WO2022072697A2 - Bispecific binding molecules 2 - Google Patents

Abstract

Bispecific binding proteins that bind IL-Ιβ or IL-1R and a checkpoint protein are provided, together with methods of making the proteins. The checkpoint protein may be PD-1 or PD-L1. The binding proteins may have immunoglobulin-like structures that contain human Fab or scFv domains. A novel bispecific antibody format is provided. Methods of making the binding proteins are provided, together with pharmaceutical compositions containing the proteins. The binding proteins and pharmaceutical compositions may be used for treating or preventing diseases such as cancer.

Description

Bispecific Binding Molecules 2

Field of the invention

Bispecific binding molecules are provided that are useful for treating cancer and other diseases.

Background

Immune checkpoints proteins act as immune system regulators and are a key part of the mechanism of self-tolerance. Inhibitory immune checkpoint proteins include: Programmed cell death protein 1 (PD-1); programmed cell death ligand 1 (PD-L1); Adenosine A2A receptor (A2AR); B7-H3 (CD276); B7-H4 (VTCN1), B and T Lymphocyte Attenuator (BTLA); Cytotoxic T-Lymphocyte- Associated protein 4 (CTLA-4, CD 152); Indoleamine 2,3 -dioxygenase (IDO); Killer-cell Immunoglobulin-like Receptor (KIR); Lymphocyte Activation Gene-3 (LAG3); Nicotinamide adenine dinucleotide phosphate NADPH oxidase isoform 2 (N0X2); T-cell Immunoglobulin domain and Mucin domain 3 (TIM-3); V-domain Ig suppressor of T cell activation (VISTA); Sialic acid-binding immunoglobulin-type lectin 7 (SIGLEC7, CD328); and Sialic acid-binding immunoglobulin- type lectin 9 (SIGLEC9, CD329).

Programmed cell death protein 1 (PD-1) is a receptor protein expressed on T-cells that acts as an immune checkpoint. PD-1 expression is upregulated on activated T cells as part of the mechanism of immune tolerance. The ligand for PD-1 is programmed cell death ligand 1 (PD-L1), and binding of PD-L1 to PD-1 transmits an inhibitory signal that reduces the activation and proliferation of antigen-specific T-cells in lymph nodes, and reduces apoptosis in regulatory T cells. Tumor cells often overexpress PD-L1 as a mechanism for avoiding immune surveillance. Monoclonal antibodies that inhibit binding between PD-1 and PD-L1 - by binding either the ligand or receptor - have been shown to be effective either as monotherapy or in combination with other agents, in subpopulations of patients with a number of cancers, while being ineffective or refractory in many other cancers. Pembrolizumab is a humanized antibody that was first approved by the FDA in 2014 and that is used for treatment of a variety of cancers where the tumor cells express elevated PD-L1. Robert et al., N Engl. J Med. 372:2521-2532 (2015) Nivolumab is a fully human antibody that was first approved by the FDA in 2014 and also is used for treating a variety of cancers.

Tumor-associated macrophages (TAMs) are a class of immune cells present in high numbers in the tumor microenvironment (TME), and are associated with cancer-related inflammation. Expression of PD-1 on TAM cells has been shown to decrease macrophage phagocytosis of tumor cells and confers “immunity” on the tumor cells. Gordon et al, Nature 545:495 (2017).

Interleukin- ip (IL-ip) is a pro-inflammatory cytokine that is associated with chronic and acute inflammation and plays an important role in multiple inflammation-associated diseases. Elevated levels of IL-1 have also been shown to recruit TAM cells and myeloid- derived suppressor cells (MDSC) to the TME and to promote tumor growth and metastasis in breast cancer. Guo et al., Sci. Rep. 6, 36107; doi: 10.1038/srep36107 (2016). In other studies, lung lesions have been shown to be populated with TAM whose pro-tumor activity is up- regulated by activation of the NLRP3 inflammasome and the release of IL-ip. (Terlizzi et al., Oncotarget 7 : 58181 (2016)). Finally, IL-ip has been shown to promote a pro-tumor phenotype of TAM cells and the level of the cytokine has been correlated to tumor size & stage in renal cell carcinoma (Chittezhath et al., Immunity 41:815 (2014)).

In mice deficient in IL-ip, fewer animals developed tumors and tumor development was slower. Apte, et al., European Journal of Cancer, 42:751 (2006). Additionally, it was shown that the lung cancer risk genotype IL-1 P-3 ITT results in increased expression of IL- ip, providing a microenvironment with elevated inflammatory stimuli and increase lung cancer risk. Bhat et al., Meta Gene 2: 123 (2014). An IL-1 receptor antagonist was shown to suppress metastasis and tumor proliferation by inhibiting angiogenic factors such as VEGF and IL-8. Konishi et al., Oncology 68: 138 (2005); Lewis et al, J. Transl. Med. 4:48 (2006). Canakinumab, a monoclonal antibody that inhibits IL-ip activity, has been shown to reduce incident lung cancer and lung cancer mortality. Ridker et al., Lancet, 390.P 1833-1842, (2017).

Engineered bispecific monoclonal antibodies are non-naturally occurring proteins containing immunoglobulin domains that can simultaneously bind to two different types of antigen. Bispecific antibodies can be made in a variety of format and have been used, for example, for cancer immunotherapy and drug delivery. See, for example, Fan etal., J. Hemat. Oncol. 8:130 (2015); Brinkmann and Kontermann, mAbs 9:182 (2017); and Spiess et al., Molecular Immunology, 67: 95-106 (2015).

Summary of the Invention

What is provided is a binding protein containing a first human immunoglobulin binding domain and a second human immunoglobulin binding domain, where the first human immunoglobulin binding domain specifically binds and inhibits activation of PD-1 or PD-L1, where the second human immunoglobulin binding domain specifically binds and inhibits the activity of IL-P, where the binding protein has a domain structure as shown in Figure 1A, Figure IB, Figure 1C, Figure 2A, Figure 2B, Figure 3 or Figure 4, and where the binding protein contains:

(a) CHI domains of the IgGl isotype and Fc domains of the IgG4 isotype and/or

(b) a disulfide bond between the VH and VL domains of the scFV when present and/or

(c) a mutant Fc region that improves antibody half-life, where the Fc region comprises: i. an N434A mutation ii. a YTE (M252Y/S254T/T256E) mutation, or iii. an M428L/N434S mutation.

The binding protein may have the domain structure as shown in Figure 1A, Figure IB, Figure 1C, Figure 2A, or Figure 2B, and the binding protein may contain a disulfide bond between the VH and VL domains of the scFV. In such molecule the disulfide bond may be between:

(a) the VH44 and VL100 residues;

(b) the VH105 and VL43 residues;

(c) the VH101 and VL46 residues;

(d) the VH45 and VL98 residues; or

(e) the VH104 and VL43 residues of the scFV moiety. The scFv moiety may contain a VH and VL domain pairing selected from the group consisting of:

VH44/VL100

VH44:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKCLEWVAVIWYDG SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWGQGTLVT vss

VL100:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGI PARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGCGTKVEIK

VH105/VL43 VH105:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKGLEWVAVIWYDG SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWGCGTLVT vss

VL43:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQCPRLLIYDASNRATGIP

ARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGQGTKVEIK

VH101/VL46

VH101:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKGLEWVAVIWYDG

SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDCYWGQGTLVT VSS

VL46:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRCLIYDASNRATGI

PARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGQGTKVEIK

VH45/VL98

VH45:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKGCEWVAVIWYDG

SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWGQGTLVT VSS

VL98:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGI PARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTCGQGTKVEIK . and

VH104/VL43

VH104:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKGLEWVAVIWYDG

SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWCQGTLVT VSS

VL43:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQCPRLLIYDASNRATGIP

ARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGQGTKVEIK The binding protein as described above may contain CHI domains of the IgGl isotype and Fc domains of the IgG4 isotype and/or may contain an N434A, YTE (M252Y/S254T/T256E), or M428L/N434S mutation in the Fc region.

Examples of the proteins described above include: ITG101-ITG110, ITG102C, ITH101, ITH102, ITH501, ITH602, ITI102, ITI202, ITI208, ITI209, ITG101-ITG110, ITG201-ITG204, ITG102C, ITH101, ITH102, ITH201-ITH204, ITH501, ITH602, ITI101- ITI103, ITI202-ITI209 and ITI212-ITI215.

Where the binding protein has a domain structure as shown in Figure IB or Figure 2B, the linker between the VH and VL domains of the scFv moiety may consist of 6-12 amino acids. Examples of such binding protein include ITH601 and ITH602.

Other embodiments of the binding proteins described above include ITG301, ITG302, ITG401-ITG406, ITH103, ITH601, ITI201, ITI210, and ITI211.

Also provided are nucleic acids encoding a binding domain of a binding protein as described above, and methods for the preparation of a binding protein as described above, including the steps of a) transforming a host cell with vectors containing nucleic acid molecules encoding the binding domains; b) culturing the host cell under conditions that allow synthesis of the binding protein; and c) recovering the binding protein from the culture. Host cells also are provided containing vectors containing nucleic acid molecules encoding one or more binding domains as described above.

Pharmaceutical compositions are provided containing a binding protein as described above and a pharmaceutically acceptable excipient. Also provided are methods of treating cancer in a subject including the step of administering a binding protein or composition as described above to a subject in need thereof. The methods of treatment may further include administering an antitumor agent to the subject. The cancer may be lung cancer, such as small cell lung cancer, combined small-cell lung carcinoma or non-small cell lung cancer. The non-small cell lung cancer maybe, for example squamous cell lung carcinoma, large cell lung carcinoma, lung adenocarcinoma, pulmonary pleomorphic carcinoma, lung carcinoid tumor, salivary gland carcinoma, or carcinoma NOS (not otherwise specified). In other embodiments the cancer may be combined small-cell lung carcinoma, extrapulmonary smallcell carcinoma, extrapulmonary small-cell carcinoma localized in the lymph nodes or smallcell carcinoma of the prostate. The cancer may have microsatellite instability.

What is also provided are methods of treatment as described above where the subject has previously been treated with cancer immune therapy or has been found to be resistant to the therapy; or where the subject has previously been treated with cancer immune therapy or has been found to be refractory to cancer immune therapy. The cancer immune therapy may be treatment with at least one immune checkpoint inhibitor. The subject may also be treated with an additional anti-tumor therapy, such as chemotherapy, immune therapy, treatment with biologies or small molecules, vaccination, or a cell therapy.

Further provided are methods preventing or reducing the risk of cancer in a subject at risk thereof, by administering to the subject an effective amount of a binding protein or a composition as described above. The cancer may be lung cancer, and the subject may previously have been diagnosed with cancer and be in remission, or have been previously treated for cancer. The subject may, for example, be considered to be at risk of cancer due to environmental exposure, tobacco use or exposure, genetic mutation, or a family history of cancer. The cancer may be, for example, esophageal, pancreatic, hepatic, colorectal, breast, and ovarian cancer, or multiple myeloma or precancerous conditions.

In a further embodiment, a binding protein is provided that contains an scFV binding domain, where the scFv binding domain contains a VH domain containing the sequence QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKCLEWVAVIWYDG SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWGQGTLVT VSS, coupled via an amino acid linker to a VL domain containing the sequence EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGI PARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGCGTKVEIK, where the VL domain is N-terminal to the VH, and where the linker consists of 14 to 25 amino acids. The binding protein may have the domain structure shown in Figure 1A, Figure IB, Figure 1C, Figure 2 A, or Figure 2B.

Other objects, features and advantages of the present disclosure will become apparent from the following detailed description. It should be understood, however, that the detailed description and the specific examples are given by way of illustration only, since various changes and modifications within the spirit and scope of the disclosure will become apparent to those skilled in the art from this detailed description.

Brief description of the Drawings

Figure 1 A shows the chain structure of a BiS2 bispecific antibody. Figure IB shows the chain structure of a diabody based on the BiS2 structure (BiS2DAB). The VL-VH linkers are too short to allow formation of a conventional single chain Fv and instead, the VH of one chain dimerizes with the VL of the other.

Figure 1C shows the structure of a BiSl bispecific antibody.

Figure 2A shows the chain structure of a BiS3 bispecific antibody.

Figure 2B shows the chain structure of a BiS3 diabody (BiS3DAB).

Figure 3 shows the chain structure of a FIT-Ig bispecific antibody.

Figure 4 shows the chain structure of a FAT-Ig bispecific antibody.

Figure 5 shows the amino acid sequences of bispecific antibodies that bind IL-ip and PD-1. Figure 5 A shows the amino acid sequence of ITA101; Figure 5B shows the amino acid sequence of ITA102; Figure 5C shows the amino acid sequence of ITA103 and ITA104; Figure 5D shows the amino acid sequence of ITA201; Figure 5E shows the amino acid sequence of ITA202; Figure 5F shows the amino acid sequence of ITA203; Figure 5G shows the amino acid sequence of ITA204; Figure 5H shows the amino acid sequence of ITA301; Figure 51 shows the amino acid sequence of ITA302; Figure 5 J shows the amino acid sequence of ITA303; Figure 5K shows the amino acid sequence of ITA304; ITA401 shows the amino acid sequence of ITC401, ITC402, ITC 403, ITC404, and ITC 405; Figure 5L shows the amino acid sequence of ITA402; Figure 5M shows the amino acid sequence of ITA403; Figure 5N shows the amino acid sequence of ITA404; Figure 50 shows the amino acid sequence of ITB101; Figure 5P shows the amino acid sequence of ITB102; Figure 5Q shows the amino acid sequence of ITB103; Figure 5R shows the amino acid sequence of ITB104; Figure 5S shows the amino acid sequence of ITB105; Figure 5T shows the amino acid sequence of ITB106; Figure 5U shows the amino acid sequence of ITB107; Figure 5V shows the amino acid sequence of ITB108; Figure 5W shows the amino acid sequence of ITB109; Figure 5X shows the amino acid sequence of ITE101 ; Figure 5Y shows the amino acid sequence of ITE102; Figure 5Z shows the amino acid sequence of ITE201; Figure 5AA shows the amino acid sequence of ITE202; Figure 5BB shows the amino acid sequence of ITE301; Figure 5CC shows the amino acid sequence of ITE302; Figure 5DD shows the amino acid sequence of ITF101; Figure 5EE shows the amino acid sequence of ITF102; Figure 5FF shows the amino acid sequence of ITF103; Figure 5GG shows the amino acid sequence of ITF201; Figure 5HH shows the amino acid sequence of ITF202; Figure 511 shows the amino acid sequence of ITF301; Figure 5JJ shows the amino acid sequence of ITF302; Figure 5KK shows the amino acid sequence of ITF401; Figure 5LL shows the amino acid sequence of ITF402; Figure 5MM shows the amino acid sequence of ITF403. Figures 6A-6C_show a table of known antibodies and binding molecules that bind IL- 1P, IL-1R, PD-1 or PD-Ll.

Figure 7 shows binding of bispecific antibodies to cell membrane-bound PD-1 and soluble IL-i simultaneously in a multiple-dose sandwich assay, as detected by flow cytometry. Variable binding affinities of ITA series of bispecific antibodies are demonstrated, and all the binding affinities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figure 8 shows binding of bispecific antibodies to cell membrane-bound PD-1 and soluble IL-i simultaneously in a multiple-dose sandwich assay, as detected by flow cytometry. Variable binding affinities of ITC, ITD and ITE series of bispecific antibodies are demonstrated, and all the binding affinities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figures 9A-C show binding of bispecific antibodies to cell membrane-bound PD-1 in a multiple-dose sandwich assay, as detected by flow cytometry. Variable binding affinities of ITB and ITF series of bispecific antibodies are demonstrated, and all the binding affinities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figures 10A and 10B show binding of bispecific antibodies to cell membrane-bound PD-1 and soluble IL-i simultaneously in a multiple-dose sandwich assay, as detected by flow cytometry. Variable binding affinities of ITB and ITF series of bispecific antibodies are demonstrated. All the binding affinities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figure 11 shows that bispecific antibodies block PD-1 activity in a PD-1/PD-L1 reporter assay. Variable blockade activities are demonstrated for the ITA series of bispecific antibodies. All the blockade activities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figures 12A and 12B show that bispecific antibodies block IL-ip activity in an IL-ip functional assay. Variable blockade activities are demonstrated of the ITA series of bispecific antibodies. All the blockade activities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figures 13 A and 13B show binding of a fixed concentration of bispecific antibodies to cell membrane-bound PD-1 in a multiple-dose soluble IL-ip sandwich assay, with detection of soluble IL-i by flow cytometry. All binding affinities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figures 14A and 14B show binding of bispecific antibodies to cell membrane-bound PD-1 in a multiple-dose of bispecific antibody sandwich assay with detection against antibody constant domains, as detected by flow cytometry. Variable binding affinities of ITG, H, and I series of bispecific antibodies are demonstrated and all binding affinities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figures 15A-H show that bispecific antibodies block IL-ip activity in an IL-ip functional assay. Variable blockade activities are demonstrated of the ITA, B, C, D, E, F, G, and H series of bispecific antibodies. All blockade activities are substantially higher than control human IgG, except ITA201 and ITG202. The table lists the EC50 values (nM) of the samples on the graph.

Figure 16A-D show that bispecific antibodies block PD-1 activity in a PD-1/PD-L1 reporter assay. Variable blockade activities are demonstrated for the ITE, F, G, and H series of bispecific antibodies. All blockade activities are substantially higher than control human IgG. The table lists the EC50 values (nM) of the samples on the graph.

Figure 17 shows the sequences of bispecific antibodies that bind to PD-1 and IL-ip. These sequences, when assembled according to Table 2 (continued), generated the molecules listed in Table 2 (continued) from the ITG, ITH, and ITI series.

Detailed Description

Bispecific binding proteins are provided that contain at least one first binding domain that specifically binds an immune checkpoint protein and at least one second binding protein that binds IL-ip. The immune checkpoint protein may be, for example, PD-1, PD-L1, A2AR, B7-H3 (CD276), B7-H4 (VTCN1), BTLA, CTLA-4 (CD152), IDO, KIR, LAG3, N0X2, TIM-3, VISTA, SIGLEC7, (CD328), or SIGLEC9 (CD329). Advantageously, the checkpoint protein is PD-1 or PD-L1.

A binding protein simultaneously binds the checkpoint protein PD-1 or PD-L1 and IL-ip or IL-1R and thereby inhibits binding between PD-1 on CD8 T-cells and PD-L1 on a target cell, such as a tumor cell, and IL-ip activity. Simultaneous inhibition of IL-ip activity and PD-1/PD-L1 binding in this fashion provides for improved methods for cancer treatment. The first and second binding domains advantageously are human antibody variable domains. Novel bispecific binding protein formats also are provided that allow for specific binding of two antigens including, but not limited to, a checkpoint protein and a cytokine such as IL-ip. Methods of using the bispecific binding proteins for treating disease, such as cancer, also are provided.

Binding domains

The binding domains that may be used in the binding proteins as described herein may be in any format that specifically binds the target protein. Advantageously, however, the binding domains are derived from the variable domains of human immunoglobulin molecules. Specifically, the binding domains may be derived from an antibody that binds a checkpoint protein and an antibody that binds IL-ip or IL-1R. Methods of making fully human antibodies that bind a preselected antigen are well known in the art. For example, human antibodies can be selected from large libraries of antibodies displayed on filamentous phage, and the heavy and light chain variable regions of the selected antibodies identified by methods that are well known. See, for example, Winter et al., Annual Review of Immunology 12:433-455 (1994). The nucleic acids encoding these variable regions are then used in constructing the genes encoding the bispecific binding proteins described herein. Alternatively, antibody variable domains from known antibodies may be used. In particular, human antibodies again IL-ip, PD-1 and PD-L1 have been approved for use in treating a variety of disease states in humans, and the variable regions from these antibodies can be used to construct the bispecific binding proteins described herein. Examples of suitable antibodies are shown below in Tables 1 and 6.

Alternatively, the CDR regions from an antibody with known specificity against IL- ip, IL-1R, PD-1 or PD-L1 can be inserted into known human framework regions using methods of CDR grafting that are well known in the art. See, for example, Williams and Matthews, “Humanising Antibodies by CDR Grafting” in Antibody Engineering (Kontermann and Dubel, Eds.) pp 319-339 (Springer, 2010). These CDR regions may be derived from the CDR regions shown in Table 1 or from the CDR regions of the antibodies described in Table 6. The skilled artisan will recognize that other antibodies that specifically bind IL-ip, IL-1R, PD-1 or PD-L1 exist in addition to those described herein, and that the CDR regions of those antibodies may be used in constructing the binding proteins as described herein.

With respect to IL-ip, canakinumab is an FDA-approved human antibody that binds IL-ip, and the amino acid sequences of the heavy and light chain variable regions are well known. See Rondeau et al., MAbs 7:1151 (2015). The entire heavy and light chain variable regions of canakinumab can be used to construct the bispecific protein proteins as described herein; alternatively, the CDR regions of canakinumab can be inserted into alternative human variable framework sequences using methods of CDR grafting that are well known in the art. See, for example, Winter and Harris, Trends in Pharmacological Sciences 14: 139-143 (1993). An alternative IL-ip binding antibody is the humanized SK48-E26 antibody described in WO1995/01997.

With respect to IL-1R, anakinra is an FDA-approved, recombinant, nonglycosylated form of human interleukin -1 receptor antagonist (IL-lRa). Compared to native human IL- IRa, anakinra contains an extra N-terminal methionine residue. Anakinra competitively inhibits binding of IL-la and IL-lp to the IL-1 receptor type 1. This IL-1R1 binding domain can be used in construction of bispecific proteins to block binding of IL-1 p to the IL-1 receptor. The sequence of the IL-1R1 binding portion of anakinra, which may be used as a suitable binding domain, is

MRPSGRKSSKMQAFR1WDVNQKTFYLRNNQLVAGYLQGPNVNLEEKSDVVP1EPHALFLG1H GGKMCLSCVKSGDETRLQLEAVNITDLSENRKQDKRFAF1RSDSGPTTSFESAACPGWFLCTAMEADQP VSLTNMPDEGVMVTKFYFQEDE (SEQ ID NO:72)

With respect to PD-1, pembrolizumab is a humanized antibody that was first approved by the FDA in 2014 and that is used for treatment of a variety of cancers where the tumor cells express elevated PD-1. Nivolumab is a fully human antibody that was first approved by the FDA in 2014 and also is used for treating a variety of cancers. Cemiplimab is a human antibody that was first approved in 2018 for treatment of metastatic cutaneous squamous cell carcinoma. The amino acid sequences for the heavy and light chain variable domains of pembrolizumab, nivolumab, and cemiplimab, are known, as are the sequences of the CDR regions.

With respect to PD-L1, durvalumab is a human antibody that was first approved in 2017 for treatment of metastatic urothelial cancer. Atezolizumab was first approved in 2016 and is used for treating lung cancer. The amino acid sequences for the heavy and light chain variable domains of durvalumab and atezolizumab are known, as are the sequences of the CDR regions.

The sequences of the variable domains and CDR regions of pembrolizumab, nivolumab, cemiplimab, atezolizumab, avelumab, durvalumab and canakinumab are shown in Table 1 below. A list of other known antibodies again IL-ip, PD-1 and PD-L1 is shown in

Figure 6.

Table 1:

Bispecific binding protein structure

Once the appropriate binding domains are selected, they are incorporated into a format that contains at least one binding domain that specifically binds IL-ip or IL-1R and at least one domain that specifically binds a checkpoint protein. Advantageously, the format of the binding protein is that of a bivalent or multivalent bispecific antibody, containing immunoglobulin variable and constant chain domains arranged to contain two different binding domains, as opposed to the naturally occurring homodimeric structure of a bivalent but monospecific human antibody.

Methods of making bispecific antibodies are well known in the art and are described in, for example, Brinkmann and Kontermann, mAbs 9:182 (2017) and Spiess et al., Molecular Immunology, 67: 95-106 (2015). The bispecific binding proteins described herein can be in any format known in the art that is stable, suitable for administering to a subject, and contains at least one binding domain the binds IL-ip or IL-1R and at least one binding domain that binds a checkpoint protein.

Examples of suitable bispecific binding protein formats known in the art include:

Fc-less bispecific antibody formats, including: two scFv molecules joined by a linker (Kontermann, Acta Pharmacol Sin 26:1-9 (2005)); bispecific single-domain antibody fusion proteins (Weidle et al., Cancer Genomics Proteomics 10:155-68 (2013;)) and diabodies (Atwell et al., Mol Immunol 33:1301-12 (1996)); Fab fusion proteins (Schoonjans et al., J Immuno , 165:7050-7 (2000); and miniantibodies (Pluckthun and Pack, Immunotechnology 3:83-105 (1997) and Muller et al., FEBS Lett 432:45 49(1998));

Asymmetric IgGs with heavy and light chains from two different antibodies (Suresh et al, Methods Enzymol; 121:210-28 (1986)); and

Bispecific IgGs with an asymmetric Fc region, e.g. asymmetric Fc regions using the “knobs into holes” method (Ridgway et al., Protein Eng, 9:617-21 (1996); Shatz et al., MAbs- 5:872-81 (2013)); Sampei et al., PLoS One; 8:e57479 (2013); Spiess et al., Biotechnol; 31:753-8 (20130; Juntilla et a/., Cancer Res. 74:5561-71 (2014); and Sun et al., J Clin Invest. 125):4077-4090 (2015)).

The skilled artisan will recognize that a large number of bispecific antibody formats in addition to the specific formats described above can be used to construct a bispecific antibody that binds IL-ip or IL-1R and a checkpoint protein. Advantageously, the bispecific antibody is either a 2+2 scFv-based structure or a 2+2 Fab-based structure as described in more detail below.

2+2 scFv-based structures

A first 2+2 scFv-based structure is the structure shown in Figure 1 A, referred to herein as BiS2. As shown in Figure 1A, the BiS2 format contains two protein chains:

(1) a heavy chain that contains (from N- to C-terminus): a single chain Fv containing a first VH domain and a first VL domain (arranged VH-VL or VL-VH, i.e. the domains can be in either order), where the scFv binds the first target (IL- 1 , IL-1R or the checkpoint protein, respectively); a second VH domain; and CHI, CH2, and CH3 domains, and

(2) a light chain that contains (from N- to C-terminus): a second VL domain and a CL domain.

The BiS2 protein assembles via non-covalent homodimeric binding of the CH3 and CH2 domains and heterodimeric binding between the CHI and CL domains on the heavy chain and the second VH and VL domains on the light chain. The binding between the CHI and CL domains and the VH and VL domains forms a Fab domain that binds the second target (the checkpoint protein or IL-ip/IL-lR, respectively). Advantageously, disulfide bonds also form between the hinge regions, and between the CHI and CL domains in the same manner as are found in naturally-occurring IgG molecules.

A second 2+2 diabody-based structure is the structure shown in Figure IB, referred to herein as BiS2DAB. In the diabody format the linker between the VL and VH domains of the scFv is too short to allow binding between the VL and VH domains of the same chain, and therefore formation of an inter-chain dimer occurs. The linker between the VL and VH of the scFv typically contains 6-12 amino acids.

As shown in Figure IB, the BiS2DAB format contains two protein chains:

(1) a heavy chain that contains (from N- to C-terminus): a half diabody containing a first VH domain and a first VL domain (arranged VH-VL or VL-VH, i.e. the domains can be in either order), whereupon interacting with the half diabody on a second heavy chain, a diabody is formed and binds the first target (IL-i , IL-1R or the checkpoint protein, respectively); a second VH domain; and CHI, CH2, and CH3 domains, and

(2) a light chain that contains (from N- to C-terminus): a second VL domain and a CL domain.

The BiS2DAB protein assembles via non-covalent homodimeric binding of the CH3 and CH2 domains and heterodimeric binding between the CHI and CL domains on the heavy chain and the second VH and VL domains on the light chain. Binding between the CHI and CL domains and the VH and VL domains forms a Fab domain that binds the second target (the checkpoint protein or IL-ip/IL-lR, respectively). Advantageously, disulfide bonds also form between cysteine residues in the hinge regions, and between the CHI and CL domains in the same manner as are found in naturally-occurring IgG molecules.

A third 2+2 scFv-based structure is the structure shown in Figure 1C, referred to herein as Bi SI. As shown in Figure 1C, the Bi SI format contains two protein chains: (1) a heavy chain that contains (fromN- to C-terminus): VH domain; and CHI, CH2, and CH3 domains, and

(2) a light chain that contains (from N- to C-terminus): a single chain Fv containing a first VH domain and a first VL domain (arranged VH-VL or VL-VH, i.e. the domains can be in either order), where the scFv binds the first target (IL-ip, IL-1R or the checkpoint protein, respectively); a second VL domain and a CL domain.

The BiSl protein assembles via non-covalent homodimeric binding of the CH3 and CH2 domains and heterodimeric binding between the CHI and CL domains on the heavy chain and the second VH and VL domains on the light chain. Binding between the CHI and CL domains and the VH and VL domains forms a Fab domain that binds the second target (the checkpoint protein or IL-ip/IL-lR, respectively). Advantageously, disulfide bonds also form between cysteine residues in the hinge regions, and between the CHI and CL domains in the same manner as are found in naturally-occurring IgG molecules.

A further 2+2 scFv-based structure is the structure shown in Figure 2A, referred to herein as BiS3. As shown in Figure 2A, the BiS3 format also contains two protein chains:

(1) a heavy chain that contains (from N- to C-terminus): a first VH domain; CHI, CH2, and CH3 domains; and a single chain Fv containing a second VH domain and a second VL domain (where the VH and VL domains can be in either order), where the scFv binds the first target (IL-ip or the checkpoint protein, respectively);

(2) a light chain that contains (from N- to C-terminus): a second VL domain and a CL domain.

The BiS3 protein also assembles via homodimeric binding of the CH3 and CH2 domains and heterodimeric binding between the CHI and CL domains on the heavy chain and the second VH and VL domains on the light chain. The binding between the CHI and CL domains and the VH and VL domains forms a Fab domain that binds the second target (the checkpoint protein or IL-ip/IL-lR, respectively). Advantageously, disulfide bonds also form between the CH2 domains, and between the CHI and CL domains in the same manner as are found in naturally-occurring IgG molecules.

The BiS3 protein also can contain a diabody structure when the linker between the VH and VL domains of the scFv is too short to allow intra-chain scFV formation. The structure of this diabody-containing binding protein is shown in Figure 2B. 2+2 Fab-based structures

A first 2+2 Fab-based structure is the structure shown in Figure 3A, referred to herein as FIT-Ig (see Gong et al., MABS, 2017, 9:1118-1128 (2017)). As shown in Figure 3A, the FIT-Ig format contains three protein chains:

(1) a heavy chain that contains (from N- to C-terminus): a first VL domain; a first CL domain, a first VH domain, a first CHI domain, and CH2 and CH3 domains;

(2) a light chain that contains (from N- to C-terminus): a second VL domain and a second CL domain; and

(3) an Fd chain that contains (fromN- to C-terminus): a second VH domain and second CHI domain.

In the heavy chain the first CL domain may be linked to the first VH domain via a peptide linker such as, for example, a flexible hydrophilic linker having the sequence (GGGGS)x, where x is 1-5 (SEQ ID NO: 123). Advantageously, however, the linker is absent.

The FIT-Ig protein assembles via: non-covalent homodimeric binding of the CH3 and CH2 domains; heterodimeric binding between the first VH and CHI domains on the heavy chain and the second VL and CL domains on the light chain; and heterodimeric binding between the first VL and CL domains and the second VH and CHI domains on the Fd chain Two identical Fab binding domains are formed by binding of the heavy chain to the light chain, and two distinct but identical Fab domains are formed by binding of the heavy chain to the Fd chain as shown in Figure 3. Advantageously, disulfide bonds also form between the CH2 domains, and between the hinge domains in the same manner as are found in naturally - occurring IgG molecules.

A second 2+2 Fab-based structure is the novel structure shown in Figure 4, referred to herein as FAT-Ig. As shown in Figure 4, the FAT-Ig format contains three protein chains:

(1) a heavy chain that contains (from N- to C-terminus): a first VH domain; a first CHI domain, and CH2 and CH3 domains; plus first VL and a first CL domain, where the first VL and first CL domains are disposed at a solvent exposed loop in the CH2 domain, the CH3 domain, or at the interface of the CH2 and CH3 domains. Advantageously, the first CL and first CL domains are disposed at a solvent-exposed loop in the CH3 domain.

(2) a light chain that contains (from N- to C-terminus): a second VL domain and a second CL domain; and

(3) an Fd chain that contains (fromN- to C-terminus): a second VH domain and second CHI domain. In the heavy chain, the first VL and CL disposed at the solvent-exposed loop are connected to the CH2 domain, CH3 domain, or interface of the CH2 and CH3 domains via flexible peptide linkers. The linkers can have 4-25 amino acids and advantageously comprise (GGGGS)x units, where x=l-5 (SEQ ID NO: 123). Other linkers may also be used, as described in more detail below.

The FAT-Ig protein assembles as shown in Figure 4 via: non-covalent homodimeric binding of the CH3 and CH2 domains; heterodimeric binding of the heavy chain CHI and VH domains with the light chain CL and VL domains; and heterodimeric binding of the heavy chain CL and VL domains with the Fd chain CHI and VH domains. Two identical Fab binding domains are formed by binding of the heavy chain to the light chain, and two distinct but identical Fab domains are formed by binding of the heavy chain to the Fd chain as shown in Figure 4. Advantageously, disulfide bonds also form between the hinge domains, and between the CHI and CL domains in the same manner as are found in naturally-occurring IgG molecules. One of ordinary skill in the art will recognize that the novel FAT-Ig antibody format can be used to bind any two desired antigens, and is not limited to IL-ip/IL-lR and a checkpoint protein.

In each of the four specific binding proteins described above the binding domains that specifically bind IL-ip/IL-lR and the checkpoint protein are disposed asymmetrically within the binding protein i.e. the structure of the binding protein is different when the first binding domain binds IL-ip/IL-lR and the second binding domain binds the checkpoint protein compared to when the first binding domain binds the checkpoint protein and the second binding domain binds IL-ip/IL-lR. Accordingly, each of the four specific binding proteins can exist in two alternative forms for any given pair of binding domains.

Polypeptide Linkers

The domains of the bispecific binding proteins described herein may be joined into contiguous protein chains using linkers. The linkers may be used, for example, to connect the variable heavy and light chains of an scFv, or to connect the CL/VL domains into the heavy chain constant domains in the FAT-Ig format.

Suitable linkers are well known in the art and, when present, advantageously contain at least four amino acids, although longer or shorter linkers may also be used. The linkers advantageously are flexible, hydrophilic and have little or no secondary structure of their own. Linkers may be approximately 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, or approximately 50 residues in length. When multiple linkers are used to interconnect portions of a bispecific protein as described herein, the linkers may be the same, or have different lengths and/or amino acid sequences.

The linker(s) facilitate formation of the desired bispecific binding protein structure. Linkers may contain (Gly-Ser)_x units, where x=l-5. Glutamic acid or lysine residues may also be placed in the linker sequences to increase solubility if necessary. The length of the linker may be varied to facilitate protein folding, target binding, and/or expression. For example, different multiples of (Gly-Ser)x units may be used to improve or optimize protein folding, target binding, and/or expression using methods that are well-known in the art.

As described above for the diabody-containing BiS2 and BiS3 structures, when the linker between the VH and VL domains of the heavy chain terminal scFv structures is short, the VH and VL domains cannot form an intrachain scFv structure, and a divalent interchain structure is formed as shown in Figures IB and 2B. Although the precise length of the linker required to allow formation of the diabody structure can vary, typically the linker contains 6- 12 amino acids.

Modifications to the binding domains

The domains described above can be modified in several ways to improve their binding, improve protein expression, and improve protein stability, and to prolong the half- life of the binding proteins in vivo. Methods for achieving these results include the following:

Chimeric binding proteins

The binding affinities of the various binding domains may be slightly reduced when placed into a bispecific antibody format. To remedy this loss of affinity it is possible to replace one or more of the constant domains with chains from a different antibody isotype. In particular, it is possible to produce chimeric binding chains where the CHI domain of the heavy chain is of the IgGl isotype and the Fc (CH2 and CH3) domains are of the IgG4 isotype. Methods of preparing antibody chains with this chimeric structure are known in the art. scFv disulfide bond formation

The stability of binding between the VL and VH domains of a scFv can be enhanced by formation of a disulfide bond between the VL and VH domains. This requires introduction of cysteine residues into each of the domains. For example, cysteine residues were introduced pairwise into the domains as follows:

(a) VH44 and VL100; (b) VH105 and VL43;

(c) VH101 and VL46;

(d) H45 and VL98; or

(e) VH104 and VL43

It was found that the VH44/VL100 cysteine pair typically was the most effective. Surprisingly it was found that for an scFv derived from an antibody containing the VL of SEQ ID 39 and the VH of SEQ ID 40, stabilization of the scFv by a disulfide bond formed by mutating position 44 of VH and 100 of VL to Cys (44/100DSB) improved yield and/or monomer percentage as measured by SEC post-protein A. Further, the VL-VH orientation appears to be preferred compared to the VH-VL orientation. For example, ITH103 (VL-VH, no disulfide bond) expressed at 11.8 mg/L and 51.9% monomer. ITA102 (VH-VL, no disulfide bond) expressed at 0.9 mg/L and 53% monomer. ITE101 (VH-VL, 44/100DSB) improved yield to 1.7 mg/L and monomer improved slightly to 57%. ITE102 (VL-VH, 44/100DSB) exhibited a yield of 7.5 mg/L and 86% monomer. This represents a significant improvement over the parent molecule ITH103 (that lacked an scFv disulfide bond) in correct assembly. In addition to these variants, ITG108 (VL-VH, 101/46DSB) and ITG109 (VH-VL, 101/46DSB) were made. Again, the VL/VH orientation was favored for yield (7.2 vs. 2.0 mg/L) and monomer percentage (71% vs. 52%). However in this case, ITG108 showed relatively low response in the sandwich FACS binding assay indicating a loss of activity, most likely in binding to PD1 (Fig. 13 A). While insertions of disulfide bonds have been shown to improve stability, it had not been shown to improve monomer percentage or expression (Weatherill, et al Protein Eng. Des. Sei. 25:321-329 (2012)).

Modifications to the Fc to improve in vivo half-life

Modifications to the Fc regions of antibodies previously have been shown to extend the half-life of immunoglobulin molecules via enhanced binding to the MHC class I-like Fc receptor FcRn. See Petkova et al, International Immunology, 18: 1759-1769 (2006). These types of modification also can be introduced into the Fc regions of the binding molecules as described herein. In particular, replacement of the asparagine residue at position 434 of the Fc by alanine can be used to enhance FcRn binding and increase the in vivo half-life of the molecules. Additional modifications that can be made include YTE (M252Y/S254T/T256E). See Dall’Acqua et al, J. Biol. Chem., 281:23514-24 (2006). Another example of half-life extension is M428L/N434S. See Zelevsky, et al, Nat. Biotechnol., 28:157-9 (2010). Preparation of bispecific binding proteins

Methods of making bispecific binding proteins such as Fc-less bispecific antibodies, asymmetric IgGs with heavy and light chains from two different antibodies and bispecific IgGs with an asymmetric Fc region, are well known in the art and are described in the references provided above.

The particular 2+2 bispecific antibodies described above may each be produced using suitable expression constructs in recombinant host cells. Nucleic acids encoding the heavy, light and Fd chains can be prepared synthetically using, for example, a commercial gene synthesis vendor such as Thermo Fisher (Carlsbad, CA). Advantageously the host cells are eukaryotic cells and the gene for each chain advantageously is synthesized with a sequence that encodes an N-terminal signal sequence that causes secretion of the translated protein from the host cell. The gene for each chain is inserted into a suitable expression vector, for example, pTT5 vector (Durocher et al., Nucleic Acids Res. 30:E9 (2002)), and the resulting expression constructs are then transfected into a culture of suitable host cells for transient expression. Methods of efficiently transfecting expression vectors into host cells are well known in the art using, for example, cationic lipids. See Feigner et al., Proc. Nat ’I Acad. Sci USA 84:7413 (1987). Other methods of delivering expression vectors into cells also are well known in the art. Methods of making host cells that provide stable expression of a desired protein by integrating expression constructs into the genome of suitable host cells also are well known, as are methods of stable expression using episomal vectors containing a mammalian origin of replication that act as extrachromosomal elements in the nucleus of the host cell.

The host cells advantageously are eukaryotic cells, for example, a single-celled eukaryote (e.g, a yeast or other fungus), a plant cell (e.g., a tobacco or tomato plant cell), an animal cell (e.g., a human cell, a monkey cell, a hamster cell, a rat cell, a mouse cell, or an insect cell) or ahybridoma. Examples of host cells include the COS-7 line of monkey kidney cells (ATCC CRL 1651) (see Gluzman etal., 1981, Cell 23:175), L cells, C127 cells, 3T3 cells (ATCC CCL 163), Chinese hamster ovary (CHO) cells or their derivatives such as Veggie CHO and related cell lines which grow in serum-free media (see Rasmussen et al., 1998, Cytotechnology 28:31) or CHO strain DX-B11, which is deficient in DHFR (see Urlaub et al., 1980, Proc. Natl. Acad. Sci. USA 77:4216-20), HeLa cells, BHK (ATCC CRL 10) cell lines, the CV1/EBNA cell line derived from the African green monkey kidney cell line CV1 (ATCC CCL 70) (see McMahan et al., 1991, EMBO J. 10:2821), human embryonic kidney cells such as 293, 293 EBNA or MSR 293, human epidermal A431 cells, human Colo205 cells, other transformed primate cell lines, normal diploid cells, cell strains derived from in vitro culture of primary tissue, primary explants, HL-60, U937, HaK or Jurkat cells. Advantageously, the host cell is a CHO cell, such as CHO-3E7.

Typically, a host cell is a cultured cell that can be transformed or transfected with a polypeptide-encoding nucleic acid, which can then be expressed in the host cell. The phrase "recombinant host cell" can be used to denote a host cell that has been transformed or transfected with a nucleic acid to be expressed. A host cell also can be a cell that comprises the nucleic acid but does not express it at a desired level unless a regulatory sequence is introduced into the host cell such that it becomes operably linked with the nucleic acid. The term “host cell” as used herein refers not only to the particular subject cell but to the progeny or potential progeny of such a cell. Because certain modifications may occur in succeeding generations due to, e.g., mutation or environmental influence, such progeny may not, in fact, be identical to the parent cell, but are still included within the scope of the term as used herein. Suitable cell culture media for growing eukaryotic host cells are well known in the art and are commercially available from, for example, Thermo Fisher (Grand Island, NY).

The host cells are cultured under suitable conditions to allow assembly of the protein chains in the endoplasmic reticulum of the host cell, followed by secretion of the bispecific binding proteins into the cell culture supernatant. The assembly of the correct structure of the binding protein (as opposed to, for example, non-specific pairing of the chains leading to formation of inactive proteins) can be improved by altering the relative ratios of the expression vectors used to transfect the host cells. This variation in the vector ratio also can be used to counteract, say, less efficient production of one of the chains compared to a different chain. One skilled in the art will recognize that methods of optimizing the vector ratio(s) are well known in the art.

After the host cells are cultured in an appropriate expression medium for a suitable length of time, the conditioned medium containing the bispecific binding protein is collected, and the bispecific binding protein is purified using methods that are well known in the art. For example, the binding protein can be purified using methods that may include ionexchange chromatography, size-exclusion chromatography, and affinity chromatography, such as protein A affinity chromatography. Methods of protein purification are described in, for example, Burgess and Deutscher (Eds) “Guide to Protein Purification, Volume 436 (Methods in Enzymology) 2nd Edition (2009). Purity of the protein can be confirmed using methods well-known in the art, such as RT-HPLC, SDS-PAGE and the like. The correct assembly of the multichain binding protein can be shown by, for example, non-denaturing gel electrophoresis, to show that the binding protein has the expected molecular weight. SDS-PAGE can be used to show that each of the expected protein chains is present and has the expected molecular weight. Western blotting using a suitable antiHuman IgG antibody can further show that the measured proteins are immunoglobulin chains.

Pharmaceutical Compositions and Methods of Administration

Methods of preparing and administering the bispecific binding molecules to a subject in need thereof are well known to or are readily determined by those skilled in the art. The route of administration of the binding molecule can be, for example, oral, parenteral, by inhalation or topical. The term parenteral as used herein includes, for example, intravenous, intraarterial, intraperitoneal, intramuscular, subcutaneous, rectal, or vaginal administration. However, in other methods compatible with the teachings herein, the binding molecules may be delivered directly to the site of the adverse cellular population thereby increasing the exposure of the diseased tissue to the therapeutic agent.

The binding molecules may be administered in a pharmaceutically effective amount for the treatment of diseases such as certain types of cancers. The pharmaceutical compositions can comprise pharmaceutically acceptable carriers, including, for example, water, ion exchangers, proteins, buffer substances, and salts. Preservatives and other additives can also be present. The carrier can be a solvent or dispersion medium. Suitable formulations for use in the therapeutic methods disclosed herein are described in Remington's Pharmaceutical Sciences (Mack Publishing Co.) 16th ed. (1980).

In any case, sterile injectable solutions can be prepared by incorporating the binding molecule(s) by itself or in combination with other active agents in an effective amount in an appropriate solvent followed by filtered sterilization. The preparations may also be packaged and sold in the form of a kit. Such articles of manufacture can have labels or package inserts indicating that the associated compositions are useful for treating a subject suffering from, or predisposed to a disease or disorder.

Parenteral formulations can be a single bolus dose, an infusion or a loading bolus dose followed with a maintenance dose. These compositions can be administered at specific fixed or variable intervals, for example, once a week or monthly, or on an "as needed" basis. The composition can be administered as a single dose, multiple doses or over an established period of time in an infusion. Dosage regimens also can be adjusted to provide the optimum desired response (for example, a therapeutic or prophylactic response).

The composition may be used for treatment of cell-mediated diseases such as certain types of cancers including for example, bone cancer, pancreatic cancer, cancer of the head and neck, cutaneous or intraocular melanoma, uterine cancer, cancer of the central nervous system (CNS), ovarian cancer, rectal cancer, cancer of the anal region, stomach cancer, colon cancer, breast cancer, melanoma, colorectal cancer, testicular cancer, Hodgkin's disease, cancer of the esophagus, cancer of the small intestine, cancer of the endocrine system, sarcomas of soft tissues, cancer of the urethra, cancer of the penis, prostate cancer, chronic or acute leukemia, solid tumors of childhood, lymphocytic lymphomas, cancer of the bladder, cancer of the kidney or ureter, adrenocortical carcinoma, AIDS-related cancers, Childhood Cerebellar Astrocytoma, Childhood Cerebral Astrocytoma, Basal Cell Carcinoma, extrahepatic bile duct cancer, osteosarcoma/malignant fibrous histiocytoma bone cancer, brain tumors, bronchial adenomas/carcinoids, carcinoid tumor, gastrointestinal carcinoid tumor, cerebellar astrocytoma, cerebral astrocytoma/malignant glioma, cervical cancer, childhood cancers, CMML, Chronic Lymphocytic Leukemia, Chronic Myelogenous Leukemia, Chronic Myeloproliferative Disorders, ependymoma, Ewing’s Family of Tumors, extracranial germ cell tumor, extragonadal germ cell tumor, extrahepatic bile duct cancer, gallbladder cancer, gastrointestinal carcinoid tumor, germ cell tumors, gestational trophoblastic tumor, glioma, hairy cell leukemia, hepatocellular cancer, hypopharyngeal cancer, hypothalamic and visual pathway glioma, islet cell carcinoma, Kaposi's Sarcoma, laryngeal cancer, leukemia, lip and oral cavity cancer, non-small cell lung cancer (including squamous cell lung carcinoma, large cell lung carcinoma, lung adenocarcinoma, pulmonary pleomorphic carcinoma, lung carcinoid tumor, salivary gland carcinoma), small cell lung cancer (including combined small-cell carcinoma), “not otherwise specified” (NOS) lung cancer, extrapulmonary small-cell carcinoma (including extrapulmonary small-cell carcinoma localized in the lymph nodes and small-cell carcinoma of the prostate), lymphoma, Waldenstrom's Macroglobulinemia, medulloblastoma, mesothelioma, metastatic squamous neck cancer with occult primary, multiple endocrine neoplasia syndrome, multiple myeloma/plasma cell neoplasm, mycosis fungoides, myelodysplastic syndromes, myelodysplastic/myeloproliferative diseases, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, oral cancer, oropharyngeal cancer, ovarian epithelial cancer, ovarian germ cell tumor, ovarian low malignant potential tumor, islet cell pancreatic cancer, parathyroid cancer, pheochromocytoma, pineoblastoma, pituitary tumor, pleuropulmonary blastoma, ureter transitional cell cancer, retinoblastoma, rhabdomyosarcoma, salivary gland cancer, Sezary Syndrome, skin cancer, non-melanoma skin cancer, Merkel Cell Skin Carcinoma, squamous cell carcinoma, testicular cancer, thymoma, gestational trophoblastic tumor, Wilms' Tumor, and cancers with microsatellite instability.

With respect to esophageal cancer, the cancer may be, for example, esophageal squamous-cell carcinomas (ESCC) or esophageal adenocarcinomas (EAC). With respect to pancreatic cancer, the cancer may be, for example, exocrine cancer, pancreatic adenocarcinoma, pancreatic ductal carcinoma, acinar cell carcinoma of the pancreas, cystadenocarcinoma, pancreatoblastoma, adenosquamous carcinomas, undifferentiated carcinomas, pancreatic mucinous cystic neoplasms, neuroendocrine, or pancreatic neuroendocrine tumors (PanNETs). With respect to hepatic cancer, the cancer may be, for example, hepatocellular carcinoma (HCC), hepatoblastoma, cholangiocarcinoma, cholangiocellular cystadenocarcinoma, angiosarcoma, or leiomyosarcoma. With respect to colorectal cancer the cancer may be adenocarcinoma, carcinoid tumors, gastrointestinal stromal tumors (GIST), lymphoma, sarcomas, adenosquamous carcinoma (Ad-SCC) or squamous carcinoma (SCC). With respect to breast cancer, the cancer may be In situ, ductal carcinoma in situ (DCIS), invasive, invasive ductal carcinoma (IDC), invasive lobular carcinoma (ILC), triple-negative breast cancer, inflammatory breast cancer, angiosarcoma, or Paget disease of the breast. With respect to ovarian cancer, the cancer may be epithelial tumors, benign epithelial ovarian tumors, borderline epithelial ovarian tumors, malignant epithelial ovarian tumors, germ cell tumors, teratoma, dysgerminoma, endodermal sinus tumor and choriocarcinoma, primary peritoneal carcinoma, fallopian tube cancer or ovarian stromal tumors. With respect to multiple myeloma & precancerous conditions, the cancer may be light chain myeloma, non-secretory myeloma, solitary plasmacytoma, extramedullary plasmacytoma, monoclonal gammopathy of undetermined significance (MGUS), smoldering multiple myeloma (SMM), Immunoglobulin D (IgD) myeloma or Immunoglobulin E (IgE) myeloma.

Therapeutically effective doses of the compositions for treating these diseases vary depending upon many different factors, including means of administration, target site, physiological state of the patient, whether the patient is human or an animal, other medications administered, and whether treatment is prophylactic or therapeutic. Usually, the patient is a human, but non-human mammals including transgenic mammals can also be treated. Treatment dosages can be titrated using routine methods known to those of skill in the art to optimize safety and efficacy.

The amount of at least one binding molecule to be administered can be readily determined by one of ordinary skill in the art without undue experimentation. Factors influencing the mode of administration and the respective amount of at least one binding molecule include, but are not limited to, the severity of the disease, the history of the disease, and the age, height, weight, health, and physical condition of the individual undergoing therapy. Similarly, the amount of binding molecule, to be administered will depend upon the mode of administration and whether the subject will undergo a single dose or multiple doses of this agent.

The binding molecule, may also be used in the manufacture of a medicament for treating a type of cancer, including, for example, the cancers listed above.

The subject treated with the compositions described herein may be treatment naive or may be pretreated with one or more other therapies (for example, at least one other anticancer therapy) prior to receiving the medicament comprising the binding molecule. It is not necessary that the subject was a responder to pretreatment with the prior therapy or therapies. Thus, the subject that receives the medicament comprising the binding molecule, could have responded, responded poorly, responded initially but subsequently failed to respond, or could have failed to respond to pretreatment with the prior therapy, or to one or more of the prior therapies where pretreatment comprised multiple therapies. Accordingly, the present disclosure provides methods to treat patients that are poor responders or non-responders to other therapies comprising administering a binding molecule as described herein. Also provided are methods to overcome or prevent resistance to cancer therapies or to prevent or delay relapse, comprising administering a binding molecule as disclosed herein, or a composition as described herein.

Even if a patient has been previously treated with an anti-cancer medicament, a person skilled in the art can determine whether a person showed no response or was refractory to that medicament. For example, a non-response to an anti-cancer medicament may be reflected in an increased suffering from cancer, such as an increased growth of a cancer/tumor and/or increase in the size of a tumor, in the formation of (or increase in) metastases or an increase in the number or size of metastases. A non-response may also be the development of a tumor or metastases, for example after resection of a tumor, in the shortening of time to disease progression, or in the increase in the size of (a) tumor(s) and/or (a) metastases, for example in neoadjuvant therapy. Based on these parameters or other parameters known in the art, a patient group can be identified that does not respond to treatment with anti-cancer medicaments and this group of patients may then be treated with the binding molecules described herein.

The binding molecules and compositions containing the binding molecules may also be used to treat patients that are, for example, poor-responders or non-responders to another therapy. The term "non-responder" as used herein can refer to an individual/patient/subject that is less likely to respond to a treatment using an anti-cancer medicament. "Less likely to respond" as used herein refers to a decreased likeliness that a pathological complete response will occur in a patient treated with an anti-cancer medicament. In some aspects, a patient can be initially a good responder, and resistance to treatment can develop during treatment with such an anti-cancer medicament, leading to poor or no-response to the treatment.

The term "good responder" as used herein refers to an individual whose tumor does not demonstrate growth, metastases, increase in number or size of metastases, etc. during or after treatment using an anti-cancer medicament, for example based on serial imaging studies, an individual that does not experience tumor growth, metastases, increase in number or size of metastases, etc. over a period of time (for example, about 1 year following initial diagnosis), and/or an individual that experiences a certain life span (for example, about 2 years or more following initial diagnosis).

The term "poor responder" as used herein refers to an individual whose tumor grows or metastasizes during or shortly thereafter standard therapy, for example using an anticancer medicament, or who experiences adverse clinical effects attributable to the tumor. The term “poor responder” also includes individuals who transitioned from “good responder” to ‘poor responder” during treatment with an anti-cancer medicament.

In cases where it is assessed that the subject is a "non-responder," a "poor-responder" or is "less likely to respond" (based, for example, on the presence of certain biomarkers in the cancer cells), the subject could be treated with the binding molecules disclosed herein.

Methods also are provided for the co-administration of a binding molecule as described herein and at least one other therapy. The binding molecule and the at least one other therapy can be co-administered together in a single composition or can be coadministered together at the same time or overlapping times in separate compositions. In some aspects, a binding molecule can be used as an adjuvant therapy.

The binding molecule may also be used in the manufacture of a medicament for treating a subject suffering from a cancer, where the binding molecule is administered before a subject has been treated with at least one other therapy. The binding molecules can also be used in methods of preventing or reducing the risk of cancer in a subject by administering to the subject an effective amount of a binding protein or composition containing the binding protein. The cancer may be lung cancer and the patient may be in remission from a previously diagnosed and/or previously treated cancer. The patient may be considered to be at risk of cancer due to environmental exposure, tobacco use or exposure, genetic mutation, or a family history of cancer.

Examples:

In the examples below the binding domains are based on heavy and light chain variable regions from human antibodies that have received regulatory approval for use in humans. “O,” “E” and “K” indicate domains that bind IL-ip, while “M” and “B” indicate binding domains that bind PD-1. The amino acid sequences of the chains of the binding proteins are shown in Figure 5

Example 1: Expression of bispecific antibodies

1. Plasmid Preparation

Target DNA sequences encoding the binding proteins were synthesized and subcloned into the pTT5 vector (Durocher et al., Nucleic Acids Res. 30:E9 (2002) for expression in CHO-3E7 cells. The amino acid sequences of the coding sequences are shown in Figure 5 and Figure 17.

2. Cell Culture and Transient Transfection

CHO-3E7 cells were grown in serum-free FreeStyle™ CHO Expression Medium (Life Technologies, Carlsbad, CA, USA). The cells were maintained in Erlenmeyer Flasks (Coming Inc., Acton, MA) at 37°C with 5% CO2 on an orbital shaker (VWR Scientific, Chester, PA). One day before transfection, the cells were seeded in Coming Erlenmeyer Flasks. On the day of transfection, DNA and transfection reagent were mixed and then added into the cells culture, during which the recombinant plasmids encoding target antibody were transiently transfected into CHO-3E7 cells. The cell culture supernatant collected on day 6 was used for purification. Table 2 lists the heavy chain, light chain and Fd chain combination of each antibody and each of the plasmid ratios that were screened. Table 2. Summary of heavy chain (HC), light chain (LC) and Fd chain (Fd) combinations and plasmid ratios for each antibody.

Table 2 (continued)

Key to table for ITG101-ITI215:

DAB: diabody (see Figures IB and 2B) having a short (10 amino acids in the examples in the table) linker between VH and VL of the 5 terminal scFv moiety.

DSB: Disulfide bond. If not specifically designated, VH 44 and VL 100 are mutated to Cys.

L(x): Linker of length (x) amino acids between: scFv and Fab (BiS2); Fc and scFv (BiS3); DAB and Fab (DAB). If the linker length is not specified, the linker is 10 amino acids long for the BiS2, BiS3, and DAB formats

N434A: Point mutant in the Fc region that extends serum half-life through enhanced pH6 FcRn interaction

G1CH1 : a chimeric heavy chain with CHI from IgGl and Fc from G4 (see ITI102)

VH15VL/VL15VH: the linker between VH and VL of the scFv moiety (when present) is 15 amino acids. If not stipulated, the linker is

20 AA

3. Purification and Results

Cell culture supernatant was centrifuged followed by filtration. The filtered supernatant was loaded onto Monofinity A Resin Prepacked Column 1 ml (GenScript, Cat.No.L00433-l 1) at 1.0 ml/min. After a washing step, the antibodies were eluted and then buffer exchanged to PBS, pH 7.2. The antibodies were further purified by standard techniques. For example, some samples were further purified by size exclusion chromatography (SEC) using a Superdex 200 Increase 10/300 GL column or by ceramic hydroxyapatite chromatography using a CHT™ ceramic hydroxyapatite column (Bio-Rad). For other proteins, Protein A affinity chromatography was used for initial purification, followed by SEC if necessary.

Each of the expressed proteins was analyzed by SDS-PAGE under reducing and nonreducing conditions and Western blot analysis (using Goat Anti-Human IgG-HRP (GenScript, Cat. N0.AOOI66), which showed that proteins and constituent chains had the desired molecular weights. Purity of the proteins was assessed using size-exclusion HPLC.

Surprisingly, when an antibody with a VL domain of SEQ ID 39 and a VH domain of SEQ ID 40 was converted into an scFv format, it was found that stabilization of the scFv by a disulfide bond, formed by mutating position 44 of VH and 100 of VL to Cys (44/100DSB), improved yield monomer percentage as measured by SEC post-protein A. Further, the VL- VH orientation appears to be preferred compared to the VH-VL orientation. For example, ITH103 (VL-VH, no disulfide bond) expressed 11.8 mg/L and 51.9% monomer. ITA102 (VH-VL, no disulfide bond) expressed at 0.9 mg/L and 53% monomer. ITE101 (VH-VL, 44/100DSB) improved yield to 1.7 mg/L and monomer improved slightly to 57%. ITE102 (VL-VH, 44/100DSB) exhibited a yield of 7.5 mg/L and 86% monomer. This represents a significant improvement over its parent ITH103 in correct assembly. In addition to these variants, ITG108 (VL-VH, 101/46DSB) and ITG109 (VH-VL, 101/46DSB) were made. Again, the VL/VH orientation was favored for yield (7.2 vs. 2.0 mg/L) and monomer percentage (71% vs. 52%). However in this case, ITG108 showed relatively low response in the sandwich FACS binding assay indicating a loss of activity, most likely in binding to PD1 (Fig. 13 A). While insertions of disulfide bonds have been shown to improve stability, they have not been shown to improve monomer percentage or protein expression (Weatherill, et al., Protein Eng. Des. Sei. 25:321-329 (2012)). Example 2

Bispecific antibodies described in Example 1 that had sufficient expression and purity were analyzed for binding to PD-1 on the surface of cells using flow cytometry (Figures 8, 9, and 14). Cell lines used were CHO-K1 cells (negative control) and a CHO-K1/PD-1 expressing line. Antibodies used were an irrelevant human IgG (negative control), and commercially available anti-PD-1 (positive controls). The secondary antibody used was a goat anti Human IgG(H+L) iFluor 647 (1 pg/ml) (data not shown). In order to demonstrate concurrent binding of the bispecific molecules to PD1 and IL-ip, a sandwich FACS assay was performed. The binding curves are shown in Figure 7, 10, and 13. CHO cells expressing PD1 were bound by bispecific antibodies at the indicated concentrations. Unbound antibody was washed away and detection was performed using biotinylated IL-ip (2 pg/ml) followed by SA-iFluor 647 (1 pg/ml). Fluorescence intensity is indicative of both BiSAb immobilization on PD1+ cells and ability to concurrently bind IL-ip. Various combinations of anti-IL-ip VH/VL binding domains with two different pairs of anti-PD-1 VH/VL binding domains were formatted as BiSl, BiS2, BiS2DAB, BiS3, FIT-IG or FAT-IG molecules. Of the constructs tested, the four binding molecules having anti-PD-1 VH/VL binding regions containing the CDR regions defined by the sequences of SEQ ID Nos: 1-6 all achieved a higher mean fluorescence intensity compared to the five binding molecules having anti-PD-1 VH/VL binding regions containing the CDR regions defined by the sequences of SEQ ID Nos: 6-12.

Figures 8-16 show additional binding data and functional activity for additional binding proteins. Figure 8 shows binding curves for binding of bispecific antibodies to cell membrane-bound PD-1 and soluble IL-i simultaneously in a multiple-dose sandwich assay, as detected by flow cytometry. Variable binding affinities of the ITC, ITD and ITE series of bispecific antibodies are shown. All the binding affinities are substantially higher than control human IgG.

Figure 9 and 14 shows binding curves for binding to cell membrane-bound PD-1 in a multiple-dose sandwich assay, as detected by flow cytometry. Variable binding affinities of the ITB and ITF series of bispecific antibodies are shown. All the binding affinities are substantially higher than control human IgG.

Figure 10 and 13 shows binding curves for binding to cell membrane-bound PD-1 and soluble IL-i simultaneously in a multiple-dose sandwich assay, as detected by flow cytometry. Variable binding affinities of bispecific antibodies are shown. All the binding affinities are substantially higher than control human IgG.

Figure 11 and 15 shows that bispecific antibodies block PD-1 activity in a PD-l/PD- L1 reporter assay. Variable blockade activities are shown for the bispecific antibodies. All the blockade activities are substantially higher than control human IgG.

Figure 12 and 16 shows that bispecific antibodies block IL-ip activity in an IL-ip functional assay. Variable blockade activities are demonstrated of the bispecific antibodies. All the blockade activities are substantially higher than control human IgG, except ITA201 and ITG202. ITE102, which contains an scFv containing a VL domain of SEQ ID 39 and a VH domain of SEQ ID 40, was also surprisingly found to be almost 3-fold more potent than ITH103 in blocking PD-dependent inhibition of reporter activity (Fig 16C). Without wishing to be bound by theory, this demonstrates that the presence of a stabilizing disulfide apparently improves not only the monomer percentage, but also improves potency, even when the molecules have both been purified to >95% monomer.

Claims

Claims What is claimed is:

1. A binding protein comprising a first human immunoglobulin binding domain and a second human immunoglobulin binding domain, wherein said first human immunoglobulin binding domain specifically binds and inhibits activation of PD-1 or PD-L1, wherein said second human immunoglobulin binding domain specifically binds and inhibits the activity of IL-P, wherein said binding protein has a domain structure as shown in Figure 1A, Figure IB, Figure 1C, Figure 2A, Figure 2B, Figure 3 or Figure 4, and wherein said binding protein comprises:

(a) CHI domains of the IgGl isotype and Fc domains of the IgG4 isotype and/or

(b) a disulfide bond between the VH and VL domains of the scFV when present and/or

(c) a mutant Fc region that improves antibody half-life, wherein said Fc region comprises: i. an N434A mutation ii. a YTE (M252Y/S254T/T256E) mutation, or iii. an M428L/N434S mutation.

2. The binding protein according to claim 1 wherein the binding protein has the domain structure as shown in Figure 1A, Figure IB, Figure 1C, Figure 2 A, or Figure 2B, and wherein said binding protein comprises a disulfide bond between the VH and VL domains of the scFV.

3. The binding protein according to claim 2 wherein the disulfide bond is between:

(a) the VH44 and VL100 residues;

42 (b) the VH105 and VL43 residues;

(c) the VH101 and VL46 residues;

(d) the VH45 and VL98 residues; or

(e) the VH104 and VL43 residues of the scFV moiety.

4. The binding protein of claim 3, wherein the scFv moiety comprises a VH and VL domain pairing selected from the group consisting of:

VH44/VL100

VH44:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKCLEWVAVIWYDG

SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWGQGTLVT vss

VL100:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGI

PARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGCGTKVEIK

VH105/VL43

VH105:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKGLEWVAVIWYDG

SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWGCGTLVT

VSS

VL43:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQCPRLLIYDASNRATGIP

ARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGQGTKVEIK

VH101/VL46

43 VH101:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKGLEWVAVIWYDG

SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDCYWGQGTLVT vss

VL46:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRCLIYDASNRATGI

PARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGQGTKVEIK

VH45/VL98

VH45:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKGCEWVAVIWYDG

SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWGQGTLVT

VSS

VL98:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGI

PARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTCGQGTKVEIK and

VH104/VL43

VH104:

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKGLEWVAVIWYDG

SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWCQGTLVT

VSS

VL43:

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQCPRLLIYDASNRATGIP

ARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGQGTKVEIK

44

5. The binding protein of any preceding claim comprising CHI domains of the

IgGl isotype and Fc domains of the IgG4 isotype.

6. The binding protein of any preceding claim comprising an N434A, YTE (M252Y/S254T/T256E), or M428L/N434S mutation in the Fc region.

7. The binding protein of claim 4 selected from the group consisting of ITG101- ITG110, ITG102C, ITH101, ITH102, ITH501, ITH602, ITI102, ITI208, and ITI209.

8. The binding protein of claim 3 selected from the group consisting of ITG101- ITG110, ITG201-ITG204, ITG102C, ITH101, ITH102, ITH201-ITH204, ITH501, ITH602, ITI101-ITI103, ITI202-ITI209 and ITI212-ITI215.

9. The binding protein according to any of claims 1-6, wherein said binding protein has a domain structure as shown in Figure IB or Figure 2B.

10. The binding protein according to claim 9 wherein the linker between the VH and VL domains of the scFv moiety consists of 6-12 amino acids.

11. The binding protein of claim 10 selected from the group consisting of ITH601 and ITH602.

12. The binding protein of claim 1 selected from the group consisting of ITG301, ITG302, ITG401-ITG406, ITH103, ITH601, ITI201, ITI210, and ITI211.

13. A method for the preparation of a binding protein according to any one of claims 1 to 12 comprising the steps of a) transforming a host cell with vectors comprising nucleic acid molecules encoding said binding domains; b) culturing the host cell under conditions that allow synthesis of said binding protein; and c) recovering said binding protein from said culture.

14. A host cell comprising vectors comprising nucleic acid molecules encoding the binding domains according to any one of claims 1 to 12.

15. A pharmaceutical composition comprising a binding protein according to any one of claims 1 to 12 and a pharmaceutically acceptable excipient.

16. A method of treating cancer in a subject comprising administering a binding protein according to any of claims 1-12 or a composition according to claim 15 to a subject in need thereof.

17. The method of claim 16, further comprising administering to said subject an antitumor agent.

18. The method of claim 16 or claim 17 wherein said cancer is lung cancer.

19. The method of claim 18, wherein said lung cancer is small cell lung cancer.

20. The method of claim 19, wherein said small cell lung cancer is combined small-cell lung carcinoma.

21. The method of claim 18, wherein said lung cancer is non-small cell lung cancer.

22. The method of claim 21, wherein said non-small cell lung cancer is selected from the group consisting of squamous cell lung carcinoma, large cell lung carcinoma, lung adenocarcinoma, pulmonary pleomorphic carcinoma, lung carcinoid tumor, salivary gland carcinoma, or carcinoma NOS (not otherwise specified).

23. The method of claim 16 or claim 17 wherein said cancer is combined smallcell lung carcinoma, extrapulmonary small-cell carcinoma, extrapulmonary small-cell carcinoma localized in the lymph nodes or small-cell carcinoma of the prostate.

24. The method of claim 16 or claim 17 wherein said cancer is a cancer with microsatellite instability.

25. The method of claim 16 or claim 17 wherein said subject has previously been treated with cancer immune therapy or has been found to be resistant to said therapy.

26. The method of claim 16 or claim 17 wherein said subject has previously been treated with cancer immune therapy or has been found to be refractory to cancer immune therapy.

27. The method of claim 25 or 26 wherein said cancer immune therapy is treatment with at least one immune checkpoint inhibitor.

28. The method of claim 16 or claim 17 further comprising administering to said subject an additional anti -tumor therapy.

29. The method of claim 28, wherein said anti-tumor therapy is chemotherapy, immune therapy, treatment with biologies or small molecules, vaccination, or a cell therapy.

47

30. A method of preventing or reducing the risk of cancer in a subject at risk thereof, comprising administering to said subject an effective amount of a binding protein according to any of claims 1-12 or a composition according to claim 15.

31. The method of claim 30, wherein said cancer is lung cancer.

32. The method of claim 30 wherein said subject previously was diagnosed with cancer and is in remission.

33. The method of claim 32 wherein said subject was previously treated for cancer.

34. The method of claim 30, wherein said subject is considered to be at risk of cancer due to environmental exposure, tobacco use or exposure, genetic mutation, or a family history of cancer.

35. A nucleic acid molecule encoding a binding domain of a binding protein according to any of claims 1-12.

36. The method of claim 16 or claim 17, wherein said cancer is selected from the group consisting of esophageal, pancreatic, hepatic, colorectal, breast, and ovarian cancer, or multiple myeloma or precancerous conditions.

37. A binding protein, comprising a scFV binding domain, wherein said scFv binding domain comprises: a VH domain comprising the sequence

QVQLVESGGGVVQPGRSLRLDCKASGITFSNSGMHWVRQAPGKCLEWVAVIWYDG

48 SKRYYADSVKGRFTISRDNSKNTLFLQMNSLRAEDTAVYYCATNDDYWGQGTLVT vss, coupled via an amino acid linker to a VL domain comprising the sequence

EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGI

PARFSGSGSGTDFTLTISSLEPEDFAVYYCQQSSNWPRTFGCGTKVEIK, wherein the VL domain is N-terminal to the VH, and wherein the linker consists of 14 to 25 amino acids.

38. A binding protein according to claim 37, wherein said binding protein has the domain structure shown in Figure 1A, Figure IB, Figure 1C, Figure 2A, or Figure 2B.

49