibits ADCC activity.

13. The composition of any one of claims 10 to 12, wherein the bispecific antibody is nonfucosylated.

14. The method of claim 14, wherein the first antigen-binding site binds to ErbB-2 and comprises the heavy chain variable region sequence QVQLVQSGAEVKKPGASVKLSCKASGYTFTAYYINWVRQAPGQGLEWIGRIYPGSGYTSYAQKFQGRATLTADESTSTAYMELSSLRSEDTAVYFCARPPVYYDSAWFAYWGQGTLVTVSS; and 14. The composition of any one of claims 10 to 13, wherein the second antigen-binding site binds to ErbB-3 and comprises the heavy chain variable region sequence QVQLVQSGAEVKKPGASVKVSCKASGYTFTGYYMHWVRQAPGQGLEWMGWINPNSGGTNYAQKFQGRVTMTRDTSISTAYMELSRLRSDDTAVYYCARDHGSRHFWSYWGFDYWGQGTLVTVSS.

15. A composition described in any one of claims 10 to 14, wherein the tumor is gastric cancer, colorectal cancer, colon cancer, gastroesophageal cancer, esophageal cancer, endometrial cancer, ovarian cancer, breast cancer, liver cancer, lung cancer including non-small cell lung cancer, clear cell sarcoma, salivary gland cancer, head and neck cancer, brain cancer, bladder cancer, pancreatic cancer, prostate cancer, kidney cancer, skin cancer or melanoma.

16. 16. The composition according to any one of claims 10 to 15, wherein the tumor is breast cancer, ovarian cancer, lung cancer such as non-small cell lung cancer, or a metastasis thereof.

17. 17. The composition of any one of claims 1 to 16, wherein the NRG1 fusion gene expresses a protein comprising an NRG1 EGF-like domain.

18. 18. The composition of any one of claims 1 to 17, wherein the NRG1 fusion gene is a fusion of NRG1 with a gene on human chromosome 8.

19. 19. The composition of claim 18, wherein the gene on human chromosome 8 encodes an excreted protein or a cell membrane-associated protein.

20. A composition described in any one of claims 1 to 19, wherein the NRG1 fusion gene is a fusion of the 3' end of the NRG1 gene with the 5' sequence of one of the genes selected from the group consisting of CD74; DOC4; TNFRSF10B; CLU; VAMP2; SLC3A2; RBPMS; WRN; SDC4; KIF13B; SLECA2; PDE7A; ATP1B1; CDK1; BMPR1B; MCPH1 and RAB2IL1.

21. 21. The composition of any one of claims 1 to 20, wherein the cells are of epithelial origin.

22. 22. The composition of any one of claims 1 to 21, wherein the treatment further comprises administering to the individual an ErbB-1 inhibitor, preferably cetuximab.

23. The composition of any one of claims 1 to 22, wherein the first antigen-binding site and the second antigen-binding site comprise the light chain variable region sequence DIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASSLQSGVPSRFSGSGSGTDFTLTISSLQPEDFATYYCQQSYSTPPTFGQGTKVEIK.

Images