DUAL PLATELET-DERIVED GROWTH FACTOR (PDGF) / VASCULAR ENDOTHELIAL GROWTH FACTOR (VEGF) ANTAGONISTS

Abstract

The invention provides a dual VEGF / PDGF antagonist comprising a VEGF antagonist linked to a PDGF antagonist. The VEGF antagonist is an antibody to a VEGF or VEGFR or is an extracellular trap segment of a VEGFR (i.e., a segment of the extracellular region of one or more VEGFR receptors that inhibits the binding of at least one VEGFR to at least one VEGF). The PDGF antagonist is an antibody to a PDGF or PDGFR or is an extracellular trap segment of a PDGFR (i.e., a segment of the extracellular region of one or more PDGFRs that inhibits the binding of at least one PDGFR to at least one PDGF). The dual antagonist is preferably conjugated to a half-life extension portion, such as a HEMA-PC polymer. The dual antagonist is particularly useful for treating wet age-related macular degeneration.

Description

DUAL PLATELET-DERIVED GROWTH FACTOR (PDGF) / VASCULAR ENDOTHELIAL GROWTH FACTOR (VEGF) ANTAGONISTS Field and Background of the Invention

[0001] Angiogenesis (the formation of blood vessels) occurs throughout the development of an organism. Actually, the first organ in an embryo is a blood vessel. Angiogenesis is also crucial for wound healing, the restoration of blood flow to damaged tissue. However, inappropriate or unregulated angiogenesis contributes to or causes many diseases including cancer, psoriasis, arthritis, and blindness. Carmeliet P. 2003. Angiogenesis in health and disease. Nature Med 9(6):653-660.

[0002] Age-related macular degeneration (AMD) is a leading cause of vision loss and blindness in the elderly. Approximately ten million Americans are afflicted with AMD. The prevalence of AMD in the population increases steadily with age; by age 40 only about 2% of the population is affected by AMD but by age 80 about 25% is. Friedman, D.S. et al. 2004. Arch. Ophthalmol. 122:564-572. In general, there are two types of AMD: dry and wet.

[0003] Dry AMD is the most common form of the disease. In dry AMD, there is depletion of the retinal pigment epithelial cell layer in the macula. Dry AMD is chronic and generally causes some vision loss. In severe cases of dry AMD, patients can develop near total blindness. Wet AMD develops in as little as 10-15% of patients with dry AMD. Wet AMD is characterized by angiogenesis, especially choroidal neovascularization (CNV). CNV is characterized by the presence of new immature blood vessels that grow into the outer retina from the choroid. These immature blood vessels leak fluid below and into the retina, causing blindness and vision loss. Blindness from wet AMD is typically acute.

[0004] Angiogenesis also plays a crucial role in cancer and tumor formation and maintenance. Recruitment of new blood vessels is an essential component of the metastatic pathway. For many tumors, vascular density can provide a prognostic indicator of metastatic potential: Highly vascular tumors have a higher incidence of metastasis than less vascular tumors.

[0005] Angiogenesis is the result of a complex interaction between growth factors, vascular endothelial cells, extracellular matrix molecules, chemokines, and cell signaling molecules. Factors identified as mediators of angiogenesis include: basic and acidic fibroblast growth factor, transforming growth factors a and β, platelet-derived growth factor (PDGF), angiogenin, platelet-derived endothelial cell growth factor, IL- 8, and vascular endothelial growth factor (VEGF). The role of VEGF in angiogenesis has been extensively reported.

[0006] VEGF signaling has been shown to be a crucial rate-limiting step in physiological angiogenesis. VEGF also plays a central role in pathological angiogenesis (eg, tumor growth). Ferrara N and Davis-Smyth T. 1997. The biology of vascular endothelial growth factor. endocr. noi i ηη / ι znz / R / v Rev 18:4-25. VEGF is also known to induce vascular leakage. Bates DO and Curry FE. 1997. Vascular endothelial growth factor increases microvascular permeability via the Ca(2+)-dependent pathway. Am J Physiol. 273: H687-H694; Roberts WG and Palade GE. 1995. Increased microvascular permeability and endothelial fenestration induced by vascular endothelial growth factor. J Cell Sci. 108:2369-2379.

[0007] Anti-VEGF therapeutics have been used successfully to treat wet AMD and cancer. Genentech's anti-VEGF monoclonal antibody bevacimumab (Avastin™) received FDA approval in 2004 for the treatment of cancer. Anti-VEGF agents have been approved for the treatment of wet AMD. In 2004, the FDA approved Eyetech Pfizer Macugen™. Genentech's LucentisMR was approved in 2006 for wet AMD. Off-label Bevacizumab is also used for the treatment of wet AMD. In 2011, Eylea™ from Regeneran was approved for the treatment of wet AMD.

[0008] Despite the success of anti-VEGF therapeutics, none of them cause regression in pathological neovascular (NV) tissue. Therefore, NV tissue remains despite continued anti-VEGF treatment and may prevent significant vision gain for treated patients. NV tissue consists of endothelial cells, inflammatory cell pericytes (ie, occasional macrophages). The presence of pericytes in the capillaries not only leads to stabilization and support of NV but also promotes endothelial cell survival through chemical signaling and physical interactions including pericyte production of VEGF. This endothelial survival signaling by integrated pericytes is critical and may explain the resistance of NV tissue to VEGF withdrawal, ie, the appearance of NV regression to anti-VEGF monotherapy treatment. Furthermore, over time, pathological NV tissue can lead to fibrosis and scarring.

[0009] Subretinal scarring develops in almost half of the eyes treated within two years of anti-VEGF therapy. Daniel E, Toth CA, Grunwald JE. 2014. Risk of scar in the comparison of age-related macular degeneration in clinical settings. Retina 32:1480-1485. The formation of subretinal fibrosis can cause permanent dysfunction of the macular system; it causes destruction of photoreceptors, retinal pigment epithelium, and colloidal vessels. Ishikawa K, Ram K, Hinton DR. 2015. Molecular mechanisms of subretinal fibrosis in age-related macular degeneration. Eye Res. xxx:1-7. While anti-VEGF therapy generally stabilizes or improves visual acuity, scarring has been identified as one of the causes of loss of visual acuity after treatment. Cohen SY, Oubraham H, Uzzan J, et al. 2012. Causes of unsuccessful ranibizumab treatment in exudative age-related macular degeneration in clinical settings. Retina 32:1480-1485.

[00010] PDGF has been reported to play a role in pericyte recruitment, maturation, and resistance to anti-VEGF-mediated regression. Animal models of corneal and conoidal neovascularization have been reported to show that administration of agents that block the PDGF-B / PDGFR-β interaction leads to detachment of pericytes from pathological neovasculature. Jo N, Mailhos C, Ju M, et al. 2006. Inhibition of Platelet-Derived Growth Factor B Signaling Enhances the Efficacy of Anti-Vascular Endothelial Growth Factor Therapy in Multiple Models of Ocular Neovascularization. American J Path. 168(6):2036-2053.

[00011] To target both routes, clinical trials are currently underway in which noi i ηη / ι znz / R / v patients receive two drugs: LucentisMR (an anti-VEGF Fab) and FovistaMRa PEGylated aptamer directed against PDGF from Ophthotech. Fovista is directed against only a single PDGF ligand: PDGF-BB. However, there are many other PDGF ligands: PDGF-AA, PDGF-CC and PDGF-DD, PDGF-DD. For example, PDGF-DD has been shown to play a crucial role in ocular angiogenesis. Kumar A, Hou X, Chunsik L, et al. 2010. Platelet-derived Growth Factor-DD Targeting Arrests Pathological Angiogenesis by Modulating Glycogen Synthase Kinase-3p Phosphorylation. J Biol Chem 285(20):15500-15510. Even Fovista does not interact with PDGF-DD. There is a need in the art for broader therapies based on anti-PDGF.

[00012] In addition, aptamer-based therapeutics generally have poor pharmacokinetic properties as aptamers undergo renal filtration and serum digestion. While these problems can be somewhat overcome with PEGylation, PEGylation tends to reduce target binding. Aptamers will typically bind with much lower affinity to targets than their antibody counterparts. PEGylation will tend to reduce binding even further. Thus, there is a need in the art for non-aptamer based anti-PDGF therapeutics.

[00013] Current clinical plans for Fovista double the number of injections patients must receive for treatment relative to currently approved anti-VEGF therapies. Fovista is formulated separately from the anti-VEGF people so twelve injections must be given instead of one. Furthermore, the injections cannot be given at the same time due to the buildup in intraocular pressure caused by a single injection.

[00014] From the point of view of both the patients and the treating physicians, intraitrial injections are not trivial. Many patients experience pain and discomfort from the injection and patient compliance is a serious issue. Common side effects of intravitrial injections include conjunctival hemorrhage, eye pain, vitreous floaters, increased intraocular pressure, and intraocular inflammation. Intravitrial injections are associated with relatively rare serious adverse events, including endophthalmitis, retinal detachment, and traumatic cataracts.

[00015] Thus, there is a need in the art for therapies that do not increase the number of intravitreal injections that patients must undergo. In addition, current anti-VEGF therapies often require once a month injections. There is also a need for therapies that are less frequently than once a month. Brief Description of the Invention

[0016] The invention provides a dual VEGF / PDGF antagonist comprising a VEGF antagonist linked to a PDGF antagonist, wherein the VEGF antagonist (a) is an antibody to VEGF or VEGFR or (b) is a segment of VEGFR extracellular trap and the PDGF antagonist (a) is an antibody to PDGF or PDGFR or (b) is a segment of PDGFR extracellular trap, with the proviso that the VEGF and PDGF antagonists are not both antibodies. Optionally, the VEGF antagonist is an antibody comprising a heavy chain and a light chain and the PDGF antagonist is the extracellular trap segment of PDGFR, and the heavy chain noi i ηη / ι znz / R / v of the antibody is fused via a linker to the C-terminus of the PDGFR extracellular trap segment, and the light chain becomes complexed with the heavy chain. Optionally, the antibody is a Fab fragment. Optionally, the antibody is an intact antibody. Optionally, the PDGF antagonist is an extracellular trap segment of a PDGFR-α or PDGFR-β receptor and the VEGF antagonist is an antibody to VEGF. Optionally, the PDGFR extracellular trap segment comprises one or more of the PDGFR-β domains D1-D5. Optionally, the PDGFR extracellular trap segment comprises the D1-D3 domains of PDGFR-β. Optionally, the PDGFR extracellular trap segment comprises amino acids 33 to 314 of SEQ ID No.: 11. Optionally, the VEGF antagonist comprises an anti-VEGF antibody. Optionally, the anti-VEGF antibody is an anti-VEGF-A antibody. Optionally, the PDGFR extracellular trap segment is located C-terminus of the heavy or light chain. Optionally, the PDGFR extracellular trap segment is located N-terminus of the heavy or light chain.

[00017] Optionally, the dual VEGF / PDGF antagonist further comprising a linker that is located between the PDGFR trap and the anti-VEGF antibody heavy chain. Optionally, the linker is GGGGSGGGGS, GG, or GGGGSGGGGSGGGGSGGGGSG.

[0018] Optionally, the anti-VEGF antibody heavy chain comprises: CDRh1: GYDFTHYGMN, CDRh2: WINTYTGEPTYAADFKR, and CDRH3: YPYYYGTSHWYFD. Optionally, the anti-VEGF light chain comprises CDRL1: SASQDISNYLN, CDRL2: FTSSLHS, and CDRL3: QQYSTVPWT.

[00019] Optionally, the anti-VEGF heavy chain isotype is lgG1 comprising a CHi, hinge, ΟΗ2 and CH3 domain and the light chain isotype is kappa. Optionally, the lgG1 constant domain has the sequence set forth in SEQ ID No.: 17 and the light chain constant region has the sequence set forth in SEQ ID No.: 18.

[00020] Optionally, the lgG1 constant domain has one or more mutations to reduce effector function. Optionally, the mutations are at one or more of the following amino acid positions (EU numbering): E233, L234, L235, G236, G237, A327, A330, and P331. Optionally, the mutations are selected from the group consisting of E233P, L234V, L234A, L235A, G237A, A327G, A330S, and P331S. Optionally, the mutations are L234A, L235A and G237A.

[00021] Optionally, the dual VEGF / PDGF antagonist comprises a heavy chain further comprising a cysteine ​​residue added by recombinant DNA technology. Optionally, the cysteine ​​residue is selected from the group consisting of (EU numbering) Q347C and L443C.

[00022] Optionally, the dual VEGF / PDGF antagonist has a heavy chain comprising the amino acid sequence of SEQ ID No.: 9 and the light chain has an amino acid sequence of SEQ ID No.: 10.

[00023] Optionally, the dual VEGF / PDGF antagonist comprises a PDGFR extracellular trap segment comprising one or more of the PDGFR-β domains D1-D5. Optionally, the PDGFR extracellular trap segment comprises the PDGFR-β domains D1-D3. Optionally, the extracellular PDGFR trap segment noi 1 ηη / ι znz / R / v comprises amino acids 33 to 314 of SEQ ID No.: 11.

[00024] Optionally, the dual VEGF / PDGF antagonist comprises a VEGF antagonist, which is an anti-VEGF antibody. Optionally, the antibody is an anti-VEGF-A Fab fragment. Optionally, the PDGFR extracellular trap segment is located C-terminally of the Fab heavy or light chain. Optionally, the PDGFR extracellular trap segment is located N-terminally of the Fab heavy or light chain.

[00025] Optionally, the dual VEGF / PDGF antagonist comprises a heavy chain comprising an anti-VEGF-A Fab fragment heavy chain and a light chain comprising an anti-VEGF-A light chain. Optionally, the dual antagonist further comprises a linker that is located between the PDGFR trap and the anti-VEGF Fab fragment heavy chain. Optionally, the linker is selected from the group consisting of GGGGSGGGGS, GG, and GGGGSGGGGSGGGGSGGGGSG. Optionally, the heavy chain of the anti-VEGF Fab fragment comprises CDRhI: GYDFTHYGMN, CDRh2: WINTYTGEPTYAADFK.R, and CDRH3: YPYYYGTSHWYFDV. Optionally, the anti-VEGF light chain comprises CDR11: SASQDISNYLN, CDR12: FTSSLHS, and CDR13: QQYSTVPWT. Optionally, the anti-VEGF heavy chain isotype is lgG1 comprising a CHi domain and the light chain isotype is kappa.

[00026] Any of the dual VEGF / PDGF antagonists may further comprise a half-life extension moiety. Optionally, the half-life extension portion comprises a polymer, which is PEG or a zwitterionic polymer. Optionally, the zwitterionic polymer comprises a monomer comprising phosphorylcholine. Optionally, the monomer comprises 2-(acryloyloxyethyl)-2'-(trimethylammoniomethyl)phosphate. Optionally, the monomer comprises 2-(methacryloxethyl)-2'-(trmethyllamoniomethyl) phosphate (HEMA-PC). Optionally, the polymer has 3 or more arms. Optionally, the polymer has 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12 arms. Optionally, the polymer has 3, 6 or 9 arms. Optionally, the polymer has 9 arms. Optionally, the polymer portion of the conjugate has a peak molecular weight of between 300,000 and 1,750,000 Da. Optionally, the polymer portion of the conjugate has a peak molecular weight between 500,000 and 1,000,000 Da. Optionally, the polymer portion of the conjugate has a peak molecular weight between 600,000 to 800,000 Da. Optionally, the dual VEGF / PDGF antagonist is covalently attached to the polymer. Optionally, the polymer is covalently bonded to at least one amino group, one hydroxyl group, one sulfhydryl group, and one carboxyl group. Optionally, the sulfhydryl group is from a naturally occurring cysteine ​​residue. Optionally, the sulfhydryl group is from a cysteine ​​residue added by recombinant DNA technology. Optionally, the polymer is covalently attached to the cysteine ​​residue at position 741 of SEQ ID No.: 9.

[00027] Optionally, the VEGF antagonist comprises a VEGFR extracellular trap segment comprising one or more extracellular segments of VEGFR-1, VEGFR-2 and VEGFR-3 and the PDGF antagonist is an anti-PDGF antibody. Optionally, the VEGFR extracellular segment comprises one or more D1-D7 domains. Optionally, the extracellular segment comprises VEGFR-1 D2 and VEGFR-2 D3. Optionally, D2 is N-terminal to D3 and further comprises a linker between the domains. Optionally, the PDGF antagonist is an intact antibody. Optionally, the PDGF antagonist is a Fab fragment. Optionally, the antinoi i ηη / ι znz / R / v antibody PDGFR is humanized 2A1E2, humanized HuM4Ts.22,1B3, humanized 2C5, anti-PDGF-BB, anti-PDGF-DD, antiPDGF-BB, or anti-PDGF-AB. Optionally, the heavy chain is lgG1 and the light chain is kappa. Optionally, the heavy chain sequence has a cysteine ​​added by recombinant DNA technology, the cysteine ​​selected from the groups consisting of Q347C or L443C. Optionally, the dual VEGF / PDGF antagonist further comprises a cysteine-conjugated half-life extension moiety. Optionally, the VEGF / PDGF dual antagonist protein has a half-life extension moiety comprising a zwitterionic polymer, the polymer comprising one or more monomer units and wherein at least one monomer unit comprises a zwitterionic group, such as phosphorylcholine . Optionally, the monomer comprises 2(acr¡lo¡lox¡et¡l)-2'-(trimethyllamon¡ometh¡l)phosphate. Optionally, the monomer comprises 2-(methacryloxyethyl)-2'(trimethylammoniomethyl)phosphate (HEMA-PC). Optionally, the polymer has 3 or more arms. Optionally, the polymer has 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12 arms. Optionally, the polymer has 3, 6 or 9 arms. Optionally, the polymer has 9 arms. Optionally, the polymer portion of the conjugate has a peak molecular weight between 300,000 and 1,750,000 Da. Optionally, the polymer portion of the conjugate has a peak molecular weight between 500,000 and 1,000,000 Da. Optionally, the polymer portion of the conjugate has a peak molecular weight between 600,000 to 800,000 Da.

[00028] In some dual VEGF / PDGF antagonists, the PDGF antagonist comprises a PDGF extracellular trap segment comprising one or more extracellular segments of a PDGFR selected from the group consisting of PDGFR-α and PDGFR-β and the antagonist VEGF is a VEGF extracellular trap segment comprising one or more extracellular segments of a VEGF selected from the group consisting of VEGFR-1, VEGFR-2 and VEGFR-3. Optionally, the VEGFR extracellular trap segment comprises one or more of the D1-D7 domains. Optionally, the extracellular trap segment comprises VEGFR-1 D2 and VEGFR-2 D3. Optionally, D2 is N-terminal to D3 and further comprises a linker between the domains. Optionally, the PDGFR trap comprises one or more of the PDGFR-β domains D1-D5. Optionally, the PDGFR trap comprises the PDGFR-β domains D1-D3. Optionally, the PDGFR trap comprises amino acids 33 to 314 of SEQ ID No.: 11. Optionally, the dual VEGF / PDGF antagonist further comprises a linker sequence between the VEGF antagonist and the PDGF antagonist. Optionally, the dual VEGF / PDGF antagonist further comprises a half-life extension portion. Optionally, the half-life extension moiety comprises a polymer selected from the group consisting of PEG and a zwitterionic polymer. Optionally, the half-life extension portion comprises a zwitterionic polymer. Optionally, the zwitterionic polymer comprises a monomer comprising phosphorylcholine. Optionally, the monomer comprises 2-(acr¡lo¡lox¡et¡l)-2'-(trimethyllamon¡ometh¡l)phosphate. Optionally, the monomer comprises 2(methacryloyloxyethyl)-2'-(trimethylammoniomethyl)phosphate (HEMA-PC). Optionally, the polymer has 3 or more arms. Optionally, the polymer has 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12 arms. Optionally, the polymer has 3, 6 or 9 arms. Optionally, the polymer portion of the conjugate has a peak molecular weight of between 300,000 and 1,750,000 Da. Optionally, the polymer portion of the conjugate has a peak molecular weight between 500,000 and noi i ηη / ι znz / R / v 1,000,000 Days. Optionally, the polymer portion of the conjugate has a peak molecular weight of between 600,000 to 800,000 Da. Optionally, the polymer has 9 arms. Optionally, the dual VEGF / PDGF antagonist is covalently attached to the polymer. Optionally, the polymer is covalently bonded to at least one of an amino group, a hydroxyl group, a sulfhydryl group, and a carboxyl group. Optionally, the sulfhydryl group is from a naturally occurring cisterna residue. Optionally, the sulfhydryl group is from a cisterna residue added by recombinant DNA technology.

[00029] Any VEGF / PDGF dual antagonist as described above may be used in the treatment or prophylaxis of disease, particularly a neovascular disorder, optionally an ocular neovascular disorder, such as wet age-related macular degeneration. Brief Description of the Figures

[00030] Figure 1: Protein sequence of human PDGFR-β.

[00031] Figure 2: VEGFR-1 protein sequence.

[00032] Figure 3: VEGFR-2 protein sequence.

[00033] Figure 4: VEGFR-3 protein sequence.

[00034] Figure 5: Sequence of bevacizumab (DrugBank DB00112).

[00035] Figure 6: ranibizumab (published by Novartis).

[00036] Figure 7A, 7B: Protein sequence of A. PDGFR-P-GS10-anti-VEGF-A light chain and B. anti-VEGF-A heavy chain.

[00037] Figures 8A, 8B: A. Protein sequence of PDGFR-p-GG-anti-VEGF-A light chain and B. Anti-VEGF-A heavy chain.

[00038] Figures 9A, 9B: Protein sequence of A. PDGFR-p-GS10-anti-VEGF-A heavy chain (Fe wild type) and B. anti-VEGF-A light chain.

[00039] Figures 10A, 10B: Protein sequence of A. PDGFR-p-GG-anti-VEGF-A heavy chain (Fe wild type) and B. anti-VEGF-A light chain.

[00040] Figures 11A, 11B: Protein sequence of A. anti-VEGF-A heavy chain (Fe wild type)-GS21PDGFRp and B. anti-VEGF-A light chain.

[00041] Figures 12A, 12B: Protein sequence of A. PDGFR-p-GS21-anti-VEGF-A heavy chain (Q347C) and B. anti-VEGF-A light chain (TAF347).

[00042] Figures 13A, 13B: Protein sequence of A. PDGFR-3-GS21-anti-VEGF-A heavy chain (L443C) and B. anti-VEGF-A light chain (TAF443).

[00043] Figures 14A, 14B: Protein sequence of A. PDGFRp-GS10-anti-VEGF-A light chain and B. antiVEGF-A Fab.

[00044] Figures 15A, 15B: Protein sequence of A. PDGFRp-GG-anti-VEGF-A light chain and B. antiVEGF-A Fab.

[00045] Figures 16A, 16B: Protein sequence of A. PDGFRp-GS10-Fab anti-VEGF-Ay B. antinoi light chain i ηη / ι znz / B / v VEGF-A.

[00046] Figures 17A, 17B: Protein sequence of A. anti-VEGF-A PDGFRp-GG-Fab and B. antiVEGF-A light chain.

[00047] Figures 18A, 18B: Protein sequence of A. anti-VEGF-A-GS21-PDGFR3 Fab and B. antiVEGF-A light chain.

[00048] Figures 19A, 19B: Protein sequence of A. anti-VEGF-A PDGFRp-GS10-Fab with certain mutations and B. anti-VEGF-A light chain.

[00049] Figures 20A, 20B: Protein sequence of A. PDGFRp-heavy chain anti-VEGF-A and B. light chain anti-VEGF-A (1a).

[00050] Figures 21A, 21B: Protein sequence of A. PDGFR-p-(D2-D3)-anti-VEGF-A heavy chain and B. anti-VEGF-A light chain (1b).

[00051] Figures 22A, 22B: Protein sequence of A. anti-VEGF-A PDGFR-p-(D2-D3)-Fab and B. anti-VEGF-A light chain (2b).

[00052] Figures 23A, 23B: Protein sequence of A. PDGFR-p-(D2-D3)-6xGS-Fab anti-VEGF-A and B. anti-VEGF-A light chain (2b').

[00053] Figures 24A, 24B: Protein sequence of A. PDGFR-P-6xGS-Fab anti-VEGF-A and B. antiVEGF-A light chain.

[00054] Figures 25A, 25B: Protein sequence of A. anti-VEGF-A Fab 6xGS-PDGFR-P-(D2-D3) and B. anti-VEGF-A light chain (3).

[00055] Figure 26 shows the chemical structure of OG1448.

[00056] Figure 27 shows compound L.

[00057] Figure 28 shows compound K.

[00058] Figure 29 shows the synthesis of OG1802 from R3707.

[00059] Figure 30 shows OG1786.

[00060] Figure 31 shows the synthesis of OG1546 from OG1150.

[00061] Figure 32 shows the synthesis of OG1784 from OG1546 and OG1563.

[00062] Figure 33 shows the synthesis of OG1405 from OG1784.

[00063] Figure 34 shows the synthesis of OG1785 from OG1405.

[00064] Figure 35 shows the synthesis of OG1786 from OG1785.

[00065] Figure 36 shows OG1802.

[00066] Figure 37 shows a graph of the percentage of grade IV laser injuries.

[00067] Figure 38 shows compound E.

[00068] Figure 39 depicts OG1448.

[00069] Figure 40 shows relative angiogenesis using OG1448, Avastin and an anti-PDGF-BB antibody and various combinations thereof. noi i ηη / ι ζπζ / β / υιλι 00070 Figure 41 shows the % grade IV lesions compared to day in the CNV monkey model for the indicated compounds.

[00071] rabbit. Figure 42 shows the ocular pharmacokinetics of OG1448 versus afhbercept and rambizumab in the vitreous Brief Description of SEQ ID Nos.

[00072]

[00073] SEQ ID No.: 1 is the protein sequence of PDGFRb-GS10-anti-light chain VEGF-A (Bevacizumab). SEQ ID No.: 2 is the anti-VEGF-A Bevacizumab heavy chain.

[00074] SEQ ID No.: 3 is a protein sequence of PDGFRb-GG-anti-VEGF-A light chain (Bevacizumab).

[00075] SEQ ID No.: 4 is PDGFRβ-GS10-anti-VEGF-A heavy chain (Bevacizumab).

[00076] SEQ ID No.: 5 is Bevacizumab anti-VEGF-A light chain.

[00077]

[00078]

[00079] SEQ ID No.: 6 is PDGFRp-GG-anti-VEGF-A heavy chain (Bevacizumab). SEQ ID No.: 7 is anti-VEGF-A heavy chain (Bevacizumab)-GS21-PDGFRp. SEQ ID No.: 8 is the amino acid sequence of the heavy chain extracellular trap segment noi i ηη / ι ζηζ / κ / γ of TAF347: PDGFR-p-anti-VEGF-A heavy chain trap (Q347C). 00080 SEQ ID No.: 9 is the amino acid sequence of the heavy chain extracellular trap segment from: TAF443: PDGFR-P-anti-VEGF-A heavy chain trap (L443C) and SEQ ID No.: 10 is the amino acid sequence of the anti-VEGF-A heavy chain.

[00081]

[00082]

[00083]

[00084]

[00085]

[00086]

[00087]

[00088]

[00089]

[00090]

[00091] SEQ ID No.: 11 is human PDGFR-β. SEQ ID No.: 12 is the ranibizumab light chain. SEQ ID No.: 13 is the ranibizumab heavy chain. SEQ ID No.: 14 is human VEGFR-1. SEQ ID No.: 15 is human VEGFR-2. SEQ ID No.: 16 is human VEGFR-3. SEQ ID No.: 17 is a constant region of human IgG1. SEQ ID No.: 18 is a human layer chain constant region. SEQ ID No.: 19 is Figure 7A. PDGFR-GS10-light chain anti-VEGF-A. SEQ ID No.: 20 is Figure 8A. PDGFR-GG-anti-VEGF-A light chain. SEQ ID No.: 21 is a Bevacizumab Fab.

[00092]

[00093]

[00094] SEQ ID No.: 22 is an anti-VEGF-A PDGFR-p-GS10-Fab. SEQ ID No.: 23 is an anti-VEGF-A PDGFR-p-GG-Fab. SEQ ID No.: 24 is an anti-VEGF-A-GS21-PDGFR-β Fab.

[00095] SEQ ID No.: 25 is an anti-VEGF-A PDGFR-GS10-Fab with certain mutations.

[00096] SEQ ID No.: 26 is a PDGFRβ-anti-VEGF-A heavy chain protein sequence (1a).

[00097] SEQ ID No.: 27 is a protein sequence of PDGFR-β(D2-D3)-anti-VEGF-A heavy chain (1b)

[00098] SEQ ID No.: 28 is a protein sequence of PDGFR ^-(D2-D3)-Fab-anti-VEGF-A (2b).

[00099] SEQ ID No.: 29 is a protein sequence of PDGFR-β(D2-D3)-6xGS-Fab anti-VEGF-A. [000100] SEQ ID No.: 30 is a protein sequence of Fab-anti-VEGF-A-6GS-PDGFR-3 (D2-D3). [000101] SEQ ID No.: 31 is a nucleic acid encoding an anti-VEGF-PDGFR heavy chain fusion. [000102] SEQ ID No.: 32 is a nucleic acid encoding an anti-VEGF light chain. [000103] GGGGS (SEQ ID NO. 37), GGGS (SEQ ID NO. 38), GGGES (SEQ ID NO. 39), GGGGSGGGGS (SEQ ID NO. 40) and GGGGSGGGGSGGGGSGGGGSG) (SEQ ID NO. 41). [000104] The CDRs of ranibizumab are: CDRH1: GYDFTHYGMN, CDRH2: WINTYTGEPTYAADFKR, and CDRH3: YPYYYGTSHWYFDV (SEQ ID NOS.: 42-44), CDRL1: SASQDISNYLN, CDRL2: FTSSLHS, and CDRL3: QQYSTVPWT (SEQ ID Nos.: 45 -47). The CDRH1 of Bevacizumab is GYTFTNYGMN (SEQ ID No.: 48) and the CDRH3 is YPHYYGSSHWYFDV (SEQ ID No.: 49). Definitions [000105] A "neovascular disorder" is a disorder or disease state characterized by impaired, dysregulated, or unregulated angiogenesis. Examples of neovascular disorders include neoplastic transformation (eg, cancer), and ocular neovascular disorders including diabetic retinopathy and age-related macular degeneration. [000106] An "ocular neovascular1" disorder is a disorder characterized by impaired, dysregulated, or unregulated angiogenesis in a patient's eye. These disorders include optic disc neovascularization, iris neovascularization, retinal neovascularization, choroidal neovascularization, corneal neovascularization, vitreal neovascularization, glaucoma, pannus, pterygium, macular edema, diabetic retinopathy, diabetic macular edema, vascular retinopathy, retinal degeneration, ureitis, inflammatory diseases of the retina and proliferative vitreoretinopathy. [000107] A "polypeptide linker" is a polypeptide comprising two or more amino acid residues linked by peptide bonds that are used to link two peptides (eg, a VH and VL domain or a VH domain and an extracellular trap segment). . Examples of these linker peptides are well known in the art (see, for example Holliger P, Prospero T, Winter G. 1993. PNAS USA. 90:6444-6448; Poljak RJ. 1994. Production and Structure of Diabodies. Structure 2 :1121-1123). Exemplary linkers include G, GG, GGGGS, GGGS, and GGGES, and oligomers of these linkers (eg, GGGGSGGGGS and GGGGSGGGGSGGGGSGGGGSG). [000108] The dual antagonists or their biologics described herein are typically provided in isolated form. This means that an antagonist is typically at least 50% w / w pure from interfering proteins and other contaminants arising from its production or purification but does not exclude the possibility that the antagonist is combined with an excess of a proposed pharmaceutically acceptable excipient to facilitate its use. Sometimes the antagonists are at least 60, 70, 80, 90, 95, or 99% w / w pure from interfering proteins and contaminants from production or purification. Frequently an antagonist is the predominant macromolecular species that remains after purification. [000109] The term "antibody" includes intact antibodies and binding fragments thereof. A noi i ηη / ι znz / R / v binding fragment refers to a molecule other than an intact antibody comprising a portion of an intact antibody that binds to the antigen to which the intact antibody binds. Examples of ligation fragments include Fv, Fab', Fab'-SH, F(ab')2; diabodies, linear antibodies; single chain antibody molecules (eg scFv); and multispecific antibodies formed from antibody fragments. scFv fragments are described in Houston JS. 1991. Methods in Enzymol. 203:46-96. In addition, antibody fragments comprise single chain polypeptides having the characteristics of a VH domain, specifically capable of co-assembly with a VL domain, or a VL domain, specifically capable of co-assembly with a VH domain, or a a functional antigen binding site and thus provide the antigen binding property of full length antibodies. [000110] Specific binding of an antibody, extracellular trap segment, or dual antagonist to its target antigens means an affinity of at least 106,107,108,109, or 1010M'1. The specific binding is detectably greater in magnitude and distinguishable from the non-specific binding that occurs in at least one unrelated target. Specific binding may be the result of bond formation between particular functional groups or particular spatial fit (eg, lock-and-key type) while non-specific binding is usually the result of van der Waals forces. Specific binding however does not necessarily imply that an antibody or fusion protein binds to one and only one target. [000111] A basic structural unit of antibody is a tetramer of subunits. Each tetramer includes two identical pairs of polypeptide chains, each pair having a "light" (approximately 25 kDa) and a "heavy" (approximately 50-70 kDa) chain. The amino-terminal portion of each chain includes a variable region of about 100 to 110 or more amino acids primarily responsible for antigen recognition. This region is usually expressed linked to a cleavable signal peptide. The variable region without the signal peptide is sometimes referred to as a mature variable region. Thus, for example, a mature light chain variable region means a light chain variable region without the light chain signal peptide. However, reference to a variable region does not mean that a signal sequence is necessarily present; and in fact the signal sequences are cleaved once the antibodies or fusion proteins of the invention have been expressed and secreted. A pair of heavy and light chain variable regions defines a binding region of an antibody. The carboxy-terminal portion of the light and heavy chains respectively define light and heavy chain constant regions. The heavy chain constant region is primarily responsible for effector function. In IgG-type antibodies, the heavy chain constant region is divided into CH1, CH2 hinge, and CH3 regions. The CHi region is linked to the light chain constant region by disulfide and non-covalent linkage. The hinge region provides flexibility between the binding and effector regions of an antibody and also provides sites for intermolecular disulfide bonding between the two heavy chain constant regions in a tetramer subunit. The CH2 and CH3 regions are the primary site of FcR binding and effector functions. [000112] Light chains are classified as either kappa and lambda. Heavy chains are classified as gamma, mu, alpha, delta, or epsilon, and define the isotype of the antibody, IgG, IgM, IgA, IgD, and IgE, respectively. Within noi 1 ηη / ι znz / B / v of the light and heavy chains, the variable and constant regions are joined by a "J" segment of approximately 12 or more amino acids, with the heavy chain including a "D" segment of approximately 10 or more amino acids. (See generally, Fundamental Immunology (Paul, W., ed., 2nd ed. Raven Press, N.Y., 1989), Ch. 7) (Incorporated by reference in its entirety for all purposes). [000113] The mature variable regions of each light / heavy chain pair form the antibody binding site. Thus, the intact antibody has two binding sites, ie it is divalent. In natural antibodies, the binding sites are the same. However, biospecific antibodies can be produced in which the two binding sites are different (see for example Songsivilai S, Lachmann PC. 1990. Bispecific antibody: a tool for diagnosis and treatment of disease. Clin Exp Immunol. 79:315- 321; Kostelny SA, Colé MS, Tso JY. 1992. Formation of bispecific antibody by the use of leucine zippers. J Immunol. 148: 1547-1553). Variable regions all exhibit the same general structure of relatively conserved least variable regions (FRs) linked by three hypervariable regions, also called complementarity determining regions or CDRs. The CDR of the two strands of each pair are aligned by the least variable regions, allowing binding to a specific epitope. From N-terminus to C-terminus, both the light and heavy chains comprise the FRI, CDRI, FR2, CDR2, FR3, CDR3, and FR4 domains. For convenience, the heavy chain CDRs may be referred to as CDRh1, CDRh2, and CDRh3; the variable light chain CDR may be referred to as CDR11, CDR12 and CDR13. The amino acid assignment to each domain is according to the definition of KabatEA, et al. 1987 and 1991. Sequences of Proteins of Immunological Interest (National Institutes of Health, Bethesda, MD) or Chothia C, Lesk AM. 1987. Canonical Structures for the Hypervariable Regions of Immunoglobulins. J Mol Biol 196:901-917; Chothia C, et al. 1989. Conformations of Immunoglobulin Hypervariable Regions. Nature 342:877-883. Kabat also provides a widely used numbering convention (Kabat numbering) in which corresponding residues between different heavy chain variable regions or between different light chain variable regions are assigned the same number. Although Kabat numbering can be used for antibody constant regions, EU numbering is more commonly used, as is the case in this application. Although specific sequences for exemplary dual antagonists are provided, it will be appreciated that upon expression of the protein chains, one to several amino acids at the amino carboxy terminus of the light and / or heavy chain, particularly a C-terminal Usin residue. heavy chain, may be absent or derived in some proportion or in all molecules. [000114] The term "epitope" refers to a site on an antigen to which an antibody or extracellular trap segment binds. An epitope on a protein can be formed from contiguous amino acids or non-contiguous amino acids juxtaposed by tertiary folding of one or more proteins. Epitopes formed from contiguous amino acids (also known as linear epitopes) are typically retained on exposure to denaturing solvents whereas epitopes formed by tertiary folding (also known as conformational epitopes) are typically lost on treatment with denaturing solvents. An epitope typically includes at least 3, and more usually at least 5 or 8-10 amino acids in a single spatial conformation. Methods for determining the spatial conformation of epitopes include, for example, X-ray crystallography and noi i ηη / ι znz / R / v two-dimensional nuclear magnetic resonance. See, for example, Epitope Mapping Protocols, in Methods in Molecular Biology, Vol. 66, Glenn E. Morris, Ed. (1996). [000115] Antibodies that recognize the same epitopes or overlapping epitopes can be identified in a simple immunoassay that shows the ability of one antibody to compete with the binding of another antibody to a target antigen. The epitope of an antibody can also be defined by X-ray crystallography of the antibody (or Fab fragment) bound to its antigen to identify contact residues. [000116] Alternatively, two antibodies have the same epitope if all amino acid mutations in the antigen that reduce or eliminate binding of one antibody reduce or eliminate binding of the other. Two antibodies have overlapping epitopes if some amino acid mutations that reduce or eliminate the binding of one antibody reduce or eliminate the binding of the other. [000117] Competition between antibodies is determined by an assay in which an antibody under test inhibits the specific binding of a reference antibody to a common antigen (see, for example, Junghans et al., Cancer Res. 50:1495,1990 ). A test antibody competes with a reference antibody if an excess of a test antibody (eg, at least 2x, 5x, 10x, 20x, or 100x) inhibits the binding of the reference antibody by at least 50% but preferably 75%, 90% or 99% as measured in a competitive binding assay. Antibodies identified by the competition assay (competing antibodies) include binding to the same epitope as the reference antibody and antibodies that bind to an adjacent epitope close enough to the epitope bound by the reference antibody for spherical hindrance to occur. [000118] The term "patient" includes human subjects and other mammalian subjects receiving either prophylactic or therapeutic treatment. [000119] For the purposes of classifying amino acid substitutions as conservative or non-conservative, amino acids are grouped as follows: group I (hydrophobic side chains): met, ala, leu, ile; group II (neutral hydrophilic side chains): cys, ser, thr; group II (acid side chains): asp,glu; group IV (basic side chains): asn, gin, his, lys, arg: group V (residues that influence chain orientation): gly, pro; and group VI (aromatic side chains): trp, tyr, phe. Conservative substitutions comprise substitutions between amino acids in the same class. Non-conservative substitutions constitute the exchange of a member of one of these classes for a member of another. [000120] Percent sequence identities are determined with maximally aligned antibody sequences by the convention of Kabat numbering for a variable region or EU numbering for a constant region. After alignment, a region of antibody is bound (for example, the complete mature variable region of a heavy or light chain) is being compared to the same region of a reference antibody, the percent sequence identity between subject and the reference antibody regions is the number of positions occupied by the same amino acid in both the subject and the reference antibody region divided by the total number of aligned positions of the two regions, not counting gaps, multiplied by 100 to convert it to percentage. Sequence identities of other sequences can be determined by aligning the sequences using algorithms, such as BESTFIT, FASTA, and TFASTA in Wisconsin Genetics Software Package Release 7.0, Genetics Computer Group, 575 Science Dr., Madison, Wl, using separation parameters. by default, or by inspection, and the best alignment (ie, resulting in the highest percentage of sequence similarity over a comparison window). Percent sequence identity is calculated by comparing two optimally aligned sequences over a comparison window, determining the number of positions at which identical residues occur in both sequences to produce the number of matched positions, divide the number of matched positions times the total number of positions in the comparison window (ie, the window size), and multiplying the result by 100 to produce the percent sequence identity. [000121] Compositions or methods "comprising" one or more cited elements may include other elements not specifically mentioned. For example, a composition comprising the antibody can contain the antibody alone or in combination with other ingredients. [000122] The term "antibody-dependent cellular cytotoxicity", or ADCC, is a mechanism for inducing cell death that depends on the interaction of antibody-coated target cells (i.e., cells with bound antibody) with immune cells possessing lytic activity. (also referred to as effector cells). These effector cells include natural killer cells, monocytes / macrophages, and neutrophils. ADCC is activated by interactions between the Fc region of a cell-bound antibody and Fcy receptors, particularly FcyRI and FcyRIII, or immune effector cells such as neutrophils, macrophages, and natural killer cells. The target cell is eliminated by phagocytosis or lysis, depending on the type of mediator effector cell. Death of the antibody-coated target cell occurs as a result of effector cell activity. [000123] The term opsonization also known as antibody-dependent cellular phagocytosis”, or ADCP, refers to the process by which antibody-coated cells are neutralized, either in whole or in part, by phagocytic immune cells (eg, macrophages, neutrophils, and dendritic cells), which bind to the Fe region of immunoglobulin. [000124] The term "complement-dependent cytotoxicity" or CDC refers to a mechanism for inducing cell death in which the Fe effector domains of a target-bound antibody activate a series of enzymatic reactions that culminate in the formation of holes in the cell. target cell membrane. Typically, antigen-antibody complexes such as those in antibody-coated target cells bind to and activate the complement component C1 q which in turn activates the complement cascade leading to target cell death. Activation of complement can also result in the deposition of complement components on the surface of the target cell that facilitate ADCC by binding to complement receptors (eg, CR3) on leukocytes. [000125] A humanized antibody is a genetically engineered antibody in which the CDRs of a non-human "donor" antibody are grafted onto sequences of the human "acceptor" antibody (see, for example, Queen, US 5,530,101 and 5,585,089; Winter, US 5,225,539 , Carter, US 6,407,213, Adair, US 5,859,205 6,881,557, Foote, US 6,881,557). The acceptor antibody sequences may be, for example, a mature human antibody sequence noi i ηη / ι znz / B / v, a composite product of these sequences, a human antibody sequence consensus sequence, a germ line. Thus, a humanized antibody is an antibody that has some or all of the CDRs completely or substantially from a donor antibody and less variable or framework regions of variable region and constant regions, if present, completely or substantially from human antibody sequences. . Similarly, a humanized heavy chain has at least one, two, and usually all three CDRs of all or substantially all of a donor antibody heavy chain, and a heavy chain variable region and heavy chain constant region framework sequence, if present. present, substantially of the human heavy chain variable region framework and constant region sequences. Similarly, a humanized light chain has at least one, two, and usually all three CDRs of all or substantially all of a donor antibody light chain, and a light chain variable region and light chain constant region framework sequence, if is present, substantially all of the human light chain variable region framework and constant region sequences. Unlike nanobodies and dAbs, a humanized antibody comprises a humanized heavy chain and a humanized light chain. A CDR is a humanized antibody that is substantially free of a corresponding CDR in a non-human antibody when at least 85%, 90%, 95%, or 100% of the corresponding residues (as defined by Kabat) are identical between the respective CDRs. . Antibody chain variable region framework sequences or antibody chain constant region framework sequences are substantially of a human variable region framework sequence or human constant region framework sequence respectively when at least 85, 90, 95 or 100% of the corresponding residues defined by Kabat are identical. [000126] Although humanized antibodies frequently incorporate all six CDRs (preferably as defined by Kabat) of a mouse antibody, they can also be made up of fewer than all CDRs (eg, at least 3, 4, or 5 CDRs of a mouse antibody) (e.g., De Pascalis R, Iwahashi M, Tamura M, et al. 2002. Grafting “Abbreviated” Complementary-Determining Regions Containing Specificity-Determining Residues Essential for Ligand Contact to Engineer a Less Immunogenic Humanized Monoclonal Antibody. J Immunol 169:30763084, Vajdos FF, Adams CW, Breece TN, Presta LG, de Vos AM, Sidhu, SS 2002. Comprehensive functional maps of the antigen-binding site of an anti-ErbB2 antibody obtained with shotgun scanning mutagenesis. J Mol Biol. 320: 415-428; Iwahashi M, Milenic DE, Padlan EA, et al. 1999. CDR substitutions of a humanized monoclonal antibody (CC49): Contributions of individual CDRs to antigen binding and immunogenicity. Mol Immunol. 36:1079- 1091; Tamura M, Milenic DE, Iwahashi M, et al. 2000. Structural correlates of an anticarcinoma antibody: Identification of specificity determining regions (SDRs) and development of a minimally immunogenic antibody variant by retention of SDRs only. J Immunol. 164:1432-1441). [000127] A chimeric antibody is an antibody in which the mature heavy and light chain variable regions of a non-human (eg, mouse) antibody are combined with human heavy and light chain constant regions. These antibodies substantially or completely retain the binding specificity of the mouse antibody, and are approximately two-thirds human in sequence. noi i ηη / ι znz / R / v [000128] An inactivated antibody is a type of humanized antibody that retains some and usually all of the CDRs and some of the non-human variable region framework residues of a non-human antibody but replaces other variable region framework residues that may contribute to B or T cell epitopes, eg Padlan EA exposed residues. 1991. A possible procedure for reducing the immunogenicity of antibody variable domains while preserving their ligand-binding properties. Mol Immunol. 28:489-98) with residues at the corresponding positions of a human antibody sequence. The result is an antibody in which the CDRs are wholly or substantially that of a non-human antibody and the variable region frameworks of the non-human antibody are made more human-like by the substitutions. A human antibody can be isolated from a human, or otherwise result from the expression of human immunoglobulin genes (eg, in a transgenic mouse, in vitro or by phage display). Methods for producing human antibodies include the trioma method of Ostberg L. Pursch E. 1983. Human X (mouse x human) hybridomas stably producing human antibodies. Hybridoma 2:361-367; Ostberg, US Patent No. 4,634,664; and Engleman et al., US Pat. US 5,770,429, US 5,661,016, US 5,633,425, US 5,625,126, US 5,569,825, US 5,545,806, Nature 148, 1547-1553 (1994), Nature Biotechnology 14, 826 (1996), Kucherlapati, WO 91 / 107 (1996) and visualization methods in phage (see, for example, Dower et al., WO 91 / 17271 and McCafferty et al., WO 92 / 01047, US 5,877,218, US 5,871,907, US 5,858,657, US 5,837,242, US 5,733,743 and US 5,565,332. [000129] "Polymer" refers to a series of monomer groups linked together. A polymer is made up of multiple units of a donor monomer (a homopolymer) or of different monomers (a heteropolymer). High MW (molecular weight) polymers are prepared from monomers including, but not limited to, acrylates, methacrylates, acrylamides, methacrylamides, styrenes, vinyl pyridine, vinyl pyrrolidone, and vinyl esters such as vinyl acetate. . The additional monomers are useful in the high MW polymers of the present invention. When two different monomers are used, the two monomers are called "comonomers", meaning that the different monomers are copolymerized to form a single polymer. The polymer can be linear or branched. When the polymer is branched, each polymer chain is referred to as a "polymer arm." The end of the polymer arm linked to the starter portion is the proximal end, and the end of the growing chain of the polymer arm is the distal end. At the end of the growing polymer arm chain, the terminal polymer arm group can be the radical scavenger, or other group. [000130] "Initiator" refers to a compound capable of initiating a polymerization using the monomers or comonomers of the present invention. The polymerization can be a conventional free radical polymerization or preferably a controlled / "active" radical polymerization, such as atom transfer radical polymerization (ATRP), reversible addition-fragmentation-termination polymerization (RAFT), or nitroxide-mediated polymerization. (NMP). The polymerization may be a controlled "pseudo" polymerization, such as degenerative transfer. When the initiator is suitable for ATRP, it contains a labile bond that can be homolytically cleaved noi i ηη / ι znz / E / v to form an initiator fragment, I, which is a radical capable of initiating radical polymerization, and a radical scavenger, I', which reacts with the radical in the growing polymer chain to reversibly terminate the polymerization. The radical scavenger 1' is typically a halogen, but can also be an organic moiety, such as nitrile. In some embodiments of the present invention, the initiator contains one or more 2-bromoisobutyrate groups as sites for ATRP polymerization. [000131] A "chemical linker" refers to a chemical moiety that binds two groups together, such as a half-life extension moiety and a protein. The linker can be cleaved or not cleaved. Cleavable linkers can be hydrolyzable, enzymatically cleavable, pH sensitive, photolabile, or disulfide linkers, among others. Other linkers include homobifunctional and heterobifunctional linkers. A "linking group" is a functional group capable of forming a covalent bond consisting of one or more links to a bioactive agent. Non-limiting examples include those illustrated in Table 1 of WO2013059137 (incorporated by reference). [000132] The term "reactive group" refers to a group that is capable of reacting with another chemical group to form a covalent bond, that is, it is covalently reactive under suitable reaction conditions and generally represses a point of attachment for another substance. The reactive group on one moiety, such as maleimide or succinimidyl ester, is capable of chemically reacting with a functional group on a different moiety to form a covalent bond. In general, reactive groups include nucleophiles, electrophiles, and photoactivatable groups. [000133] “Phosphorylcholine”, also denoted as “PC, refers to the following: noi i ηη / ι znz / R / v where * denotes the junction point. Phosphorylcholine is a zwitterionic group and includes salts (such as inner salts), and protonated and deprotonated forms thereof. [000134] "Phosphorylcholine-containing polymer" is a phosphorylcholine-containing polymer. "Zwitterion-containing polymer" refers to a polymer that contains a zwitterion. [000135] Polymer containing pol¡(acrylo¡lox¡ethylphosphorylcholine) refers to a polymer containing 2(acrylo¡lox¡)eth¡l-2-(trimethylammon¡o) ethyl phosphate (HEA-PC shown below in Example 51) as monomer. [000136] Poly(methacryloyloxyethylphosphorylcholine)-containing polymer refers to a polymer containing 2(methacryloyloxy)ethyl-2-(trimethylammonium)ethyl phosphate (HEMA- PC) as monomer. [000137] "Molecular weight" in the context of the polymer can be expressed as either a number average molecular weight, or a weight average molecular weight or as a peak molecular weight. Unless otherwise indicated, all references to molecular weight herein refer to peak molecular weight. These number average (Mn), weight average (Mw) and peak (Mp) molecular weight determinations can be measured using size exclusion chromatography or other liquid chromatography techniques. Other methods for measuring molecular weight values ​​can also be used, such as the use of end group analysis or the measurement of colligative properties (for example, freezing point depression, boiling point elevation, or osmotic pressure) to determine the number average molecular weight, or the use of light scattering, ultracentrifugation, or viscometry techniques to determine the weight average molecular weight. In a preferred embodiment of the present invention, molecular weight is measured by SEC-MALS (size exclusion chromatography-multi angle light scattering). The polymeric reagents of the invention are typically polydisperse (ie, number average molecular weight and weight average molecular weight of the polymers are not the same), preferably possessing low polydispersity values ​​of, for example, less than about 1.5, as judged, for example, by the PDI value derived from the SEC-MALS measurement. In other embodiments, the polydispersities (PDI) are more preferably in the range of about 1.4 to about 1.2, even more preferably less than about 1.15, and even more preferably less than about 1.10, even more preferably less than about 1.05, and more preferably less than about 1.03. [000138] The phrase "a" or "an" entity refers to one or more of that entity; for example, a compound refers to one or more compounds or at least one compound. As such, the terms "a" (or "to one"), "one or more", and "at least one" may be used interchangeably herein. [000139] "Approximately" means variation that can be seen in measurements taken between different instruments, samples and sample preparations. [000140] "Protected", "protected form", "protecting group" and "protecting group" refer to the presence of a group (i.e., the protecting group) that prevents or blocks the reaction of a particular chemically reactive functional group in a molecule under certain reaction conditions. Protecting groups vary depending on the type of chemically reactive group being protected as well as the reaction conditions to be employed and the presence of additional reactive or protecting groups on the molecule, if any. Suitable protecting groups include those as found in Greene et al., "Protective Groups In Organic Synthesis," 3rd Edition, John Wiley and Sons, Inc., New York, 1999. [000141] "Alkyl" refers to a saturated, branched, straight aliphatic radical having the indicated number of carbon atoms. For example, CrCe alkyl includes, but is not limited to, methyl, ethyl, propyl, isopropyl, butyl, isobutyl, sec-butyl, tert-butyl, pentyl, isopentyl, hexyl, etc. Other alkyl groups include, but are not limited to, heptyl, octyl, nonyl, decyl, etc. Alkyl can include any number of carbons, such as 1-2,1-3, 1-4,1-5,1-6,1-7,1-8, 1-9, 1-10, 2-3, 2-4, 2-5, 2-6, 3-4, 3-5, 3-6, 4-5, 4-6 and 5-6. The alkyl group is typically monovalent, but can be noi i ηη / ι znz / B / v divalent, such as when the alkyl group is bonded together to two moieties. [000142] The term "lower" referred to above and hereinafter with respect to organic radicals or compounds respectively defines a radical compound which may be branched or unbranched with up to and including 7, preferably up to and including 4 and (as unbranched) one or two carbon atoms. [000143] "Alkylene" refers to an alkyl group, as defined above, that links at least two different groups, ie, a divalent hydrocarbon radical. The two moieties attached to the alkylene can be attached to the same or different alkylene atoms. For example, a straight chain alkylene can be a divalent radical of (CH2)n, where n is 1, 2, 3, 4, 5, or 6. Alkylene groups include, but are not limited to, methylene, ethylene, propylene, , isopropylene, butylene, isobutylene, sec-butylene, pentylene and hexylene. [000144] Substituents for alkyl and heteroalkyl radicals (including those groups frequently referred to as alkylene, alkenyl, heteroalkylene, heteroalkenyl, alkynyl, cycloalkyl, heterocycloalkyl, cycloalkenyl, and heterocycloalkenyl) may be a variety of groups selected from: -OR ', =0, =NR', =N-OR', -NR'R”, -SR', -halogen, -SIRW”, -OC(O)R', -C(O)R', -CO2R ', -CONR'R, -OC(O)NR'R”, -NR”C(O)R', -N R'-C(O)NR”R'”, -NR”C(O)2R ', -NH-C(NH2)=NH, -NR'C(NH2)=NH, -NH-C(NH2)=NR', -S(O)R', -S(O)2R', - S(O)2NR'R”, CN and -NO2in a number ranging from zero to (2m' +1), where m' is the total number of carbon atoms in this radical, R', R” and R' ” each independently refer to hydrogen, unsubstituted (Ci-CiC4)alkyl and heteroalkyl, unsubstituted aryl, aryl substituted with 1-3 halogens, unsubstituted alkyl, alkoxy and thioalkoxy groups, or aryl-(CiC4)alkyl groups. When R' and R'' are attached to the same nitrogen atom, they can combine with the nitrogen atom to form a 5-, 6-, or 7-membered ring. For example, -NR'R'' is intended to include 1-pyrrolidinyl and 4-morpholinyl. The term "alkyl" includes groups such as haloalkyl (eg, -CF3 and -CH2CF3) and acyl (eg, -C(O)CH3, -C(O)CF3, -C(O)CH2OCH3, and the like). Preferably, the substituted alkyl and heteroalkyl groups have 1 to 4 substituents, more preferably 1, 2 or 3 substituents. The exceptions being perhalo-alkyl groups (eg pentafluoroethyl and the like) are also preferred and contemplated by the present invention. [000145] Substituents for alkyl and heteroalkyl radicals (including those groups referred to as alkylene, alkenyl, heteroalkylene, heteroalkenyl, alkynyl, cycloalkyl, heterocycloalkyl, cycloalkenyl, and heteroacycloalkenyl) may be one or more of a variety of groups selected from , but not limited to: -OR', =0, =NR', =N-OR', -NR'R”, -SR', -halogen, -SÍR'R”R”', -OC(O) R', -C(O)R', -CO2R', -CONR'R”, -OC(O)NR'R”, -NR”C(O)R', -NR' -C(O)NR ”R'”, -NR”C(O)2R', -NR-C(NR'R”R'”)=NR””, -NR-C(NR'R”)=NR'”, -S (O)R', -S(O)2R', -S(O)2NR'R”, -NRSO2R', CN, and -NO2in a number ranging from zero to (2m' +1), where m' is the total number of carbon atoms in this radical. R', R", R' and R"" each refer preferably independently to hydrogen, substituted or unsubstituted heteroalkyl, substituted or unsubstituted aryl, for example, aryl substituted with 1-3 halogen, alkyl, alkoxy or thioalkoxy groups. substituted or unsubstituted, or arylalkyl groups. When a compound of the invention includes more than one R group, for example, each of the R groups are independently selected as are each R', R", noi 1 ηη / ι znz / R / v group R"' and R"" when more than one of these groups is present. When R' and R" are attached to the same nitrogen atom, they can combine with the nitrogen atom to form a 5-, 6-, or 7-membered ring. For example, -NR'R" is intended to include, but is not limited to, 1-pyrrolidinyl and 4-morpholinyl. From the foregoing discussion of substituents, one skilled in the art will understand that the term "alkyl" is intended to include groups including carbon atoms attached to groups other than hydrogen groups, such as haloalkyl (for example -CF3 and -CH2CF3), and acyl (eg, -C(O)CH3, -C(O)CF3, -CIOjCHZOCHa, and the like). [000146] "Alkoxy" refers to an alkyl group having an oxygen atom that either connects the alkoxy group to the point of attachment or is bonded to two carbons of the alkoxy group. Alkoxy groups include, for example, methoxy, ethoxy, propoxy, iso-propoxy, butoxy, 2-butoxy, iso-butoxy, sec-butoxy, tert-butoxy, pentoxy, hexoxy, etc. Alkoxy groups may be further substituted with a variety of substituents described herein. For example, alkoxy groups may be substituted with halogens to form a "halo-alkoxy" group. [000147] "Carboxyalkyl" means an alkyl group (as defined herein) substituted with a carboxy group. The term "carboxycylalkyl" means a cycloalkyl group (as defined herein) substituted with a carboxy group. The term "alkoxyalkyl" means an alkyl group (as defined herein) substituted with an alkoxy group. The term "carboxy" used herein refers to carboxylic acids and their esters. [000148] "Haloalkyl" refers to alkyl as defined above where some or all of the hydrogen atoms are substituted with halogen atoms. Halogen (halo) preferably represents chloro or fluoro, but can also be bromo or iodo. For example, haloalkyl includes trifluoromethyl, fluoromethyl, 1,2,3,4,5-pentafluorophenyl, etc. The term "perfluoro" defines a radical compound having all available hydrogens replaced with fluorine. For example, perfluorophenyl refers to 1,2,3,4,5-pentafluorophenyl, perfluoromethyl refers to 1,1,1-trifluoromethyl, and perfluoromethoxy refers to 1,1,1-trifluoromethoxy. [000149] "Fluoro-substituted alkyl" refers to an alkyl group in which one, some, or all of the hydrogen atoms have been replaced by fluorine. [000150] "Cytosine" in the context of this invention is a member of a group of protein signaling molecules that may participate in cell-cell communication in immune and inflammatory responses. Cytokines are typically small, water-soluble glycoproteins that have a mass of approximately 8-35 kDa. [000151] "Cycloalkyl" refers to a cyclic hydrocarbon group containing from about 3 to 12, 3 to 10, or 3 to 7 endocyclic carbon atoms. Cycloalkyl groups include fused ring, bridged, and spiro structures. [000152] "Endocyclic" refers to an atom or group of atoms comprising part of a cyclic ring structure. [000153] "Exocyclic" refers to an atom or group of atoms that bind but do not define the cyclic ring structure. [000154] "Cyclic ether alkyl" refers to a 4- or 5-membered cyclic alkyl group having 3 or 4 endocyclic carbon atoms and one endocyclic sulfur or oxygen atom (for example, oxetane, thiethane, tetrahydrofuran, noi 1 ηη / ι znz / R / v tetrahydrothiophene); or a 6 to 7 membered cyclic alkyl group having 1 or 2 endocyclic oxygen or sulfur atoms (for example, tetrahydropyran, 1,3-dioxane, 1,4-dioxane, tetrahydrothiopyran, 1,3-dithiane, 1,4 -dithiane, 1,4-oxathiane). [000155] "Alkenyl" refers to either a straight or branched chain hydrocarbon of 2 to 6 carbon atoms, having at least one double bond. Examples of alkenyl groups include, but are not limited to, vinyl, propenyl, isopropenyl, 1-butenyl, 2-butenyl, isobutenyl, butadienyl, 1-pentenyl, 2-pentenyl, isopentenyl, 1,3-pentadienyl, 1,4 -pentadienyl, 1-hexenyl, 2-hexenyl, 3-hexenyl, 1,3-hexadienyl, 1,4-hexadienyl, 1,5-hexadienyl, 2,4-hexadienyl, or 1,3,5-hexatrienyl. Alkenyl groups can also have 2 to 3, 2 to 4, 2 to 5, 3 to 4, 3 to 5, 3 to 6, 4 to 5, 4 to 6, and 5 to 6 carbons. The alkenyl group is typically monovalent, but can be divalent, such as when the alkenyl group links two moieties together. [000156] "Alkenylene" refers to an alkenyl group, as defined above, that links at least two other groups, ie, a divalent hydrocarbon radical. The two alkenylene-linked moieties can be linked to the same atom or to different alkenylene atoms. Alkenylene groups include, but are not limited to, ethenylene, propenylene, isopropenylene, butenylene, isobutenylene, sec-butenylene, pentenylene, and hexenylene. [000157] "Alkynyl" refers to either a straight or branched chain group of 2 to 6 carbon atoms, having at least one triple bond. Examples of an alkynyl group include, but are not limited to acetylenyl, propynyl, 1-butynyl, 2-butynyl, isobutynyl, sec-butynyl, butadinyl, 1-pentynyl, 2-pentynyl, isopentinyl, 1,3-pentadinyl, 1,4-pentadinyl, 1-hexynyl, 2-hexynyl, 3-hexynyl, 1,3-hexadinyl, 1,4-hexadiynyl, 1,5-hexadiynyl, 2,4-hexadiynyl, or 1,3,5hexatriynyl. Alkynyl groups can have 2-3, 2-4, 2-5, 3-4, 3-5, 3-6, 4-5, 4-6, and 5-6 carbons. . The alkynyl group is typically monovalent, but can be divalent, such as when the alkynyl group links two moieties together. [000158] "Alkynylene" refers to an alkynyl group, as defined above, that links at least two other groups, ie, a divalent hydrocarbon radical. The two alkynylene bonded moieties can be bonded to the same or different alkynylene atoms. Alkynylene groups include, but are not limited to, ethynylene, propynylene, butynylene, sec-butynylene, pentynylene, and hexynylene. [000159] "Cycloalkyl" refers to a saturated or partially unsaturated monocyclic bridged polycyclic or fused bicyclic ring assembly containing from 3 to 12 ring atoms, or the indicated number of atoms. Monocyclic rings include, for example, cyclopropyl, cyclobutyl, cyclopentyl, cyclohexyl, and cyclooctyl. Bicyclic and polycyclic rings include, for example, norbornane, decahydronaphthalene, and adamantane. For example, C3-8cycloalkyl includes cyclopropyl, cyclobutyl, cyclopentyl, cyclohexyl, cyclooctyl, and norbornane. [000160] "Cycloalkylene" refers to a cycloalkyl group, as defined above, that links at least two other groups, ie, a divalent hydrocarbon radical. The two moieties attached to the cycloalkylene can be attached to the same or different cycloalkylene atoms. Cycloalkylene groups include, but are not limited to, cyclopropylene, cyclobutylene, cyclopentylene, cyclohexylene, and cyclooctylene. [000161] "Heterocycloalkyl" refers to a ring system having from 3 to about 20 ring members and from 1 to about 5 heteroatoms such as N, O and S. The noi 1 ηη / ι znz / B / v Additional heteroatoms may also be useful, including but not limited to B, Al, Si, and P. Heteroatoms may also be oxidized, such as, but not limited to -S(O)- and -S(O) 2-. For example, the heterocycle includes, but is not limited to, tetrahydrofuranyl, tetrahydrothiophenyl, morpholino, pyrrolidinyl, pyrrolinyl, imidazolidinyl, imidazolinyl, pyrazolidinyl, pyrazolinyl, piperazinyl, piperidinyl, indolinyl, quinuclidinyl, and 1,4-dioxa-8-aza-spiro[ 4.5]dec-8-¡lo. [000162] "Heterocycloalkylene" refers to a heterocycloalkyl group, as defined above, that links at least two other groups. The two moieties attached to the heterocycloalkylene can be attached to the same or different heterocycloalkylene atoms. [000163] "Aryl" refers to an aromatic, monocyclic or fused bicyclic, tricyclic or larger ring assembly containing from 6 to 16 ring carbon atoms. For example, aryl can be phenyl, benzyl or naphthyl, preferably phenyl. "Arylene" means a divalent radical derived from an aryl group. Aryl groups may be mono-, di-, or tri-substituted by one, two, or three radicals selected from alkyl, alkoxy, aryl, hydroxy, halogen, cyano, amino, amino-alkyl, trifluoromethyl, alkylenedioxy, and oxy-Cz- Cg-alkylene; all of which are further optionally substituted, for example as defined herein above; or 1- or 2-naphthyl; or 1- or 2-phenanthrenyl. Alkylenedioxy is a divalent substituent attached to two adjacent phenyl carbon atoms, eg, methylenedioxy or ethylenedioxy. Oxy-C2-C3-alkylene is also a divalent substituent attached to two adjacent phenyl carbon atoms, for example, oxyethylene or oxypropylene. An example for oxy-C2-C3-alkylene-phenyl is 2,3-dihydrobenzofuran-5-yl. [000164] Preferred aryl is naphthyl, phenyl or phenyl mono- or di-substituted by alkoxy, phenyl, halogen, alkyl or trifluoromethyl, especially phenyl or phenyl-mono- or di-substituted by alkoxy, halogen or trifluoromethyl, and in particular phenyl. [000165] Examples of substituted phenyl groups as R are for example, 4-chlorophen-1-yl, 3,4-dichlorophen-1-yl, 4-methoxyphen-1-yl, 4-methylphen-1 -yl, 4-am¡nomethylfen-1-¡lo, 4-methoxy¡lam¡nomethylfen-1-¡lo, 4-hydroxyethylaminomethylphen-l-yl, 4-hydroxy¡ethyl-( methyl)-aminomethylphen-1-yl, 3-aminomethylphen-1-yl, 4-N-acetylaminomethylphen-1-¡lo, 4-am¡nophen-1-¡lo, 3-am¡nophen-1-¡lo, 2-am¡nophen-1-yl, 4-phenyl -phen-1-¡lo, 4-(¡midazol-1-¡l)-phenyl, 4-(¡midazol-1-¡lmethyl)-phen-1-¡lo, 4-(morpholin- 1-¡l)-phen-1-¡lo, 4-(morpholin-1-¡lmethyl)-phen-1-¡lo, 4-(2-methoxyethylam¡nomethyl)-phen-1-¡ lo and 4-(pyrrol¡din-1-ylmethyl)-phen-1-¡lo, 4-(thiophenyl)-phen-1-¡lo, 4-(3-thiophenyl)-phen-1- ¡lo, 4-(4-methylp¡peraz¡n-1-yl)-phen-1-¡lo, and optionally substituted 4-(piperidinyl)-phenyl and 4-(pyrádin¡l)-phenyl in the heterocyclic ring. [000166] "Arylene" refers to an aryl group, as defined above, that links at least two other groups. The two moieties bonded to the arylene bond to different arylene atoms. Arylene groups include, but are not limited to, phenylene. [000167] "Arylene-oxy" refers to an arylene group, as defined above, where one of the moieties bonded to the arylene is bonded through an oxygen atom. Arylene-oxy groups include, but are not limited to, phenylene-oxy. [000168] Similarly, the substituents for the aryl and heteroaryl groups are varied and selected from noi i ηη / ι znz / B / v of halogen -OR', -OC(O)R', -NR'R ”, -SR', -R', -CN, -NO2, -CO2R', -CONR'R”, -C(O)R', -OC(O)NR'R”, -NR”C(O )R', -NR” C(O)2R', ,-NR'-C(O)NR”R”', -NH-C(NH2)=NH, -NR'C(NH2)=NH, - NH-C(NH2)=NR', -S(O)R', -S(O)2R', -S(O)2NR'R”, -N3, -CH(Ph)2, perfluoro(Ci- C4)alkoxy¡, and perfluoro(Ci-C4)alkyl, in a number ranging from zero to the total number of open valences to the aromatic ring system; and where R', R” and R'” are independently selected from hydrogen, (CiCsjalkyl and heteroalkyl, aryl and unsubstituted heteroaryl, (unsubstituted aryls)-(Ci-C4)alkyl, and (unsubstituted aryl)ox ¡(Ci-C4)alkyl. [000169] Two of the substituents on adjacent aryl or heteroaryl ring atoms may be optionally replaced with a substituent of the formula -T-C(O)-(CH2)q-U-, where T and U are independently -NH-, O-, -CH2- or a double bond, and q is an integer from 0 to 2. Alternatively, two of the substituents on adjacent aryl or heteroaryl ring atoms may optionally be replaced with a substituent of the formula- A-(CH2)rB-, where A and B are independently -CH2-, -O-, -NH-, -S-, -S(O)-, -S(O)2-, -S(O) 2NR'-o a double bond and r is an integer from 1 to 3. One of the individual bonds in the new ring thus formed can optionally be replaced with a double bond. Alternatively, two of the substituents on adjacent aryl or heteroaryl ring atoms may optionally be replaced with a substituent of the formula (CH2)s-X-(CH2)t-, where s and t are independently integers from 0 to 3, and X is -O-, -NR'R'”-, -S-, -S(O)-, S(O)2-, or -S(O)2NR'-. The R' substituent in -NR'- and -S(O)2NR' is selected from hydrogen or unsubstituted (Ci-Cejalkyl. [000170] "Heteroaryl" refers to a monocyclic or fused bicyclic or tricyclic aromatic ring assembly containing 5 to 16 ring atoms, where 1 to 4 of the ring atoms are one heteroatom each of N, O, or S. For example, heteroaryl includes pyridyl, indolyl, indazolyl, quinoxalinyl, quinolinyl, isoquinolinyl, benzothienyl, benzofuranyl, furanyl, pyrrolyl, thiazolyl, benzothiazolyl, oxazolyl, isoxazolyl, triazolyl, tetrazolyl, pyrazolyl, imidazolyl, thienyl, or any other substituted radical, especially mono- or di-substituted, for example by alkyl, nitro or halogen Pyridyl represents 2 -, 3- or 4-pyridyl, advantageously 2- or 3-pyridyl Thienyl represents 2- or 3-thienyl Quinolinyl preferably represents 2-, 3-, or 4-quinolinyl Isoquinoline preferably represents 1-, 3-, or 4-Isoquinolinyl Benzopyranyl, benzothiopyranyl preferably represents 3-benzopyranyl or 3-benzothiopyranyl, respectively. Thiazolyl preferably represents 2- or 4-thiazolyl, and more preferred 4-thiazolyl. Triazolyl is preferably 1-, 2- or 5-(1,2,4-triazolyl). Tetrazolyl is preferably 5-tetrazolyl. [000171] Preferably, heteroaryl is pyridyl, indolyl, quinolinyl, pyrrolyl, thiazolyl, isoxazolyl, triazolyl, tetrazolyl, pyrazolyl, imidazolyl, thienyl, furanyl, benzothiazolyl, benzofuranyl, isoquinolinyl, benzothienyl, oxazolyl, indazolyl, or any of the radicals substituted, especially mono- or di-substituted. [000172] The term "heteroalkyl" refers to an alkyl group having 1 to 3 heteroatoms such as N, O, and S. Additional heteroatoms may also be useful, including, but not limited to B, Al, Si, and P. Heteroatoms can also be oxidized, such as, but not limited to -S(O)- and -S(O)2-, For example, heteroalkyl can include ethers, thioethers, alkyl amines, and alkyl thiols. noi i ηη / ι 7Π7 / Β / Υ [000173] The term "heteroalkylene" refers to a heteroalkyl group, as defined above, that links at least two other groups. The two moieties attached to the heteroalkylene can be attached to the same or different heteroalkylene atoms. [000174] "Electrophile" refers to an ion or atom or collection of atoms, which may be ionic, having an electrophilic center, ie, a center that is seeking electrons, capable of reacting with a nucleophile. An electrophile (or electrophilic reagent) is a reactant that forms a bond to its reaction partner (the nucleophile) by accepting both bonding electrons from that reaction partner. [000175] "Nucleophile" refers to an ion or atom or collection of atoms, which may be ionic, that has a nucleophilic center, ie, a center that is seeking an electrophilic center or capable of reacting with an electrophile. A nucleophile (or nucleophilic reagent) is a reagent that forms a bond to its reaction partner (the electrophile) by donating both bonding electrons. A "nucleophilic group" refers to a nucleophile after it has reacted with a reactive group. Non-limiting examples include amino, hydroxyl, alkoxy, haloalkoxy, and the like. [000176] "Maleimido" refers to a pyrrol-2,5-dione-1-yl group having the structure: noi i nn / ι 7Π7 / Ε / Υ which on reaction with a sulfhydryl (for example, a thioalkyl) forms an -S-maleimido group having the structure: where "·" indicates the point of attachment of the maleimido group and indicates the point of attachment of the sulfur atom of the thiol to the rest of the group having the original sulfhydryl. [000177] For the purpose of this disclosure, "naturally occurring amino acids" found in proteins and polypeptides are L-alanine, L-arginine, L-asparagine, L-aspartic acid, L-cysteine, L-glutamine, L -glutamic acid, L-glycine, L-histidine, L-isoleucine, L-leucine, L-lysine, L-methionine, L-phenylalanine, L-proline, L-serine, L-threonine, L-tryptophan, L-tyrosine , and or L-valine. "Non-naturally occurring amino acids" found in proteins are any amino acids other than those referred to as naturally occurring amino acids. Non-naturally occurring amino acids include, without limitation, the D-isomers of naturally occurring amino acids, and mixtures of D- and L-isomers of naturally occurring amino acids. Other amino acids, such as N-alpha-methyl-amino acids (for example sarcosine), 4-hydroxyproline, desmosine, isodesmosine, 5-hydroxylysine, epsilon-N-methyl-lysine, 3-methylhistidine, although they are found in proteins that occur naturally occurring, non-naturally occurring amino acids found in proteins are considered for the purpose of this disclosure as being generally introduced by a means other than ribosomal translation of mRNA. [000178] "Linear" in reference to the overall geometry, architecture or structure of a polymer, refers to a polymer having a single polymer arm. [000179] “Branched, referring to the general geometry, architecture, or structure of a polymer, refers to a polymer that has 2 or more polymer “arms” extending from a core structure contained within an initiator. The initiator can be scaled up in an atom transfer radical polymerization (ATRP) reaction. A branched polymer has 2 polymer chains (arms), 3 polymer arms, 4 polymer arms, 5 polymer arms, 6 polymer arms, 7 polymer arms, 8 polymer arms, 9 polymer arms or more. Each polymer arm extends from a polymer start site. Each polymer start site is capable of being a site for the growth of a polymer chain by addition of monomers. For example and not by way of limitation, using ATRP, the polymer initiation site in an initiator is typically an organic halide that undergoes a reversible redox process catalyzed by a transition metal compound such as cuprous halide. Preferably, the halide is a bromide. [000180] "Pharmaceutically acceptable excipient" refers to an excipient that can be included in the composition of the invention and that does not cause significant adverse toxicological effect in the patient and is tested or approved by the FDA for therapeutic use, particularly in humans . Non-limiting examples of pharmaceutically acceptable excipients include water, NaCI, normal saline, lactated Ringer's, normal sucrose, normal glucose, and the like. [000181] Dual antagonists are administered in an effective regimen meaning a dose, route of administration, and frequency of administration that delays onset, reduces severity, inhibits further deterioration, and / or ameliorate at least one sign or symptom of a disorder. If a patient is already suffering from a disorder, the regimen may be referred to as a therapeutically effective regimen. If the patient is at elevated risk of the disorder relative to the general population but is not yet experiencing symptoms, the regimen may be referred to as a prophylactically effective regimen. In some cases, therapeutic or prophylactic efficacy may be observed in an individual patient relative to historical controls or prior experience in the same patient. In other cases, therapeutic or prophylactic efficacy may be demonstrated in a pre-clinical or clinical trial in a population of treated patients relative to a control population of untreated patients. [000182] The "biological half-life" of a substance is a pharmacokinetic parameter that specifies the time required for half of the substance to be removed from a tissue or an organism after introduction of the substance. [000183] "HEMA-PC" is 2-(methacryloyloxyethyl)-2'-(trimethylammoniomethyl)phosphate. [000184] "TAF" means PDGFRp-GS10-anti-VEGF-A heavy chain / anti-VEGF-A light chain where amino acids 1-282 of the heavy chain correspond to amino acids 33-314 of human PDGFR-β ( UniProtKB / Swiss-Prot: P09619.1), fused as a single open reading frame via a glycine-serine linker (GGGGSGGGG) linked to the N-terminus of a Bevacizumab heavy chain sequence having the following mutations in the region variable: T28D, N31H, H97Y, S100aT (Ferrara N, Damico L, Shams N, et al. 2006. Development of Ranibizumab, an anti-vascular endothelial growth factor antigen binding fragment, as therapy for neovascular age-related macular degeneration. Retina 26(8):859-870); and the following in the Fe region: L234A, L235A, and G237A (EU numbering) (Strohl WR. 2009. Optimization of Fc-mediated effector functions of monoclonal antibodies. Curr Opin in Biotech. 20:685-691). The light chain is the bevacizumab light chain that has an M4L mutation. TAF normally exists as a dimer having two heavy chains and two light chains. TAF may or may not have carbohydrate or other post-translational modifications after it is expressed in cells. TAF is also sometimes called TAFwt or TAFWT, indicating that the molecule in question does not have either the Q347 or L443C mutations in the heavy chain (Fe region) or in TAF347 or TAF443, as defined infra. [000185] "TAF347" is the same as TAF except that it has the Q347C mutation. [000186] "TAF433" is the same as TAF, except that it has the L443C mutation. TAF443 is sometimes referred to as noi i ηη / ι znz / R / v herein as OG1321. [000187] OG1786” is a 9-arm primer used for the synthesis of polymer with the structure shown in Figure 35, which represents that the salt form of OG1786 with trifluoroacetic acid. OG1786 can be used in accordance with the present invention as other salts or as the free base. [000188] OG1801” is an approximately (+ / - 15%) 750 kDa (either by Mn or Mp) polymer produced using OG1786 as a primer for ATRP synthesis using the HEMA-PC monomer. [000189] OG1802” is OG1801 with a maleimide functionality added and is shown in Figure 36 where ni, n2, na, Π4, ns, ne, n?, na and ng are each an integer (positive) (from 0 to about 3000) such that the total molecular weight of the polymer is (Mw) 750,000 ± 15% daltons. [000190] OG1448" is TAF443 conjugated to biopolymer OG1802. Detailed description of the invention I. General Points [000191] The present invention provides a dual VEGF / PDGF antagonist comprising a VEGF antagonist linked to a PDGF antagonist. The VEGF antagonist is an antibody to VEGF or VEGFR or is a VEGFR extracellular trap segment (i.e., a segment of the extracellular region of one or more VEGFR receptors that inhibits the binding of at least one VEGFR to at least one VEGF The PDGF antagonist is an antibody to PDGF or PDGFR or is a PDGFR extracellular trap segment (i.e., segment of the extracellular region of one or more PDGFRs, which inhibits the binding of at least one PDGFR and at least one PDGF At least one of the antagonists is not an antibody, or put another way, at least one of the antagonists is an extracellular trap segment Preferably, the dual antagonist includes an antibody antagonist and an extracellular trap segment antagonist. extracellular trap.In this dual antagonist, the extracellular trap segment is preferentially fused, optionally via a linker, to the N-terminus of the antibody heavy chain.The antibody light chain is complexed with the antibody heavy chain in a similar manner to that in a natural antibody. These dual antagonists are preferably provided in the form of conjugates with a half-life extension moiety conjugated to the dual antagonist. Preferably, a cysteine ​​residue that has been introduced into the antagonist is used for conjugation. More preferably, the cysteine ​​residue is at positions 347 to 443 of an IgG 1 heavy chain. It is preferred that the half-life extension portion is a zwitterionic polymer. More preferably, the zwitterionic polymer is a phosphorylcholine-containing polymer. [000192] Angiogenesis is the process by which new blood vessels are created and plays a crucial role in development (from embryo to adult) and wound healing (restoring blood flow to damaged or injured tissue) . However, when angiogenesis is dysregulated, it contributes to the pathologies of many disorders, including cancer, psoriasis, arthritis, and blindness. Carmeliet P. 2003. Angiogenosis in health and disease. Nature Med 9(6):653-660. [000193] Abnormal angiogenesis is associated with age-related macular degeneration (a leading cause of blindness in the elderly) and with cancer. Angiogenesis is characterized by an increase in proliferating stromal and endothelial cells and morphologically altered vasculature. See generally Folkman J. 2007. Angiogenosis: an organizing principle for drug discovery?. Nat Rev Drug 6:273-286 and Baluk P, Hashizume H, McDonald DM. 2005. Cellular abnormalities of blood vessels as targets in cancer. Curr Opin Genot Dev. 15:102-111. [000194] As mentioned above, neovascularization (NV) is a normal process that occurs in both development and wound healing but can become pathological when angiogenesis is dysregulated and occurs in tumor-associated tissues ( cancer), corneal to vascular or subretinal space (wet AMD). The proliferation, invasion and migration of NV vessels is controlled by a complex interplay between growth factors, vascular endothelial cells, extracellular matrix molecules, chemokines and cell signaling molecules. [000195] NV tissue is composed of endothelial cells (EC), pericytes, and inflammatory cells (eg, macrophages). Pericytes are derived by mast cell differentiation. The process of neovascularization first comprises the formation of angiogenic sprouts composed of EC from existing capillaries in the avascular space. VEGF signaling is understood to be the primary change for this NV process. In this regard, VEGF has been localized to tip cell filopodia leading to angiogenic sprouting. [000196] After bud formation, the newly formed vessels are lined by pericytes, which leads to VN maturation. Pericyte lining of the NV leads to stabilization and support of the NV both physically and through signaling, including pericyte production of VEGF Armulik A, Abramsson A, BetsholtzC. 2005. Endothelial / Pericyte Interactions. Turn Res. 97:512-523. [000197] The approved therapies for wet AMD are all directed at the suppression of VEGF signaling. These therapies include pegaptanib (Macugen™), approved in 2004, Genentech's Bevacizumab (Avastina™), approved in 2004 for cancer, used without a prescription for AMD, Genentech's ranibizumab (Lucentis™), approved in 2006, and Regeneren's aflibercept (Eylea™) approved in 2011. Pegaptanib is an aptamer-based therapeutic, but with a limited market compared to protein-based therapeutics perhaps due to limited gains in visual acuity by patients. Bevacizumab is an IgG1 anti-VEGFA antibody approved for cancer treatment, but is widely used without prescription for the treatment of AMD. Ranibizumab is a Fab that is affinity matured from Bevacizumab and has been approved for AMD. However, the market for [Ranibizumab is substantially skewed by the use of the much cheaper Bevacizumab. Finally, aflibercept is a VEGF trap, employing a soluble receptor fragment decoy. [000198] Anti-VEGF monotherapy has not led to disease-modifying regression of pathologic VN. Brown DM, Kaiser PK, Michels M, et al. 2006. ANCHOR Study Group. Ranibizumab versus verteporfin for neovascular age-related macular degeneration N Engl J Med 355(14): 1432-1444; Rosenfeld PJ, Brown DM, Heier JS, et al. 2006. MARINA study group. Ranibizumab for neovascular age-related macular degeneration N Engl J Med 355(14):1419-1431; Regillo CD, Brown DM, Abraham P, et al. 2008. Randomized, noi i ηη / ι znz / R / v double-mask, sham-controlled trial of ranibizumab for neovascular age-related macular degeneration: PIER study year 1. Am J Ophthalmol. 145:239-248. Instead, most of the therapeutic benefit or efficiency of anti-VEGF therapies is due to their anti-permeability property. Zebrowski BK, Yano S. Liu W, et al. 1999. Vascular endothelial growth factor levels and induction of permeability in malignant pleural effusions. Clin Cancer Res 5:3364-3368. [000199] Because conventional anti-VEGF therapies do not cause pathologic NV regression, visual acuity gains for many patients have been severely limited. However, neovasculature can also lead to subretinal fibrosis which is a cause of blindness in patients with wet AMD. [000200] Subretinal scarring develops in almost half of the eyes treated within two years of anti-VEGF therapy. Daniel E. Toth CA, Grunwald JE. 2014. Scarring risk in comparison of age-related macular degeneration in clinical settings Retina 32:1480-1485. Formation of subretinal fibrosis can lead to permanent dysfunction of the macular system; causes destruction of photoreceptors, retinal pigment epithelium, and colloidal vessels Ishikawa K, Ram K, Hinton DR. 2015, Molecular mechanisms of subretinal fibrosis in age-related macular degeneration Eye Res. Mar 13, 2015 Epub 1-7. Although anti-VEGF therapy generally better stabilizes visual acuity, scarring has been identified as one of the causes of visual acuity loss after treatment Cohen SY, Oubraham H. Uzzan J, et al. 2012. Causes of unsuccessful treatment of subsequent outcome in oozing age-related macular degeneration in clinical settings. Retina 32:1480-1485. [000201] In general, pro-angiogenic factors are favored for expression in pathological angiogenesis, including two members of the vascular endothelial growth factor (VEGF) family: VEGF-A and placental growth factor (PGF). ). VEGF-A and PGF activate quiescent endothelial cells, promote cell proliferation and vascular permeability. VEGF-A has been identified as a major factor in vascular leakage in wet AMD. Dvorak HF, Nagy JA, Feng D, Brown LF, Dvorak AM. 1999. Vascular permeability factor / vascular endothelial growth factor and the significance of microvascular hyperpermeability in angiogenesis Curr Top Microbiol Immunol. 237:97-132. [000202] The platelet-derived growth factor "PDGF" plays an important role in the maturation of NV and in particular in the lining of NV by pericytes. The lining of NV endothelial cells by pericytes begins with EC expression of the paracrine platelet-derived growth factor B, which forms the PDGF-BB homodimer. PDGF-BB is highly retained in the tip cells of angiogenic sprouts by heparin-sulfate-proteoglycan. This PDGF-BB is then recognized by the pericyte-bound receptor PDGFR-β, which initiates pericyte migration proliferation along with increasing neovascularization. [000203] PDGF-DD has also been found to play a central role in pathological angiogenesis Kumar A, Hou X, Chunsik L, et al. 2010. Targeting of platelet-derived growth factor-DD having pathological angiogenesis by modulating glycogen synthase-kinase-3β phosphorylation J Biol Chem 285(20):15500-15510. PDGF-DD overexpression induces blood vessel maturation during angiogenesis Kong D. Wang Z, noi i ηη / ι znz / R / v Sarkar FH, et al. 2008. Overexpression of platelet-derived growth factor-D in transitional epithelial-mesenchymal PC3 prostate cancer cells. Stem Cells 26:1425-1435. PDGF-DD is highly expressed in eye Ray S. Gao C, Wyatt K, et al. 2005. Platelet-derived growth factor D, tissue-specific expression in the eye and key role in control of lens epithelial cell proliferation J Biol Chem. 280:8494-8502. Kumar et al. (2010) found that PDGF-DD expression is favored during pathological angiogenesis and that inhibition of PDGF-DD signaling decreased retinal and choroidal neovascularization. [000204] The term "PDGF" as used herein means any member of the growth factor class that (i) binds to a PDGF receptor such as PDGFR-β, or PDGFR-α; (ii) activates a PDGF receptor-associated tyrosine kinase activity; and (II) thereby affects angiogenesis or an angiogenic process. The term "PDGF" refers generally to those members of the class of growth factors that induce DNA synthesis and mitogenesis through the binding and activation of a cell surface receptor for platelet-derived growth factor (i.e. say PDGFR) in a sensitive cell type. PDGFs effect specific biological effects including, for example: directed cell migration (chemotaxis) and cell activation; phospholipase activation; increased production of phosphatidylinositol and increased metabolism of prostaglandins; stimulation of both collagen and collagenase synthesis by responsive cells; alteration of cellular metabolic activities, including matrix synthesis, cytokine production, and lipoprotein uptake; induction, indirectly, of a proliferative response in cells lacking PDGF receptors; fibrosis and potent vasoconstrictor activity. The term "PDGF" is intended to include both a "PDGF" peptide and its corresponding nucleic acid or gene encoding "PDGF". [000205] The PDGF family consists of disulfide-linked homo dimers of PDGF-A (Swiss Protein P04085), -Β (P01127), -C (Q9NRA1) and -D (Q9GZP0) and the PDGF-AB heterodimer. The various PDGF isoforms exert their effect by binding to β- and β-tyrosine kinase receptors (PDGFR-α (P16234) and PDGFR-β (P09619) respectively). See generally US Patent No. 5,872,218 which is incorporated herein by reference for all purposes. The α and β receptors are structurally similar: both have extracellular domains with five immunoglobulin (Ig)-like domains and intracellular domains with a kinase function. PDGF binding occurs primarily through domains 2 and 3 of the receptors and causes receptor dimerization. Ig-like domain 4 is involved in receptor dimerization. Receptor dimerization is a key component of PDGF signaling: receptor dimerization leads to receptor autophosphorylation. Auto-phosphorylation in turn causes a conformational change in the result and removes the kinase from the receptor. PDGF-A, -B, -C and -D bind to the two different receptors with different affinities and effects. PDGF-AA, -AB, -BB and -CC induce αα receptor homodimers, PDGF-BB and -DD induced ββ homodimers and PDGF-AB, -BB, -CC and -DD produce αβ receptor homodimers. [000206] In terms of function, PDGFR-α and PDGFR-β appear to have substantially different roles. PDGFR-α signaling is involved in gratulation and in the development of cranial and cardiac neural crests, gonads, gut, skin, CNS, and skeleton. PDGFR-β signaling is involved in the noi i ηη / ι znz / B / v formation of blood vessels and early hematopoiesis. Andrae J, Radiosa G, Betsholtz C. 2008. Role of platelet-derived growth factors in physiology and medicine. Genes Develop 22:1276-1312. In terms of the interaction of the various PDGF ligands with receptors, PDGF-AA and PDGF-CC exclusively bind to and interact with PDGFR-α. PDGF-BB and PDGF-AB bind to α and β receptors. PDGF-DD interacts exclusively with PDGFRβ. Raica M, Cimpean AM. 2010. Platelet-Derived Growth Factor (PDGF) / PDGF Receptors (PDGFR) Axis as Target for Antitumor and Antiangiogenic Therapy. Pharmaceut. 3:572-599. [000207] Unless otherwise apparent from the context, reference to a PDGF means any of PDGF-A, -B, -C and -D in any of the natural isoforms or natural variants or induced variants having at least 90, 95, 98 or 99% sequence identity to a natural form. Preferably, these PDGFs are human PDGFs. Likewise, reference to a PDGFR means PDGFR-A (P16234) or PDGF-B including any natural isoform or natural variant, or an induced variant that has at least 90, 95, 98, or 99% or 100% sequence identity to a natural sequence. [000208] The amino acid sequence of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1) is set forth in Figure 1, which shows a full length human PDGFR-β (including leader sequence), a protein of 1106 amino acids. Amino acids 1-32 are part of the guide peptide that is cleaved in the mature protein. PDGFR-β has five Ig-like extracellular domains D1-D5. Williams AF, Barclay AN. 1988. The immunoglobulin superfamily—domains for cell surface recognition. Annu Rev Immunol. 6:381-405. The full-length extracellular region runs from approximately amino acid 33 to 532, the transmembrane domain from approximately residue 533 to 553, and the cytoplasmic domain is from approximately residue 554 to 1106. The extracellular region includes five immunoglobulin-like domains, D1-D5 . The D1 domain is typically considered to be from about amino acid 33 (Leu) to about 123 (Pro). According to the present invention, D1 can also be from 33 to 122 (Val). Typically, D2 is considered to be from about 124 (Thr) to about 213 (Ser). According to the present invention, D2 can be 129 (Pro) to 210 (Gln). Typically, D3 is considered to be from about amino acid 214 (lie) to 314 (Gly). According to the present invention, D3 can be from 214 (lie) to 309 (Thr). D4 is typically considered to be from about amino acid 315 (Tyr) to 416 (Pro). D5 is typically considered to be from about amino acid 417 (Val) to 531 (Lys). [000209] The exact boundaries of the D1-D5 domains may vary depending on how the analysis is performed. Preferably, the limits vary by 9 amino acids or less. Typically, they vary by 7 or fewer amino acids, more typically by 5 or fewer amino acids. Usually, the cutoff variance is 3 amino acids or less. More typically, the limits vary by only one amino acid. The essential feature of each domain is its ability to bind to its cognate ligands. [000210] A "PDGF antagonist" or a molecule that "PDGF antagonist" is an agent that reduces, or inhibits, either partially or completely, at least one activity of a PDGF including its ability to specifically bind to a PDGFR, and consequent cellular responses, such as proliferation. PDGF antagonists include antibodies that specifically bind to a PDGF or PDGFR and noi i ηη / ι znz / R / v trap segments from an extracellular PDGFR. [000211] One or more portions of a PDGFR-β extracellular receptor sequence can be used as an antagonist for PDGF-PDGFR-β signaling. The term "extracellular trap segment" refers to a full-length extracellular region or any portion thereof, or combination of portions from different PDGF receptors that can recognize PDGF-PDGFR-β signaling. These portions are typically used free of the PDGFR transmembrane and intracellular sequence and are therefore referred to as being soluble. The portions antagonize by acting as a trap or lure for a cognate PDGF. PDGF binds to the soluble trap segment of PDGFR-β and is unable to bind to the corresponding membrane-bound receptor. Preferably, these traps include one or more of the PDGFR-β domains D1-D5. Preferably, the trap contains at least one of D2 and D3. More preferably, the trap contains D1, D2 and D3. More preferably, the trap is a contiguous segment corresponding to amino acids 33 to 314 of Figure 8A-8B that contains D1-D3. Likewise PDGFR-alpha induces domains D1 through D5 and extracellular trap segments incorporating the corresponding domains of PDGFR-β can equally be used in place of PDGFR-β. [000212] Antibodies can also be used as PDGFR-β antagonists, including antibodies that bind to the receptor (for example 2A1E2 [US Patent No. 7,060,271]: HuM4Ts.22 [US Patent No. 5,882,644] : or 1B3 or 2C5 [US Patent No. 7,740,850]), and anti-PDGF antibodies such as anti-PDGF BB, anti-PDGF-DD, anti-PDGF-BB and anti-PDGF-AB. [000213] "VEGF" or "vascular endothelial growth factor" is a human vascular endothelial growth factor that affects angiogenesis or an angiogenic process. In particular, the term "VEGF" means any member of the class of growth factors that (i) binds to a VEGF receptor such as VEGFR-1 (Flt-1), VEGFR2 (KDR / Flk-1), or VEGFR- 3 (FLT-4); (ii) activates a VEGF receptor-associated tyrosine kinase activity; and (iii) thereby affects angiogenesis or an angiogenic process. [000214] The VEGF family of factors is made up of five related glycoproteins: VEGF-A also known as VPE), -B, -C, -D and PGF (placental growth factor). Of these, VEGF-A is the best studied and is the target of anti-angiogenic therapy. Ferrara et al, (2003) Nat. Med. 9:669-676. VEGF-A exists as several different isotypes that are generated by both alternative splicing and proteolysis. VEGF-A206, VEGF-Aiss, VEGFAi65, and VEGF-A121. The isoforms differ in their ability to bind heparin and non-signaling binding proteins called neurophilins. The isoforms are all biologically active as dimers. [000215] The various effects of VEGF were measured by the binding of a VEGF, eg, VEGF-A, (P15692), B(P49766), -C (P49767) and -D (Q43915), to the tyrosine kinase of the receiver (RTK). The VEGF family receptors correspond to the V RTK class and each have seven Ig-like domains in the extracellular domain (ECD). In humans, VEGF binds to three RTK isotypes: VEGFR-1 (Flt-1) (P17948), VEGFR-2 (KDR, Flk-1) (P935968), and VEGFR-3 (Flt-4) (P35916). A sequence of VEGFR-1 is shown in Figure 2. Unless otherwise apparent from the context, reference to a VEGF means any of VEGF-A, -B, -C, -D, and PGF, in any of the naturally occurring isoforms or naturally occurring variants or induced variants having at least 90, 95, 98 or 99% or 100% sequence identity to noi 1 ηη / ι znz / R / v a natural form. Preferably, these VEGFs are human VEGFs. Likewise, reference to a VEGFR means any one of VEGFR-1, R-2 or R-3, including any natural isoform or wild-type vanant, or an induced vanant that is at least 90, 95, 98 or 99% or 100% identical. sequence identity to a natural sequence. [000216] The extracellular region runs from about amino acid 27-758, the transmembrane domain from about amino acid 759 to 780, and the extracellular region from about 781-1338. The extracellular region includes seven immunoglobulin-like domains, D1-D7. VEGFR-1 domain 1 is 32 (P) to 128 (I), domain 2 is 134 (P) to 125 (Q), domain 3 is 232 (V) to 331 (K), domain 4 is 333 (F) to 428 (P), domain 5 from 431 (Y) to 553 (T), domain 6 558 (G) to 656 (R), and domain 7 from 662 (Y) to 751 (T) . See generally US Patent No. 8,273,353, incorporated herein by reference for all purposes. The exact boundaries of the VEGFR-1 D1-D7 domains can vary depending on how the analysis is done. Preferably, the boundaries vary by 9 or fewer amino acids, typically by 7 or fewer amino acids, more typically by 5 or fewer amino acids. Usually, the cutoff variance is 3 amino acids or less. More typically, the limits vary by only one amino acid. [000217] The protein sequence of VEGFR-2 is shown below in Figure 3. [000218] The extracellular region runs from approximately residues 20-764, the transmembrane domain from approximately residues 765-785, and the intracellular domain from approximately residues 786-1356. The extracellular region includes seven immunoglobulin-like domains, D1-D7. VEGFR-2 domain 1 is from 32(P) to 118(V), domain 2 is from 124(P) to 200(G), domain 3 is from 226(V) to 327(K), domain 4 is from 329 (F) to 421 (P), domain 5 is from 424 (G) to 548 (T), domain 6 is from 553 (I) to 662 (L), and domain 7 is from 668(T) to 757(A). In general see US Patent No. 8,273,353, incorporated herein by reference for all purposes. The exact boundaries of the VEGFR-2 D1-D7 domains may vary depending on how the analysis is performed. Preferably, the limits vary by 9 amino acids or less. Typically, they vary by 7 or fewer amino acids, more typically by 5 or fewer amino acids. Usually, the cutoff variation is by 3 amino acids or less. Typically, the limits vary by only one amino acid. [000219] The protein sequence of VEGFR-3 is shown below in Figure 4. The extracellular region runs from residues 25-775, the transmembrane domain from approximately residues 776-796, and the intracellular domain from approximately residues 797 -1363. The extracellular domain includes seven immunoglobulin-like domains, D1-D7. VEGFR-3 domain 1 is from 30(P) to 132(V), domain 2 is from 138(P) to 226(G), domain 3 is from 232(I) to 330(N), domain 4 is from 332 (F) to 422 (P), domain 5 is from 425 (H) to 552 (T), domain 6 is from 557 (G)l to 673 (Q), and domain 7 is from 679(R) to 768(S). See generally US Patent No. 8,273,353, incorporated herein by reference for all purposes. The exact boundaries of the VEGFR-3 domains D1-D7 may vary depending on how the analysis is performed. Preferably, the limits vary by 9 amino acids or less. Typically, they vary by 7 or fewer amino acids, more typically by 5 or fewer amino acids. Usually, the cutoff variation is by 3 amino acids or less. More typically, the limits vary by only one amino acid. noi i ηη / ι znz / R / v [000220] VEGFR-2 is expressed predominantly on vascular endothelial cells. VEGFR-1 is also expressed in the vascular endothelium, but in addition it is also expressed by several other cell types: neutrophils, monocytes, macrophages, wall cells, and endothelial progenitor cells. VEGFR-1 has a higher affinity for VEGF-A but VEGFR-2 does not. However, when VEGFR-1 binds to VEGF-A on endothelial cells, VEGFR1 exhibits only very weak tyrosine phosphorylation. Therefore, it is believed that the effects of VEGF-A (including its various isoforms) on the vascular endothelium were mediated by the binding of VEGF-A to VEGFR-2. [000221] PGF and VEGF-B bind only to VEGFR-1. PGF and VEGF-B have been implicated in pathogenic vascular remodeling. Carmeliet P, Moons L, Lutten A, et al. 2001. Synergism between vascular endothelial growth factor and placental growth factor contributes to angiogenesis and plasma extravasation in pathological conditions. Nat Med. 7(5):575-583. VEGF-C and -VEGF-D bind with high affinity to VEGFR-3, which is found primarily on lymphatic endothelial cells in the adult. VEGF-C and -D are thought to play a role with respect to lymphagiogenesis. [000222] A "VEGF antagonist" or a "VEGF antagonizing" molecule is an agent that reduces, or inhibits, either partially or completely, an activity of a VEGF including its ability to specifically bind to its receptor a VEGFR and consequent cellular responses, such as angiogenesis and cell proliferation. VEGF antagonists include antibodies that specifically bind to a VEGF or a VEGFR or a VEGFR extracellular trap segment. [000223] The term "extracellular trap segment" refers to a full-length extracellular region or any portion thereof, or combination of portions from different VEGFR receptors that can antagonize signaling between at least one VEGF and VEGFR. Preferably, the extracellular trap segment includes at least one domain from one of VEGFR-1, -2 or -3 defined above, and more preferably at least two contiguous domains, such as D2 and D3. Optionally, an extracellular domain includes at least one domain as defined above from at least two different VEGFRs. A preferred extracellular domain comprises or consists essentially of VEGFR-1 D2 and VEGFR-2 D3. [000224] VEGF antagonist therapies have been approved for the treatment of certain cancers and wet AMD. Bevacizumab (AVASTIN™, Genentech / Roche) is a humanized mouse monoclonal antibody that binds to and neutralizes human VEGF, particulate to all VEGF-A isoforms and to bioactive proteolytic fragments of VEGF-A. See, for example, Ferrara N, Hillan KJ, Gerber HP, Novotny W. 2004. Discovery and development of bevacizumab, an antl-VEGF antibody for treating cancer. Nat Rev Drug Discov. 3(5):391-400. Bevacizumab has been tested for the treatment of certain cancers. The protein sequence of the heavy and light chains of Bevacizumab (DrugBank DB00112) are shown below in Figure 5 with the CDRs underlined (see also SEQ ID Nos.: 2 and 5). [000225] Bevacizumab variable light chain CDRs are CDR11: SASQDISNILN, CDR12: FTSSLHS, and CDR13: QQYSTVPWT. Bevacizumab variable heavy chain CDRs are CDRh1: GYTFTNYGMN, CDRh2: WINTITGEPTIAADFKR, and CDRH3: YPHYYGSSHWYFDV. CDRs are defined by Kabat except CDRH1 is the Kabat / Chothia composite definition. noi i ηη / ι znz / E / v [000226] Another anti-VEGF molecule, derived from the same mouse monoclonal antibody as Bevacizumab has been tested as a treatment for wet AMD: ranibizumab (LUCENTIS™, Genentech / Roche). Ranibizumab is an antibody fragment or Fab. Ranibizumab was produced by affinity maturation of Bevacizumab variable heavy and light chains. The sequence of the ranibizumab heavy and light chains is shown below (as published by Novartis) in Figure 6 (see also SEQ ID Nos.: 12 and 13). [000227] Ranibizumab variable light chain CDRs are CDR11: SASQDISNILN, CDR12: FTSSLHS, and CDR13: QQYSTVPWT. The variable heavy chain CDRs for ranibizumab are CDRh1: GYDFTIGMN, CDRh2: WINTITGEPTIAADFKR, and CDRH3: YPIYYGTSHWYFDV. [000228] Antibodies that compete with Bevacizumab for binding to VEGF-A or binding to the same epitope on VEGF-A as Bevacizumab can also be used. [000229] Another anti-VEGF therapy is a VEGF trap. For example, aflibercept (EyleaMRRegeneren), consists of the second Ig-like domain of VEGFR-1 and the third Ig-like domain of VEGFR-2 expressed as a fusion in line with the constant (Fe) region of human IgG1. PapadopoulosN, et al. 2012. Binding and neutralization of vascular endothelial growth factor (VEGF) and related ligands by VEGF Trap, ranibizumab and bevacizumab. Angiogenesis 15:171-185. In theory, aflibercept binds only VEGF-A, but also VEGF-B and PGF, thus antagonizing their interaction with VEGFR-1. [000230] In accordance with the present invention, there is provided a dual VEGF / PDGF antagonist comprising a VEGF antagonist linked to a PDGF antagonist. The linkage preferably includes a fusion of protein strands to form a hybrid strand made up of components from both antagonists. Alternatively, the components can be joined by chemical crosslinking. As an example, of fusion linkage, if the dual antagonist is formed from an antibody and an extracellular trap segment, then either a heavy or light chain of the antibody can be fused to the extracellular trap segment. Preferably, the extracellular trap segment is fused directly or indirectly via a linker to the L-terminus of the heavy or light chain of the antibody. Any chain that does not fuse to the extracellular trap segment can be associated with the chain that is in a manner similar to the light-heavy chain association in a natural antibody. For example, an exemplary format has an extracellular trap segment fused to the N-terminus of an antibody heavy chain by a linker and the antibody light chain rendered complexed with the antibody heavy chain. The antibody to that dual antagonist can be an intact antibody or any of the binding fragments described above, such as a Fab fragment. Preferably, in these dual antagonists, the VEGF antagonist is an antibody to VEGF-A, such as Bevacizumab or ranibizumab, and the PDGF antagonist is an extracellular trap segment of PDGR-1. [000231] In an alternative format, the VEGF antagonist and the PDGF antagonist are both extracellular trap segments. The two segments can be fused in any orientation with respect to each other, directly or by means of a ligator. That is, the VEGFR extracellular trap region can be attached to either the N-terminus or the C-terminus of the PDGFR extracellular trap region. The C-terminus of this fusion protein can be linked noi i ηη / ι znz / R / v to an Fe region of an antibody that forms Fe fusion proteins. [000232] In preferred embodiments, the PDGFR is PDGFR-β and the extracellular trap segment comprises one or more of the PDGFR-β domains D1-D5. More preferably, the extracellular trap segment comprises the D1-D3 domains of PDGFR-β. Even more preferably, the extracellular trap segment comprises or consists of amino acids 33 to 314 of SEQ ID No.: 11. In preferred embodiments, the VEGF antagonist is an anti-VEGF antibody, preferably an anti-VEGF-antibody. TO. [000233] In dual antagonists that have extracellular trap and antibody components fused together, the respective components, typically the antibody heavy chain and extracellular trap segment are separated by a linker sequence. The linker is preferably GGGGSGGGGS, GG, or GGGGSGGGGSGGGGSGGGGSG or an oligomer of any of these. Preferably, GGGGSGGGGS linkers. [000234] According to one aspect of the present invention, the anti-VEGF-A antibody heavy chain has at least the following CDR sequences: CDRh1: GYDFTIGMN, CDRh2: WINTITGEPTIAADFKR, and CDRh3: YPIYYGTSHWYFDV. Preferably, the anti-VEGF-A light chain has at least the following CDRs: CDR11: SASQDISNILN, CDR12: FTSSLHS and CDR13: QQYSTVPWT. In the case of the anti-VEGFA antibody heavy chain, it is preferred that its isotype be lgG1 and have a CHi, hinge, CH2 and CH3 domain. It is also preferred that the isotype of the light chain is kappa. The constant region of the lgG1 preferred sequence is set forth in SEQ ID No.: 17. The sequence of the light chain constant region is set forth preferably in SEQ ID No.: 18. [000235] The lgG1 domain of the anti-VEGF-A antibody preferably has one or more mutations to reduce or diminish effector function. Preferred amino acids for use of effector function-reducing mutations include (EU numbering) E233P, L234V, L235, G236, G237, delG236, D270A, K322A, A327G, P329A, A330, A330S, P331S, and P331A, in which the second amino acid mentioned is the mutation. Preferably, the mutations include one or more of the following: E233P, L234V, L234A, L235A, G237A, A327G, A330S and P331A (EU numbering). More preferably, the anti-VEGF-A heavy chain has the following mutations: L234A, L235A and G237A. The number of these mutations relative to a natural human IgG1 sequence is not more than 10, and preferably not more than 5, 4, 3, 2 or 1. [000236] Alternatively, the IgG domain can be IgG2, IgG3 or IgG4, preferably human IgG2, IgG3 or IgG4, or a composite product in which a constant region of more than one of these isotypes (eg example, CHi region of IgG2 or IgG4, hinge, CH2 and CH3 regions of IgG1). These domains may contain mutations to reduce effector function at one or more of the EU positions mentioned for lgG1. Human IgG2 and IgG4 have reduced effector functions relative to human IgG1 and IgG3. [000237] The anti-VEGF-A heavy chain may also contain an added cysteine ​​residue as a mutation by recombinant DNA technology that can be used to conjugate a half-life extension moiety. Preferably, the mutation is (EU numbering) Q347C and / or L443C. More preferably, the L443C. Preferably, the stoichiometry of the dual antagonist to polymer is 1:1; in other words, noi 1 ηη / ι znz / B / v a conjugate consists essentially of molecules each comprising a dual antagonist molecule conjugated to a polymer molecule. [000238] A preferred dual antagonist that includes an antibody to VEGF-A and a PDGFR extracellular trap segment comprises a fusion protein of the antibody heavy chain and the PDGFR extracellular trap segment having the amino acid sequence of SEQ ID No.: 9 and the antibody light chain having the amino acid sequence of SEQ ID No.: 10, or variants thereof including sequences that each differ from SEQ ID No.: 9 and 10 by no more than 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 amino acids. [000239] In another aspect of the present invention, a dual VEGF / PDGF antagonist is provided having a PDGF antagonist constituting one or more segments of a PDGFR as described above and a VEGF antagonist constituting an anti Fab fragment. -VEGF. For this aspect of the present invention, the PDGFR extracellular trap comprises one or more of the PDGFR-β domains D1-D5 More preferably, the PDGFR trap comprises the PDGFR-β domains D1-D3. More preferably, the PDGFR trap is amino acids 33 to 314 of SEQ ID No.: 11. [000240] The PDGFR trap is preferably located C-terminus of the Fab heavy or light chain. The PDGFR trap is also preferentially located N-terminus of the VEGF heavy or light chain. Preferably, the dual antagonist includes an anti-VEGF-A Fab fragment heavy chain fused via a linker to a PDGFR extracellular trap segment and an anti-VEGF-A light chain. [000241] In another aspect of the invention, a dual VEGF / PDGF antagonist is provided wherein the extracellular trap segment binds one or more of PDGF-AA, PDGF-BB, PDGF-AB, PDGF-CC, and PDGF -DD. Preferably, the extracellular trap binds PDGF-AB, PDGF-BB, and PDGF-DD. Even more preferably, the extracellular trap inhibits PDGF-AB, PDGF-BB, and PDGF-DD from binding to any of the PDGFR-αα, PDGFR-αβ, and PDGFR-ββ receptors. [000242] A linker is preferentially located between the PDGFR trap and the anti-VEGF Fab fragment heavy chain. Preferably, the linker is selected from the group consisting of GGGGSGGGGS, GG, and GGGGSGGGGSGGGGSGGGGSG, and oligomers of any of these. More preferably, the linker is GGGGSGGGGS. [000243] The anti-VEGF Fab fragment heavy chain preferably has at least the following CDRs: CDRs: CDRh1: GYDFTIGMN, CDRH2: WINTITGEPTIAADFKR, and CDRH3: YPIYYGTSHWYFDV. The antiVEGF-A light chain preferably has at least the following CDRs: CDR11: SASQDISNILN, CDR12: FTSSLHS and CDR13: QQYSTVPWT. [000244] A preferred anti-VEGF Fab fragment heavy chain isotype is lgG1 and comprises a CHi domain and the light chain isotype is kappa. [000245] The dual VEGF / PDGF antagonist may have an attached half-life extension moiety. Preferably, the half-life extension portion is a zwitterionic polymer but PEG or other half-life extenders discussed below may alternatively be used. More preferably, the zwitterionic noi i nn / ι 7Π7 / Ε / Υ polymer is formed from monomers having a phosphorylcholine group. Preferably, the monomer is 2(acr¡lo¡lox¡et¡l)-2'-(tnmethylammon¡omethyl)phosphate. More preferably, the monomer is 2-(methacryloyloxyethyl)-2'(trmethyllamonmethyl)phosphate (HEMA-PC). [000246] A dual antagonist conjugated polymer preferably has at least 2 and more preferably 3 or more arms. Some polymers have 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12 arms. Even more preferably, the polymer has 3, 6 or 9 arms. More preferably, the polymer has 9 arms. Preferably, the peak molecular weight of the polymer is between 300,000 and 1,750,000 Da. More preferably, the polymer has a peak molecular weight between 500,000 and 1,000,000 Da. Even more preferably, the polymer has a peak molecular weight between 600,000 to 800,000 Da. [000247] The polymer can be covalently attached to the dual antagonist by conjugation. Preferably, the polymer is conjugated to the VEGF / PDGF dual antagonist via a group such as an amino group, hydroxyl group, sulfhydryl group, or carboxyl group. The sulfhydryl group may be from a naturally occurring cysteine ​​residue. The sulfhydryl group may also be from a cysteine ​​residue added by recombinant DNA technology. [000248] In a preferred aspect of the present invention, the polymer is conjugated to the cysteine ​​residue at position 741 of SEQ ID No.: 9 or aligned position of any variant of SEQ ID No.: 9 discussed herein. [000249] In another aspect of the present invention, a dual VEGF / PDGF antagonist having a VEGFR trap containing one or more extracellular segments of a VEGFR, such as VEGFR-1, VEGFR-2 or VEGFR-3, fused to an anti-PDGF antibody or Fab fragment light or heavy chain and an anti-PDGF antibody or Fab fragment light or heavy chain not included in the fusion. [000250] According to one aspect of the present invention, the extracellular segment of VEGFR is preferably one or more of the D1-D7 domains. More preferably, the extracellular segment comprises VEGFR-1 D2 and VEGFR-2 D3. Even more preferably, D2 is N-terminal to D3 and further comprises a linker between the domains. [000251] In preferred embodiments of this aspect of the present invention, the PDGF antagonist is an antibody. More preferably, the antibody is selected from the group consisting of humanized 2A1E2, humanized HuM4Ts.22.1B3, humanized 2C5, anti-PDGF-BB, anti-PDGF-DD, anti-PDGFR-BB, and anti-PDGF-AB . The PDGF antagonist is also preferably a Fab fragment. [000252] According to this aspect of the present invention, the antibody heavy chain is preferably IgG1, more preferably human IgG1 and the light chain is preferably kappa, human kappa. The heavy chain may have a cysteine ​​added by recombinant DNA technology. Preferably, the cysteine ​​is selected from the group consisting of Q347C and L443C. Preferably, there is a half-life extension moiety conjugated to cysteine. [000253] Preferably, the half-life extension moiety is a zwitterionic polymer having one or more monomer units and wherein at least one monomer unit has a zwitterionic group. Preferably noi i ηη / ι 7Π7 / Ε / Υ, the zwitterionic group is phosphorylcholine. The monomer is preferably 2-(acryloyloxyethyl)-2,(trimethylammoniomethyl)phosphate. More preferably, the monomer is 2-(methacr¡lo¡lox¡et¡l)-2'(trimethylammon¡ometh¡l)phosphate (HEMA-PC). [000254] According to this aspect of the present invention, the polymer preferably has at least 2 and more preferably 3 or more arms. Some polymers have 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12 arms. Even more preferably, the first one has 3, 6 or 9 arms. More preferably, the polymer has 9 arms. According to one aspect of the present invention, the peak molecular weight of the polymer is between 300,000 and 1,750,000 Da. More preferably, the polymer has a peak molecular weight between 500,000 and 1,000,000 Da. Even more preferably, the polymer has a peak molecular weight between 600,000 to 800,000 Da. [000255] According to one aspect of the present invention, the polymer is covalently attached to the polymer by conjugation. Preferably, the polymer is conjugated to the dual VEGF / PDGF antagonist via a group selected from the group consisting of an amino group, a hydroxyl group, a sulfhydryl group, and a carboxyl group. Preferably, the sulfhydryl group is from a naturally occurring cysteine ​​residue. In other preferred embodiments, the sulfhydryl group is from a cysteine ​​residue added by recombinant DNA technology. [000256] In preferred aspects of the present invention, the PDGF trap-VEGF trap is conjugated to a half-life extension moiety as discussed with other dual antagonists. [000257] Preferably, the half-life extension moiety is a zwitterionic polymer having one or more monomer units and wherein at least one monomer unit has a zwitterionic group. Preferably, the zwitterionic group is phosphorylcholine. The monomer is preferably 2-(acryloyloxyethyl)-2'(trimethylammoniomethylphosphate. More preferably, the monomer is 2-(methacryloyloxyethyl)-2'(trimethylammoniummethylphosphate (HEMA-PC) . [000258] According to this aspect of the present invention, the polymer preferably has at least 2 and more preferably 3 or more arms. Some polymers have 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12 arms. Even more preferably the polymer has 3, 6, or 9 arms. More preferably, the polymer has 9 arms. According to one aspect of the present invention, the peak molecular weight of the polymer is between 300,000 and 1,750,000 Da. More preferably, the polymer has a peak molecular weight between 500,000 and 1,000,000 Da. Even more preferably, the polymer has a peak molecular weight between 600,000 to 800,000 Da. [000259] According to one aspect of the present invention, the polymer is covalently attached to the polymer by conjugation. Preferably, the polymer is conjugated to the VEGF / PDGF dual antagonist via a group such as an amino group, hydroxyl group, sulfhydryl group, or carboxyl group. In some conjugates, the sulfhydryl group is from a naturally occurring cysteine ​​residue. In some conjugates, the sulfhydryl group is from a cysteine ​​residue added by recombinant DNA technology. [000260] Dual VEGF / PDGF antagonists can be produced by recombinant expression including (i) production of recombinant DNA by genetic engineering, (ii) introduction of recombinant DNA into noi i ηη / ι znz / E / v cells prokaryotic or eukaryotic cells by, for example and without limitation, transfection, electroporation or microinjection, (iii) culturing the transformed cells, (iv) expressing dual antagonists, for example, constitutively or inducingly, and (v) isolating the antagonist. dual, for example, the culture medium or by harvesting the transformed cells, in order to (vi) obtain the purified dual antagonist. [000261] Dual antagonists can be produced by expression in a suitable prokaryotic or eukaryotic host system characterized by producing a pharmacologically acceptable dual antagonist molecule. Examples of eukaryotic cells are mammalian cells, such as CHO, COS, HEK 293, BHK, SK-Hip, and HepG2. Other suitable expression systems are prokaryotic cells (eg E. coli with pET / BL21 expression system), yeast (Saccharomyces cerevisiae and / or Pichia pastoris systems) and insectiles. [000262] A wide variety of vectors can be used for preparation of the dual antagonist and are selected from eukaryotic and prokaryotic expression vectors. Examples of vectors for prokaryotic expression include plasmids such as, and not limited to, preset, pet, and pad. Wherein the promoters used in the prokaryotic expression vectors include one or more of, and without limitation, lac, trc, trp, recA, or araBAD. Examples of vectors for eukaryotic expression include: (i) for yeast expression, vectors such as, and not limited to, pAO, pPIC, pYES, or pMET, using promoters such as, and not limited to, AOXI1, GAP, GAL1, or AUG1 : (ii) for expression in insectile cells, vectors such as and without limitation, PMT, pAc5, pIB, pMIB, or pBAC using promoters such as and without limitation PH, p10, MT, Ac5, OplE2, gp64, or polh, and (iii) for expression in mammalian cells, vectors such as, and without limitation, pSVL, pCMV, pRc / RSV, pcDNA3, or pBPV, and vectors derived from, in one aspect, viral systems such as, and without limitation, vaccinia virus. , adeno-associated virus, herpes virus, or retrovirus, using promoters such as and without limitation CMV, SV40, EF-1, UbC, RSV, ADV, BPV, and beta-actin. [000263] The half-life of the dual antagonists can be prolonged by the attachment of a "half-life extension moiety" or "half-life extension groups", terms used interchangeably herein to refer to one or more chemical groups attached to one or more amino acid site chain functionalities such as -SH, -OH, -COOH, -CONH2, -NH2, or one or more N- and / or O-glycan structures and may increase the in vivo circulating half-life of proteins / peptides when conjugated to these proteins / peptides. Examples of half-life extension moieties include polymers described herein, particularly those of zwitterionic monomers, such as HEMA-phosphorylcholine, PEG, biocompatible fatty acids and derivatives thereof, hydroxy-alkyl-starch (HAS), for example, hydroxy-ethyl-starch (HES), polyethylene glycol (PEG), poly (Glix-Seri) (HAP), hyaluronic acid (HA), heparosan polymers (HEP), fleximers, dextran, polysialic acids (PSA), domains Fe, transferrin, albumin, elastin-like peptides (ELP), XTEN polymers, PAS polymers, PA polymers, albumin-binding polymers, CTP peptides, FcRn-binding polymers, and any combination of these. [000264] In one embodiment, a half-life extension moiety can be conjugated to a dual antagonist via free amino groups of the protein using N-hydroxysuccinimide (NHS) esters. Reagents targeting conjugation to amine groups can randomly react to the ε-amine group of lysines, α-amine group of N-terminal amino acids, and δ-amine group of histidines. noi i ηη / ι znz / R / v [000265] However, the present dual antagonists have many amine groups available for polymer conjugation. Conjugation of polymers to free amino groups, in this way, can negatively impact the ability of dual antagonist proteins to bind VEGF and / or PDGF. [000266] In another embodiment, a half-life extension moiety is coupled to one or more free SH groups using any appropriate thiol-reactive chemistry including, without limitation, maleimide chemistry, or the coupling of polymer hydrazides or polymer amines. to carbohydrate portions of the dual antagonist after the above oxidation. The use of maleimide coupling is a particularly preferred embodiment of the present invention. Coupling occurs preferentially on cysteines naturally present or introduced by genetic engineering. [000267] The polymers preferentially bind covalently to introduced cysteine ​​residues in the dual antagonist by site-directed mutagenesis. It is particularly preferred to employ cysteine ​​residues in the Fe portion of the dual antagonist. For preferred sites for introducing cysteine ​​residues into an Fe region see WO 2013 / 093809, US 7,521,541, WO 2008 / 020827, US 8,008,453, US 8,455,622 and US2012 / 0213705, incorporated herein by reference for all purposes. Particularly preferred cysteine ​​mutations are Q347C and L443C which refer to the human IgG heavy chain by EU numbering. [000268] The invention provides dual antagonist conjugates and high MW polymers that serve as half-life extenders. A preferred conjugate comprises a dual antagonist that is coupled to a zwitterionic polymer where the polymer is formed from one or more monomer units and where at least one monomer unit has a zwitterionic group. Preferably, the zwitterionic group is phosphorylcholine. [000269] Preferably, one of the monomer units is 2-(acryloyloxyethyl)-2'-(trimethylammon¡omethyl) phosphate or 2-(methachnloyloxyethyl)-2'-(trimethylammon¡omethyl) l) phosphate (HEMA-PC). In other preferred embodiments, the polymer is synthesized from a single monomer that is preferably 2-(acryloyloxyethyl)-2'-(tnmethylammoniomethyl) phosphate or 2-(methacryloyloxyethyl) )-2'-(trimethylammoniomethyl) phosphate. [000270] Some dual antagonist conjugates have 2 or more preferably 3 or more polymer arms where the monomer is HEMA-PC. More preferably, the conjugates have 2, 3, 4, 5, 6, 7,8, 9,10,11 or 12 polymer arms where the monomer is HEMA-PC. More preferably, the conjugates have 3, 6 or 9 arms. More preferably, the conjugate has 9 arms. [000271] Polymer-dual antagonist conjugates preferably have a polymer moiety with a molecular weight between 100,000 and 1,500,000 Da. More preferably, the conjugate has a polymer moiety with a molecular weight between 500,000 and 1,000,000 Da. Even more preferably, the conjugate has a polymer moiety with a molecular weight between 600,000 to 800,000 Da. More preferably, the conjugate has a polymer moiety with a molecular weight between 600,000 to 850,000 Da and has 9 arms. When a molecular weight is given for a dual VEGF / PDGF antagonist conjugated to a polymer, the molecular weight will be the addition of the molecular weight of the protein, including any carbohydrate moieties associated with it, and the molecular weight of the polymer. noi i ηη / ι znz / R / v [000272] According to one aspect of the present invention, a dual VEGF / PDGF antagonist having a HEMA-PC polymer having a molecular weight as measured by Mw of between about 100 kDa and 1500 kDa. More preferably, the molecular weight of the polymer as measured by Mw is between about 500 kDa and 1000 kDa. More preferably, the molecular weight of the polymer as measured by Mw is between about 600 kDa to about 900 kDa. More preferably, the molecular weight of the polymer portion as measured by Mw is 750 kDa plus or minus 15%. [000273] In this aspect of the present invention, the polymer is preferably produced from a suitable initiator for ATRP having one or more polymer initiation sites. Preferably, the polymer initiation site has a 2-bromoisobutyrate site. Preferably, the initiator has 3 or more polymer initiation sites. More preferably, the initiator has 3, 4, 5, 6,7, 8,9,10,11 or 12 polymer initiation sites. More preferably, the initiator has 3, 6, or 9 polymer initiation sites. Even more preferably, the initiator has 9 polymer initiation sites. More preferably, the primer OG1786. [000274] The invention provides methods for synthesizing a zwitterionic polymer-dual antagonist conjugate, the conjugate having one or more functional agents and one or more polymer arms wherein each of the polymer arms has one or more monomer units where at least one of the units has a zwitterion. The method can have the steps of: to. providing an initiator having one or more sites for monomer polymerization and a first linker having an amine group wherein the initiator is a salt of trifluoroacetic acid; b. providing one or more monomers suitable for polymerization wherein at least one of the monomers is zwitterionic; either. reacting the monomers with the initiator to form one or more polymer arms each corresponding to sites for monomer polymerization to provide an initiator-polymer conjugate having the first linker with the amine group; d. providing a second linker having at least second and third reactive groups; and. coupling one of the second and third reactive groups of the second linker to the amine group of the first linker of the initiator-polymer conjugate to provide a linker-initiator-polymer conjugate having one or more reactive groups that were not used in the coupling step; and F. coupling one or more functional agents to one or more of the unreacted reactive groups of the linker-initiator-polymer portion to provide the polymer-functional agent conjugate. [000275] Prior to the present invention, the initiator entity or molecule had contained a deprotectable functional group and would allow coupling of the functional agent. An example of this primer having a protected maleimide is shown below: noi i ηη / ι znz / R / v noi i nn / ι zoz / r / yiai [000276] After polymer synthesis, the protected maleimide is deprotected with heat to allow generation of maleimide that can be used to couple the functional agent. If it is desired to vary the nature of the chemical entity between the maleimide and the polymer initiation site, a complete new initiator would have to be synthesized. [000277] Each time the primer is changed or altered in any way, a new upscale synthesis procedure would have to be developed. Each change in the nature of the initiator molecule can have a wide variety of effects on the synthesis of the polymer. However, according to the present invention, a method is presented whereby the conjugation group (eg, maleimide) is added after the synthesis of the polymer. This is sometimes referred to as a "snap approach" or "universal polymer approach." A single portion of primer can be used for large-scale polymer and bio-conjugate development and discovery. In this way, conditions can be developed to scale up the optimal synthesis of the polymer. This polymer can then be adapted to various types of functional agents by "pressure" from various types of linkers and functional conjugation chemistries. [000278] For example, if it is desired to conjugate a larger functional agent or polymer of the present invention, such as an antibody to even a Fab fragment, a longer linker sequence can be crimped to the polymer. In contrast, smaller functional agents may require relatively shorter linker sequences. [000279] In preferred embodiments of the methods, the initiator has 1, 2,3,4, 5,6, 7,8,9,10,11 or 12 sites for polymerization initiation. Preferably, the initiator has 3, 6 or 9 polymer initiation sites. [000280] According to one aspect of the present invention, a second linker has second, third, fourth, fifth, and sixth reactive groups. More preferably, a second linker has only second and third reactive groups. [000281] In accordance with one aspect of the present invention, each polymer arm has from about 20 to about 2000 monomer units. Preferably, each arm has from about 100 to 500 monomer units or from about 500 to 1000 monomer units or from about 1000 to 1500 monomer units or about 1500 to 2000 monomer units. [000282] According to one aspect of the present invention, the peak molecular weight of the functional-polymer conjugate is about 100,000 to 1,500,000 Da. Preferably, the peak molecular weight of the polymer-functional agent conjugate is from about 200,000 to about 300,000 Da, from about 400,000 to about 600,000 Da, or from about 650,000 to about 850,000 Da. [000283] According to another aspect of the present invention, the first linker is preferably alkyl, substituted alkyl, alkylene, alkoxy, carboxyalkyl, haloalkyl, cycloalkyl, cyclic alkyl ether, alkenyl, alkenylene, alkynyl, alkynylene, cycloalkylene, heterocycloalkyl, heterocycloalkylene , aryl, arylene, arylene-oxy, heteroaryl, amino, amido, or any combination thereof. Preferably, the first linker has the formula: noi i ηη / ι znz / R / v where m is from 1 to 10, more preferably, the first linker has the above formula and m is 4. [000284] In still other aspects of the present invention, the initiator preferably includes a structure selected from the group consisting of: where X is selected from the group consisting of NCS, F, C1, Bre I. More preferably, X is Br. [000285] In preferred embodiments of the present invention, the monomer is selected from the group consisting of where R7 is H or Ci-6 alkyl and t is from 1 to 6. [000286] Preferably, the monomer is selected from the group consisting of 2-(methacrylo¡lox¡et¡l)-2'(trimethylammoniomethyl) phosphate (HEMA-PC) and 2-(acrylo¡lox¡et ¡l)-2'-(trimethylammon¡ometh¡l) phosphate. [000287] More preferably, the monomer is 2-(methacryloyloxyethyl)-2'-(trmethyllamonmethyl)phosphate. [000288] The second linker portion preferably comprises an activated ester having the structure where R8 is selected from the group consisting of where p is from 1 to 12. [000289] In more preferred embodiments of the present invention, the polymer has 9 arms, m of R2 is 2-4, R9 is and p is from 4 to 15. Even more preferably, m is 4 and p is 12. [000290] When a polymer is to be conjugated via a cisternae (or other specified residue), the polymer can be linked directly or indirectly to the residue (eg, with an intervening primer, and / or spacer or the like). [000291] Dual antagonists can be incorporated into a pharmaceutical composition with a pharmaceutically acceptable excipient. Pharmaceutical compositions adapted for oral administration may be presented as discrete units such as capsules, as solutions, syrups, or suspensions (in aqueous or non-aqueous liquids; or as foams or edible desserts; or as emulsions). Suitable excipients for hard gelatin tablets or capsules include lactose, corn starch or derivatives thereof, stearic acid or salts thereof. Excipients suitable for use with soft gelatin capsules include for example vegetable oils, waxes, fats, liquid or semi-solid polyols, etc. For the preparation of solutions and syrups, excipients that can be used include, for example, water, polyols, and sugars. For the preparation of suspensions, oils (eg vegetable oils) can be used to provide oil-in-water or water-in-oil suspensions. [000292] Pharmaceutical compositions can be adapted for salt administration where the excipient is a solid including a coarse powder having a particle size for example in the range of 20 to 500 microns which is administered in a manner in which a sniff is taken, ie, by rapid inhalation through the nasal passage from a container of the powder held close to the nose. Suitable compositions where the excipient is a liquid, for administration as a nasal spray or as nasal drops, include aqueous or oily solutions of the active ingredient. Pharmaceutical compositions adapted for administration by inhalation include fine particle dusts or mists which can be generated by means of various types of pressurized aerosols, nebulizers or metered dose insufflators. [000293] Pharmaceutical compositions adapted for parenteral dual antagonist include aqueous and non-aqueous sterile injection solution which may contain dissolved antioxidants, buffers, bacteriostats that render the formulation substantially isotonic with the blood of the intended recipient; and aqueous and non-aqueous sterile suspensions that may include functional agents and thickening agents. Excipients that can be used for injectable solutions include water, alcohols, polyols, glycerin, and vegetable oils, by way of example. The compositions can be presented in unit-dose or multi-dose containers, for example closed vials and ampoules, and can be stored in a freeze-dried (lyophilized) condition requiring only the addition of the sterile transported liquid, for example water for injections. , immediately before use. Suspensions and solutions for temporary injections can be prepared from sterile powders, granules, and tablets. The pharmaceutical compositions can be substantially isotonic, implying an osmolarity of approximately 250-400 mOsm / kg of water. [000294] Pharmaceutical compositions may contain preservatives, solubilizing agents, stabilizing agents, wetting agents, emulsifiers, sweeteners, colorants, odorants, salts (the substances of the present invention may be provided by themselves in the form of a pharmaceutically acceptable salt ), buffers, coating agents or antioxidants. They may also contain therapeutically active agents in addition to the substance of the present invention. The pharmaceutical compositions of the noi i ηη / ι znz / R / v invention can be used in combination with one or more pharmaceutically acceptable excipients. These excipients may include, but are not limited to, saline, buffered saline (such as phosphate buffered saline), dextrose, liposomes, water, glycerol, ethanol, and combinations thereof. [000295] The dual antagonists and pharmaceutical compositions containing them can be administered in an effective regimen for treatment or prophylaxis of a patient's disease including, for example, administration by oral, intravitrial, intravenous, subcutaneous, intramuscular, intraosseous, intranasal, topical, intraperitoneal and intralesional. Parenteral infusions include intramuscular, intravenous, intraarterial, intraperitoneal, or subcutaneous administration or routes in between. In therapy or as a prophylactic, the active agent may be administered to an individual as an injectable composition, for example as a sterile aqueous dispersion, preferably isotonic or substantially isotonic. [000296] For administration to mammals, and particularly humans, the dose of the active agent is expected to be 0.01 mg / kg of body weight, typically about 1 mg / kg. The practitioner can determine the most suitable actual dose for an individual depending on factors including the individual's age, weight, gender and response, the disease or disorder being treated, and the age and condition of the individual being treated. The above dosages are exemplary of the average case. Of course, there may be cases where higher or lower doses are worthwhile. [000297] This dose can be repeated as often as appropriate (eg, weekly, bi-weekly, monthly, quarterly). If side effects develop, the amount and / or frequency of the dose may be reduced, in accordance with normal clinical practice. In one embodiment, the pharmaceutical composition can be administered once every thirty days. [000298] The dual antagonists and pharmaceutical compositions of the invention may be employed alone or in conjunction with other compounds, such as therapeutic compounds or therapeutic molecules, eg, anti-inflammatory drugs, analgesics, or antibiotics. This administration with other compounds can be simultaneous, separate or sequential. The components can be prepared in the form of a kit that can comprise instructions as appropriate. [000299] The dual antagonists and pharmaceutical compositions described herein can be used for the treatment or prophylaxis of disease, particularly the ocular diseases or conditions described herein. Although both modalities of antagonist within the dual antagonist are believed to contribute to efficiency as discussed above and shown in Example 40, an understanding of the mechanism is not required for practice of the invention. Preferably, a dual antagonist is more effective than an equimolar concentration of each antagonist administered alone, or a 1:1 combination of the antagonists administered as separate molecules. [000300] Used in this way, the conjugates are typically formulated for, and administered by, ocular, infraocular, and / or intravitrial injection, and / or juxtascleral injection, and / or subretinal injection, and / or subtendon injection. , and / or super-choroidal injection and / or topical administration in the form of eye drops and / or ointment. These dual antagonists and compositions can be delivered by a variety of methods, for example, intravitrially as noi i ηη / ι znz / R / v a device and / or a reservoir that allows slow release of the compound into the glass, including those described in references such as Intraocular Drug Delivery, Jaffe, Ashton, and Pearson, editors, Tailor & Francis (March 2006). In one example, a device may be in the form of a minipump and / or a matrix and / or a passive diffusion system and / or encapsulated cells that release the compound over an extended period of time (Infraocular Drug Delivery, Jaffe, Ashton , and Pearson, editors, Tailor & Francis (March 2006). [000301] Formulations for ocular, infraocular or intravitrial administration may be prepared by methods and using ingredients known in the art. A primary requirement for efficient treatment is proper penetration through the eye. Unlike diseases of the front of the eye, where drugs can be delivered topically, retinal diseases require a more site-specific approach. Eye drops and ointments rarely penetrate the back of the eye, and the blood-ocular barrier prevents systemically administered drugs from penetrating ocular tissue. Therefore, usually the method of choice for drug delivery to treat retinal disease, such as AMD and CNV, is direct intravitrial injection. Intravitrial injections are usually repeated at intervals that depend on the patient's condition, and the properties and half-life of the administered drug. [000302] The related conjugate and therapeutic dual antagonists according to the present invention are generally placed in a container having a sterile access port, for example, an intravenous solution bottle or bag having a stopper pierced by a needle. hypodermic injection. These compositions can also be supplied in the form of pre-filled syringes. [000303] A "stable formulation" is one in which the protein or protein conjugated to a polymer of another half-life extension portion therein essentially retains its physical and / or chemical stability and / or biological activity on storage. By "stable" is also meant a formulation that exhibits little or no signs of instability, including aggregation and / or deamidation. For example, according to one aspect of the present invention, the formulations provided by the present invention may remain stable for at least two years, when stored as indicated at a temperature of 5-8°C. [000304] Various analytical techniques for measuring protein stability are available in the art and are reviewed in Peptide and Protein Drug Delivery, 247-301 (Vincent Lee ed., New York, N.Y., 1991) and Jones, 1993 Adv. Drug Delivery Rev. 10: 29-90, for examples. Stability can be measured at a selected temperature for a selected period of time. Storage of stable formulations is preferably for at least 6 months, more preferably 12 months, more preferably 12-18 months, and most preferably 2 or more years. [000305] A protein, such as an antibody or fragment thereof, retains its physical stability" in a pharmaceutical formulation if it does not show signs of aggregation, precipitation, deamidation and / or denaturation on visual examination for color and / or clarity, or as measured by UV light scattering or by size exclusion chromatography. [000306] A protein retains its "chemical stability" in a pharmaceutical formulation, if the chemical stability at a given time is such that the protein is considered to still retain its biological activity. Chemical stability can be assessed by detecting and quantifying chemically altered forms of the protein. Chemical alteration may comprise size modification (e.g., clipping), which can be assessed using size exclusion chromatography, SDS-PAGE, and / or matrix-assisted laser desorption ionization / time-of-flight mass spectrometry (MALDI). / TOF MS), for examples. Other types of chemical alteration include charge alteration (eg occurring as a result of deamidation), which can be assessed by ion exchange chromatography, for example. An antibody "retains its biological activity" in a pharmaceutical formulation if the biological activity of the antibody at a given time is within approximately 10% (within assay errors) of the biological activity exhibited at the time the antibody was prepared. pharmaceutical formulation as determined for example in an antigen binding assay. [000307] A protein-polymer binding conjugate "retains its chemical stability" if the chemical bond between the protein and the polymer remains intact, eg, is not hydrolyzed or otherwise disrupted. The protein part of the conjugate retains its chemical stability as described above. [000308] By "isotonic" it is meant that the formulation of interest has essentially the same osmotic pressure as human blood or vitreous humor for intravitrial injections. In general, isotonic formulations will have an osmotic pressure of approximately 250 to 400 mOsm. Isotonicity can be measured using an ice-freeze or vapor pressure type osmometer, as an example. [000309] As used herein, "buffer" refers to a buffered solution that resists changes in pH by the action of its acid-base conjugate components. The buffer of this invention has a pH in the range preferably from about 3.0 to about 8.0; for example from about 4.5 to 8; or from about pH 6 to about 7.5; or about 6.0 to about 7.0, or about 6.5-7.0, or about pH 7.0 to about 7.5; or about 7.1 to about 7.4. A pH anywhere between the above ranges is also contemplated. [000310] "PBS" phosphate buffered saline, Tris-based buffers and histidine-based buffers are particularly preferred buffers for the currently invented dual antagonists. In the case of OG1448, PBS is particularly preferred. More particularly, in the case of OG1448, the PBS buffer has a pH of 7-8 and the concentration of OG1448 is from about 10 mg / ml to about 100 mg / ml. Even more preferably, the OG1448 is from about 25 to about 65 mg / ml and the pH is about 7.4. In the most preferred embodiments of the present invention, the concentration of OG1448 is from 50 mg / ml to 60 mg / ml. [000311] In preferred embodiments of the present invention, the PBS buffer consists of at least Na2HPO4, KH2PO4, and NaCI adjusted to provide the appropriate pH. In particularly preferred embodiments of the present invention, the buffer may contain other pharmaceutical excipients such as KCI and other salts, detergents and / or preservatives to provide a storage stable solution. noi i ηη / ι znz / R / v [000312] A "preservative" is a compound that can be included in the formulation to essentially reduce bacterial action therein, thereby facilitating the production of a multi-use formulation, by way of example. Examples of potential preservatives include octadecyldimethylbenzyl ammonium chloride, hexamethonium chloride, benzalkonium chloride (a mixture of alkylbenzyldimethylammonium chlorides in which the alkyl groups are long chain compounds), and benzethonium chloride. Other types of preservatives include aromatic alcohols such as phenol, benzyl butyl alcohol, alkyl parabens such as methyl- or propyl-paraben, catechol, resorcinol, cyclohexanol, 3-pentanol, and m-cresol. [000313] According to one aspect of the present invention, formulations of the dual PDGF / VEGF antagonists according to the present invention are safe for human use or for animal testing must have sufficiently low levels of endotoxin. "Endotoxiná" is lipopolysaccharide (LPS) derived from the cell membrane of gram-negative bacteria. Endotoxin is composed of a hydrophilic polysaccharide moiety covalently linked to a hydrophobic lipid moiety (lipid A). Raetz CR, Ulevitch RJ, Wright SD, Sibley OH, Ding A, Nathan CF. 1991. Gram-negative endotoxin: an extraordinary lipid with profound effects on eukaryotic signal transduction. FASEB J. 5(12):2652-2660. Lipid A is responsible for most of the biological activities of endotoxin, that is, its toxicity. Endotoxins are secreted in large quantities at the death of bacterial cells as well as during growth and division. They are highly heat stable and do not destroy under regular sterilization conditions. Extreme treatments with heat or pH, for example 180-250°C and more than 0.1 M acid base should be used (Petsch D, Anspach F. 2000. Endotoxin removal from protein Solutions. J Biotechnol. 76: 97-119) . Of course, these conditions will be highly detrimental to biologic drugs. [000314] In the biotechnology and pharmaceutical industries, it is possible to find the endotoxin during both production processes in final products. Since bacteria can grow in nutrient-poor media, including water, saline, and buffers, endotoxins are prevalent unless precautions are taken. Injection of endotoxin into an animal or human causes a wide variety of pathophysiological effects, including endotoxin shock, tissue injury, or even death. Ogikubo Y, Ogikubo Y, Norimatsu M, Noda K, Takahashi J, Inotsume M, Tsuchiya M, Tamura Y. 2004. Evaluation of the bacterial endotoxin test for quantifications of endotoxin contamination of porcine vaccines. Biologics 32:88-93. [000315] Pyrogenic reactions and shock are induced in mammals by intravenous injection of endotoxin at low concentrations (1 ng / mL) (Fiske JM, Ross A, VanDerMeid RK, McMichael JC, Arumugham. 2001. Method for reducing endotoxin in Moraxella catarrhalis UspA2 protein preparations. J Chrom B. 753:269-278). The maximum endotoxin level for intravenous applications of pharmaceuticals and biologics is set to 5 endotoxin units (EU) per kg body weight per hour by all pharmacopoeias (Daneshiam M, Guenther A, Wendel A, Hartung T, Von Aulock S 2006. In vitro pyrogen test for toxic or immunomodulatory drugs. J Immunol Method 313:169175). EU is a measure of the biological activity of an endotoxin. For example, 100 pg of the standard endotoxin EC-5 and 120 pg of the endotoxin from Escherichia coli0111:B4 have activity of one EU (Hirayama C, Sakata M. 2002. Chromatographic removal of endotoxin from protein Solutions by polymer partitions. J Chrom B 781:419-432). The noi i ηη / ι znz / R / v compliance with this threshold level has always been a challenge in the pharmaceutical industry and in biological research (Berthold W, Walter J. 1994. Protein Purification: Aspects of Processes for Pharmaceutical Products. Biologicals 22:135-150, Petsch D, Anspach FB. 2000. Endotoxin removal from protein Solutions. J Biotech 76:97-119). [000316] The presence of endotoxin in drugs to be administered by intravenous injection is of particular interest. Intravitrial injection of drug (penicillin) was first performed in 1945 by Rycroft. Rycroft BW. 1945. Penicillin and the control of deep intra-ocular infection. British J Ophthalmol 29(2): 57-87. The vitreous humor is a chamber where a high level of drug can be introduced and maintained for relatively long periods of time. The concentration of drug that can be achieved by intravitrial injection greatly exceeds that which can be generated by topical administration or by systemic (eg, intravenous) administration. [000317] One of the most dangerous complications potentially arising from intravitrial injections is endophthalmitis. Endophthalmitis falls into two classes: infectious and sterile. Infectious endophthalmitis is usually caused by bacteria, fungi, or parasites. Symptoms of infective endophthalmitis include severe pain, loss of vision, and redness of the conjunctiva and underlying episclera. Infectious endophthalmitis requires urgent diagnosis and treatment. Possible treatments include intravitrial injection of antibiotics and pars plana vitrectomy in some cases. Eluculation may be required to remove a painful blind eye. See, for example, Christy NE, Sommer A. 1979. Antibiotic prophylaxis of postoperative endophthalmitis. Ann Ophthalmol 11(8): 1261-1265. [000318] Sterile endophthalmitis in contrast does not comprise an infectious agent and can be defined as acute infraocular inflammation of the vitreous cavity that resolves without the need for intravitrial antibiotics and / or vitreo-retinal surgery. If a microbiological study of the vitreous humor has been done, it needs to be proven negative culture to support a diagnosis of sterile endophthalmitis. Marticorena J, Romano V, Gomez-Ulla F. 2012 “Sterile Endophthalmitis after Intravitrial Injections” Med Inflam. 928123. [000319] It has been observed that intravitrial injection of endotoxin-contaminated biologic drugs can result in sterile endophthalmitis. Marticorena, et al. Bevacizumab (Avastin) is approved by the US Food and Drug Administration for the treatment of glioblastoma and metastatic colorectal cancer, advanced non-squamous non-small cell lung cancer, and metastatic kidney cancer. Over-the-counter bevacizumab is also widely used as a treatment for wet AMD. Bevacizumab comes from the manufacturer as 100 mg / 4 ml. This solution cannot be used directly for intravitrial injection and must be mixed by a pharmacist. Clusters of sterile endophthalmitis have been observed and are theorized to be caused by inadvertent endotoxin contamination of Bevacizumab by the combining pharmacist. [000320] Given the appalling clinical results of intravitrial injection of endotoxin, the total amount of endotoxin that can be given to a patient via intravitrial dosing is highly limited. According to one aspect of the present invention, there is provided a solution having a dual VEGF / PDGF antagonist according to the present invention having an endotoxin level not exceeding 5.0 EU / ml. More preferably, the endotoxin level does not exceed 1.0 EU / ml. Even more preferably, the endotoxin level does not exceed 0.5 EU / ml. Even more preferably, the endotoxin level does not exceed 0.2 EU / ml. In even more preferred modalities, the endotoxin level does not exceed 0.1, 0.09, 0.08, 0.07, 0.06, 0.05, 0.04, 0.03, 0.02 or 0.01 EU / ml. [000321] Two commonly used EDA-approved tests for the presence of endotoxin are the rabbit pyrogen test and the Limululs Amoebodyte (LAL) Assay (Hoffman S, et al. 2005. International validation of novel pyrogen tests based on human monocytoid cells J. Immunol. Methods 298:161-173; Ding JL, Ho BA. 2001. New era in pyrogen testing. Biotech. 19:277-281). The rabbit pyrogen test was developed in the 1920s and involves monitoring the temperature rise in a rabbit injected with a test solution. However, the use of the rabbit pyrogen test has declined considerably over the years due to expense and long lead times. Much more common is the LAL test. LAL is derived from the blood of a horseshoe crab and coagulates on exposure to endotoxin. [000322] One of the simplest LAL assays is the LAL gel-clot assay. Essentially, the LAL coagulation assay is combined with a serial dilution of the sample in question. The formation of the gel is to provide the amount of endotoxin in the sample. Serial dilutions of the sample are prepared and each dilution is assessed for its ability to form the LAL gel. At some point, a negative reaction is contained. The amount of endotoxin in the original sample can be estimated by the dilution assay. [000323] Other LAL tests have also been developed, including the turbidimetric LAL test (Ong KG, Leían JM, Zeng KF, Barrett G, Aourob M, Grimes CA. 2006. A rapid highly-sensitive endotoxin detection system. Blosensors and Bioelectronics 21:2270-2274) and the chromogenic LAL assay (Haishima Y, Hasegawa C, Yagami T, Tsuchiya T, Matsuda R, Hayashi Y. 2003. Estimation of uncertainty in kinetic-colorimetric assay of bacterial endotoxins. J Pharm Biomed Analysis 32:495-503). Turbidimetric and chromogenic assays are much more sensitive and quantitative than the simple gel-clot dilution assay. [000324] The present invention provides a method for reducing the amount of endotoxin in a composition having a dual VEGF / PDGF antagonist, the method having the steps of contacting the composition with an affinity chromatography resin that binds the dual VEGF / PDGF antagonist; diluting the VEGF / PDGF dual antagonist from the affinity chromatography resin to form an affinity chromatography eluent having the antagonist; contacting the affinity chromatography eluent with an ion exchange resin that binds the dual VEGF / PDGF antagonist; and eluting the VEGF / PDGF dual antagonist from the ion exchange resin, wherein the VEGF / PDGF dual antagonist eluted on the ion exchange resin is substantially free of endotoxin. [000325] The above method for reducing the amount of endotoxin, or another method or process cited herein, can be performed in the order described in the previous steps or can optionally be performed by varying the order of the steps or even repeating one or more of the steps. In one embodiment, the method of reducing the amount of endotoxin in a composition is performed in the order of the steps described. In some embodiments, the affinity chromatography resin contacting, washing, and elution steps are repeated in the same order more than once before contacting the affinity chromatography diluent with the ion exchange resin. . The method may also include a filtration step using, for example, a 0.1 micron, 0.22 micron, or 0.44 micron filter, which can be performed on either one or more of the eluents removed. after each resin bonding step. [000326] In certain cases, the steps of contacting the composition with affinity chromatography resin, washing, and eluting the antibody from the affinity chromatography resin can be repeated more than once before contacting the first eluent with an ion exchange resin. In one embodiment, the affinity chromatography resin comprises a recombinant protein A ("rProteinA") resin. An example of a suitable recombinant protein A resin is rProteinA Sepharose FFMR resin (Amersham, Piscataway, N.J.). In another embodiment, a suitable affinity chromatography resin will comprise a protein G chromatography resin. In other embodiments, a suitable affinity chromatography resin comprises a mixed protein A / protein G resin. Affinity chromatography comprises a hydrophobic charge induction resin comprising a 4-mercaptoethylpyridine ligand such as a MEP HyperCel™ resin (BioSepra, Cergy, Saint Christophe, France). [000327] In some embodiments, it is preferred that the ion exchange resin comprises an anion exchange resin. As will be known to the person skilled in the art, ion exchangers can be based on various materials with respect to the matrix as well as the attached charged groups. For example, the following matrices can be used, in which the mentioned materials can be more or less crosslinked: MacroCap Q (GE Healthcare Biosciences, Piscataway, NJ), agarose-based (such as Sepharose CL-6BMR, Sepharose Fast FlowMRand Sepharose High Performance™), cellulose-based (such as DAEA Sephacel™), dextran-based (such as Sephadex™), silica-based, and synthetic polymer-based. For the ion exchange resin, the charged groups, which are covalently attached to the matrix, can be, for example, diethylaminoethyl, tertiary aminoethyl and / or quaternary ammonium. It is preferred that the anion exchange resin comprises a quaternary amine group. An exemplary anion exchange resin having a quaternary amine group for binding to the anti-MCSF antibody is a Q Sepharose™ resin (Amerscham, Piscataway, N.J.) [000328] In other aspects, if endotoxin levels are higher than desired after subjecting the composition to the aforementioned anion exchange chromatography step, the composition may alternatively be subjected to a cation exchange resin. In accordance with this aspect of the present invention, any endotoxin in the composition must have differential binding to the ion exchange resin such that the protein in question allows purification of the protein from the endotoxin. In this regard, the endotoxin is negatively charged and will generally bind to an anion exchange resin. If both the protein and the endotoxin bind to the ion exchange resin, purification of one another can be accomplished by using a salt gradient to elute the two in different fractions. The relative binding of the protein to a particular resin can also be effected by changing the pH of the buffer relative to the pI of the protein. In a preferred aspect of the present invention, cation exchange chromatography is the only ion exchange chromatography employed. [000329] According to another aspect of the present invention, if the endotoxin levels are too high noi i ηη / ι znz / R / v after the anion exchange resin, the composition can be further subjected to a second step of ion exchange, for example, by contacting the compositions with a cation exchange resin and followed by a washing step, then elution from the ion exchange resin. In preferred embodiments, the cation exchange resin comprises a sulfonic group for attachment. Exemplary cation exchange resins are SP SepharoseMRFE resin (Amersham, Piscataway, N.J.) (Poros XS (CEX) (Life) Technology, Grand Island, New York). [000330] According to one aspect of the invention, after the PDGF / VEGF dual antagonist solution is produced having the specified level of endotoxin, there are several steps before final formulation of the protein. In some embodiments of the present invention, a half-life extension moiety is conjugated to the protein. The conjugate is then formulated into a final drug formulation that is injected into patients. In some embodiments, the conjugate is further purified on an ion exchange resin which may preferably be a cation exchange resin. In other embodiments, the protein is formulated. In all cases, normal laboratory procedures should be used to prevent the introduction of endotoxin contaminants into the protein sample or protein-polymer conjugate. examples [000331] Example 1.- Protein sequence of PDGFR3-GS10-anti-VEGF-A light chain / antiVEGF-A heavy chain (wild type Fe). [000332] A PDGFR-β-anti-VEGF-A light chain / anti-VEGFA heavy chain trap chain was constructed having the sequence set forth below in Figures 7A, 7B. Amino acids 1-282 of PDGFR-GS10 anti-VEGF-A light chain correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the linker sequence GGGGSGGGGS and the light chain sequence of Bevacizumab. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2 (as noted above), x3, x4, of GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs known to those skilled in the art may also be used in accordance with the present invention, including G, GG, GGGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The sequence of Figure 7A is set forth in SEQ ID No.: 19. Figure 7B shows the Bevacizumab heavy chain sequence (SEQ ID No.: 2). The Bevacizumab light chain optionally has an M4L mutation (Kabat numbering). Bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, S100aT (Kabat numbering), L234A, L235A, G237A, Q347C and L443C (EU numbering). [000333] Example 2,- Protein sequence PDGFRB-GG-anti-VEGF-A light chain / anti-VEGFA heavy chain (Wild type Fe). [000334] Another PDGFR-β-anti-VEGF-A light chain / anti-VEGF-A heavy chain trap was constructed having the sequence set forth below in Figures 8A, 8B. Figure 8A shows amino acids 1-282 corresponding to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the linker sequence noi i ηη / ι 7Π7 / Ε / Υ GG and the Bevacizumab light chain sequence. Alternatively, the linker can be the motif x1, x2, x3, x4, GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs can also be used in accordance with the present invention, including G, GG (as noted above), GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The protein sequence of Figure 8A is set forth in SEQ ID No.: 3. Figure 8B shows the heavy chain sequence of Bevacizumab (SEQ ID No.: 2). The Bevacizumab light chain of Figure 8A optionally has an M4L mutation. Bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, S100aT (Kabat numbering), L234A, L235A, G237A, Q347C and L443C (EU numbering). [000335] Example 3. Protein sequence of PDGFR3-GS10-anti-VEGF-A heavy chain (Wild-type Fe) / anti-VEGF-A light chain. [000336] Another PDGFR-p-anti-VEGF-A heavy chain (Fe wild-type) / anti-VEGF-A light chain trap was constructed having the sequence set forth in Figures 9A, 9B. Amino acids 1-282 in Figure 9A correspond to 33 to 314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the linker sequence GGGGSGGGGS and the Bevacizumab heavy chain sequence, optionally having Q347C or L443C (EU numbering). Alternatively, the linker can be the motif x1, x2 (as noted above), x3, x4, GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The protein sequence of Figure 9A is set forth in SEQ ID No.: 4. The protein of Figure 9B is the Bevacizumab light chain sequence (SEQ ID No.: 5). The Bevacizumab light chain optionally has an M4L mutation. Bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, S100aT (Kabat numbering), L234A, L235A, G237A, Q347C and L443C (EU numbering). [000337] Example 4.- Protein sequence of PDGFR3-GG-heavy chain anti-VEGF-A (Fe wild type) / light chain anti-VEGF-A. [000338] Another PDGFR-p-anti-VEGF-A heavy chain (Fe wild type) / anti-VEGF-A light chain trap was constructed having the sequence set forth below in Figures 10A, 10B. Amino acids 1-282 in Figure 10A correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the GG linker sequence and the Bevacizumab heavy chain sequence, optionally having Q347C or L443C. Alternatively, the linker can be the motif x1, x2, x3, x4 GGGGS, etc. such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG (as noted above), GGGGS and GGGES X1, x2, x3, x4, etc. The linker may be combinations of the above. The protein sequence of Figure 10A is set forth in SEQ ID No.: 6. The protein of Figure 10B is the Bevacizumab light chain sequence (SEQ ID No.: 5). The Bevacizumab light chain has an M4L mutation. The Bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, S100aT (Kabat numbering), L234A, L235A, G237A, Q347C and L443C (EU numbering). [000339] Example 5.- Anti-VEGF-A heavy chain protein sequence (Fe wild type)-GS21noi i ηη / ι 7Π7 / Ε / Υ PDGFRB / anti-VEGF-A light chain. [000340] An anti-VEGF-A antibody PDGFR-p-trap construct was constructed with the anti-VEGF-A heavy chain being 5' or N-terminal to the PDGFR-β trap having the sequence set forth below in Figures 11A, 11B. Amino acids 1-451 in Figure 11A correspond to the Bevacizumab heavy chain sequence, optionally having Q347C or L443C, followed by the linker sequence GGGGSGGGGSGGGGSGGGGSG. Alternatively, the linker can be the motif x1, x2, (as noted above), x3, x4 GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES X1, x2, x3, x4, etc. The linker may be combinations of the above. The linker is followed by the amino acid sequences 33 to 314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1). The protein sequence of Figure 11A is set forth in SEQ ID No.: 7. Figure 11B shows the Bevacizumab light chain sequence (SEQ ID No.: 5). The Bevacizumab light chain optionally has an M4L mutation. Bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, S100aT (Kabat numbering), L234A, L235A, G237A, Q347C and L443C (EU numbering). [000341] Example 6. Protein sequence of PDGFRS-GS10-anti-VEGF-A heavy chain (Q347C) / anti-VEGF-A light chain (TAF347). [000342] Another PDGFR-P-anti-VEGF-A heavy chain (Q347C) / anti-VEGFA light chain trap was constructed having the sequence set forth below in Figures 12A, 12B. Amino acids 1-282 in Figure 12A correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1). Immediately following the PDGFR sequence is a 10 amino acid linker GGGGSGGGGS. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. The linker may be combinations of the above. Attached to the carboxyl terminus of the linkers serine residue is the heavy chain of Bevacizumab with the following amino acids: T28D, N31H, H97Y, S100aT (Kabat numbering), L234A, L235A, G237A, Q347C, and L443C (EU numbering). The protein sequence of Figure 12A is set forth in SEQ ID No.: 8. The protein of Figure 12B is ranibizumab light chain (Bevacizumab w / M4L) (SEQ ID No.: 12). [000343] Example 7. Protein sequence of PDGFRB-GS1Q-anti-VEGF-A heavy chain (L433C) / anti-VEGF-A light chain. [000344] Another PDGFR^-anti-VEGF-A heavy chain (L443C)) / anti-VEGFA light chain trap was constructed having the sequence set forth below in Figures 13A, 13B. Amino acids 1-282 in Figure 13A correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1). Immediately following the PDGFR sequence is a 10 amino acid linker GGGGSGGGGS. Attached to the carboxyl terminus of the serine residue of the linker is the heavy chain of Bevacizumab with the following amino acids: T28D, N31H, H97Y, S100aT (Kabat numbering), L234A, L235A, G237A, Q347C, and L443C (EU numbering). The TAF443 light chain is the same as Bevacizumab except for an M4L change (Kabat numbering). The protein sequence of Figure 13A is set forth in SEQ ID No.: 9. Figure 13B shows the ranibizumab light chain (Bevacizumab w / M4L) (SEQ ID No.: noi i ηη / ι znz / B / v 12). [000345] Example 8. Protein sequence of PDGFRB-GS10-anti-VEGF-A light chain / anti-VEGF-A Fab. [000346] A PDGFR-B-anti-VEGF-A light chain / anti-VEGF-A Fab trap was constructed having the sequence set forth below in Figures 14A, 14B. Amino acids 1-282 in Figure 14A correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the linker sequence GGGGSGGGGS and the Bevacizumab light chain sequence. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2 (as noted above), x3, x4 GGGGS, etc., such that the activity of the two proteins is optimized. Two other linker motifs can also be used according to the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The protein sequence of Figure 14A is set forth in SEQ ID No.: 1. The protein of Figure 14B is the Bevacizumab Fab (SEQ ID No.: 21). The Bevacizumab light chain of Figure 14A optionally has an M4L mutation. The Bevacizumab Fab of the second protein optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT. The second chain Bevacizumab Fab optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CAA. Alternatively, SGGGC or CAA can be added to the C-terminus of the light chain. [000347] Example 9.- PDGFRB-GG protein sequence anti-VEGF-A light chain / anti-VEGF-A Fab [000348] A PDGFR-B-anti-VEGF-A light chain / anti-VEGF-A Fab trap was constructed having the sequence set forth below in Figures 15A, 15B. Amino acids 1-282 in Figure 15A correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the GG linker sequence and the bevacizumab light chain sequence. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2, x3, x4, GGGGS etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG (as above), GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The protein sequence of Figure 15A is set forth in SEQ ID NO. 3. Figure 15B shows the bevacizumab Fab heavy chain (SEQ ID NO. 21). The bevacizumab light chain of Figure 15A optionally has an M4L mutation (Kabat numbering). The bevacizumab Fab of Figure 15B optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT (Kabat numbering). The bevacizumab Fab of Figure 15B optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CAA. Alternatively, SGGGC or CAA can be added to the C-terminus of the light chain. [000349] Example 10. Protein sequence of PDGFRB-GS1Q-Fab anti-VEGF-A / anti-VEGF-A light chain. [000350] Another anti-VEGF-A PDGFR-B-Fab / anti-VEGF-A light chain trap was constructed having the sequence set forth below in Figures 16A, 16B. Amino acids 1-282 in Figure 16A correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the linker sequence GGGGSGGGGS noi i ηη / ι znz / E / v and the sequence of Fab of bevacizumab. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2 (as above), x3, x4, GGGGS etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The protein sequence of Figure 16A is set forth in SEQ ID NO. 22. Figure 16B shows the sequence of bevacizumab light chain (SEQ ID NO. 5). The bevacizumab light chain optionally has an M4L mutation. The bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT (Kabat numbering). The heavy chain optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CAA. Alternatively, SGGGC or CAA can be added to the C-terminus of the light chain. [000351] Example 11. Protein sequence of anti-VEGF-A PDGFRB-GG-Fab / anti-VEGF-A light chain. [000352] Another anti-VEGF-A PDGFR-p-Fab / anti-VEGF-A light chain trap was constructed having the sequence set forth below in Figures 17A, 17B. Amino acids 1-282 in Figure 17A correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the linker sequence GGGGSGGGGS and the bevacizumab Fab sequence. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2 (as above), x3, x4, GGGGS etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The protein sequence of Figure 17A is set forth in SEQ ID NO. 23. Figure 17B shows the sequence of bevacizumab light chain (SEQ ID NO. 5). The bevacizumab light chain optionally has an M4L mutation. The bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT (Kabat numbering). The bevacizumab Fab heavy chain optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CAA. Alternatively, SGGGC or CAA can be added to the C-terminus of the light chain. [000353] Example 12. Protein sequence of anti-VEGF-A-GS21-PDGFRB / anti-VEGF-A light chain Fab. [000354] A PDGFR-B-anti-VEGF-A antibody trap construct was constructed with the anti-VEGF-A heavy chain being 5' or N-terminal to the PDGFR-β trap having the sequence set forth below in Figures 18A, 18B. Amino acids 1-231 in Figure 18A correspond to the bevacizumab Fab followed by the linker sequence GGGGSGGGGSGGGGSGGGGSG. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2, x3, x4, GGGGS etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The linker is followed by the amino acid sequences 33 to 314 of human PDGFR noi i ηη / ι znz / B / v β (UniProtKB / Swiss-Prot: P09619.1). The protein sequence of Figure 18A is set forth in SEQ ID NO. 24. Figure 18B shows the sequence of bevacizumab light chain (SEQ ID NO. 5). The bevacizumab light chain optionally has an M4L mutation. The bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT. The protein of Figure 18A optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CAA. Alternatively, SGGGC or CAA can be added to the C-terminus of the light chain of Figure 18B. [000355] Example 13. Protein Sequence of PDGFR3-GS10-Fab anti-VEGF-A / anti-VEGF-A light chain. [000356] Another anti-VEGF-A PDGFR-p-Fab / anti-VEGF-A light chain trap was constructed having the sequence set forth below in Figures 19A, 19B. Amino acids 1-282 in Figure 19A correspond to 33-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1). Immediately following the PDGFR sequence is a 10 amino acid linker GGGGSGGGGS. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2 (as above), x3, x4, GGGGS etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. Attached to the carboxyl terminus of the serine residue of the linker is the bevacizumab Fab having the mutations T28D, N31H, H97Y, and S100aT (Kabat numbering). The protein sequence of Figure 19A is set forth in SEQ ID NO. 25. The protein in Figure 19B is ranibizumab light chain (bevacizumab w / M4L) (SEQ ID NO. 12). The protein of Figure 19A optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CAA. Alternatively, SGGGC or CAA can be added to the C-terminus of the light chain of Figure 19B. [000357] Example 14. Protein sequence of anti-VEGF-A PDGFR8-Fab / anti-VEGF-A light chain (1a). [000358] Another anti-VEGF-A PDGFR^-Fab / anti-VEGF-A light chain trap was constructed having the sequence set forth in Figures 20A, 20B. Amino acids 1-283 in Figure 20A correspond to 32-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the bevacizumab heavy chain. The protein sequence of Figure 20A is set forth in SEQ ID NO. 26. The protein in Figure 20B is the bevacizumab light chain sequence (SEQ ID NO. 5). As set forth in this example, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2, x3, x4, GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The bevacizumab light chain optionally has an M4L mutation. The bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, S100aT (Kabat numbering), Q347C and L443C (EU numbering). [000359] Example 15. Protein Sequence PDGFR-β (D2-D3)-anti-VEGF-A heavy chain / light chain (1b) noi i nn / ι zoz / r / yiai anti-VEGF-A. [000360] Another PDGFR-p-anti-VEGF-A heavy chain / anti-VEGFA light chain trap (D2-D3) was constructed having the sequence set forth below in Figures 21A, 21B. Amino acids 1-190 in Figure 21A correspond to 125-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the bevacizumab heavy chain. The protein sequence of Figure 21A is set forth in SEQ ID NO. 27. As set forth in this example, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2, x3, x4, GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. Figure 21B shows the sequence of bevacizumab light chain (SEQ ID NO. 5). The bevacizumab light chain optionally has an M4L mutation. The bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, S100aT (Kabat numbering), Q347C, and L443C (Eu numbering). [000361] Example 16. Protein Sequence of PDGFR-β (D2-D3)-Fab anti-VEGF-A / light chain (2b) anti-VEGF-A. [000362] Another trap (D2-D3) of anti-VEGF-A PDGFR^-Fab / anti-VEGF-A light chain was constructed having the sequence set forth in Figures 22A, 22B. Amino acids 1-190 in Figure 22A correspond to 125-314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the bevacizumab Fab. As set forth in this example, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2, x3, x4, GGGGS, etc., such that the activity of the two proteins is optimized. The linker GGGGSGGGGS is particularly preferred. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The sequence of Figure 22A is set forth in SEQ ID NO. 28. Figure 22B shows the sequence of bevacizumab light chain (SEQ ID NO. 5). The bevacizumab light chain optionally has an M4L mutation. The bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT (Kabat numbering). The bevacizumab Fab of Figure 22A optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CAA. Alternatively, SGGGC or CAA can be added to the C-terminus of the light chain of Figure 22B. [000363] Example 17. Protein sequence of anti-VEGF-A PDGFR-β (D2-D3)-Fab / anti-VEGF-A light chain (2b1). [000364] Another trap (D2-D3) of anti-VEGF-A PDGFR^-6xGS-Fab / anti-VEGF-A light chain having the sequence set forth below in Figures 23A, 23B was constructed. Amino acids 1-190 in Figure 23A correspond to 125 to 314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the linker GGGSGGGGSGGGGSGGGGSGGGGSGGGGS and then by the bevacizumab Fab. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the noi i ηη / ι 7Π7 / Β / Υ motif χ1, χ2, χ3, χ4, GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs known to those skilled in the art may also be used in accordance with the present invention, including G, GG, GGGS, and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The protein sequence of Figure 23A is set forth in SEQ ID NO. 29. Figure 23B shows the sequence of bevacizumab light chain (SEQ ID NO. 5). The bevacizumab light chain optionally has an M4L mutation. The bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT (Kabat numbering). The bevacizumab Fab heavy chain optionally has a cisternae portion added to the C-terminus to conjugate a half-life extension portion. Preferably, the cistern portion is added via SGGGC or CM. Alternatively, SGGGC or CM can be added to the C-terminus of the light chain. [000365] Example 18. Protein sequence of anti-VEGF-A PDGFR-β (D2-D3)-Fab / anti-VEGF-A light chain (2b1). [000366] Another anti-VEGF-A-6xGS-trap (D2-D3) PDGFR-p / anti-VEGF-A light chain Fab was constructed having the sequence set forth below in Figures 24A, 24B. Amino acids 1-190 in Figure 24A correspond to 125 to 314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1), followed by the linker GGGSGGGGSGGGGSGGGGSGGGGSGGGGS and then by the bevacizumab Fab. Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2, x3, x4, GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be combinations of the above. The sequence of Figure 24A is set forth in SEQ ID NO. 29. Figure 24B shows the sequence of bevacizumab light chain (SEQ ID NO. 5). The bevacizumab light chain of Figure 24B optionally has an M4L mutation. The bevacizumab heavy chain of the first protein optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT (Kabat numbering). The bevacizumab Fab of Figure 24A optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CM. Alternatively, SGGGC or CM can be added to the C-terminus of the light chain. [000367] Example 19. Anti-VEGF-A-6xGS-PDGFR-S (D2-D3) / anti-VEGF-A light chain (3) Fab Protein Sequence. [000368] Another anti-VEGF-A-6xGS-PDGFR-p(D2-D3) / anti-VEGF-A light chain Fab was constructed having the sequence set forth below in Figures 25A, 25B. Amino acids 1-231 in Figure 25A correspond to the bevacizumab Fab, followed by the linker GGGSGGGGSGGGGSGGGGSGGGGSGGGGS and then 125 to 314 of human PDGFR-β (UniProtKB / Swiss-Prot: P09619.1). Optionally, no linker need be used between the PDGFR-β segment and the anti-VEGF segment. Alternatively, the linker can be the motif x1, x2, x3, x4, GGGGS, etc., such that the activity of the two proteins is optimized. Other linker motifs may also be used in accordance with the present invention, including G, GG, GGGS and GGGES x1, x2, x3, x4, etc. The linker may be noi i ηη / ι znz / R / v combinations of the above. The sequence of Figure 25A is set forth in SEQ ID NO. 30. Figure 25B shows the sequence of bevacizumab light chain (SEQ ID NO. 5). The bevacizumab light chain optionally has an M4L mutation. The bevacizumab heavy chain optionally has one or more of the following mutations: T28D, N31H, H97Y, and S100aT (Kabat numbering). The PDGFR-β of Figure 25A optionally has a cysteine ​​moiety added to the C-terminus to conjugate a half-life extension moiety. Preferably, the cysteine ​​portion is added via SGGGC or CAA. Alternatively, SGGGC or CAA can be added to the C-terminus of the light chain. [000369] Example 20. Production of PDGFR / VEGF dual antagonist protein. [000370] TAF443 heavy and light chains were cloned into expression plasmids and transfected into CHO cells. Cells were grown in appropriate medium and harvested. TAF443 was purified as follows. 10L of culture medium for CHO cells expressing SEQ ID NOS. 31 and 32 were adjusted with 5% (v / v) 1.1 M HEPES, 0.22 M EDTA, pH 6.7 or 10% 0.55 M Hepes, 0.11M EDTA, 5.5% Triton X-100, pH 6.7, and loaded. on a 167 / 400ml protein column (2 runs) packed with Mab Select Sure resin equilibrated in 50mM Tris, 150mM NaCI, pH 7.5 (5-CV). The column was washed with 50mM Tris, 150mM NaCI, pH 7.5 (2-CV), 50mM Tris, 20.5M CaCI, pH 7.5 (5-CV), and then 10mM Tris, 10mM NaCI, pH7.5 (3-CV). before protein was eluted using 150mM Glycine, 40mM NaCI, pH 3.5 (4-CV). Fractions were pooled, adjusted to pH 3.5 using 2M Glycine, pH 2.7, and then neutralized to pH 7 using 2M HEPES, pH 8.0. The protein A mixture was loaded onto a 274ml TMAE column equilibrated in 50mM Hepes, 65mM NaCI, pH 7.0 (5-CV). The column was washed with 50 mM Hepes, 65 mM NaCI, pH 7.0 (3-CV), and then eluted with 50 mM Tris, 200 mM NaCI, pH 7.5 (5-CV). The elution fractions were pooled and buffer-exchanged on a 1150 mL Sephadex G-25 Coarse column equilibrated with PBS-CMF, pH 7.2. The mixture was filtered, concentrated to >5mg / ml via 30k MWCO VivaFlow200. The concentrated protein was filtered through a 0.22um filter, and then characterized by SDS-PAGE, analytical SEC, O.D.280 / 320, LAL endotoxin assay, Protein A ELISA, IEF, and Freeze / Thaw Analysis. noi i ηη / ι znz / R / v [000371] The table below summarizes the properties of an example batch of purified TAF443 Characteristics of Purified Lot of TAF443 Concentration (UV) 5.69 mg / ml Purity (SEC) 98.6% MW (SDS-PAGE) -200 kDa (NR) Pl (IEF) 4.2-4.5 Endotoxin (LAL) 0.1 EU / mg Protein A ( Elisa) <10 ng / ml Final Yield -700 mg / L (CM) [000372] Example 21.- Stability of the TAF bifunctional molecule at high concentration in representative formulations [000373] TAF bi-functionals were concentrated to 50-85 mg / ml in a series of standard formulation buffers ranging from pH 4.5 to 7.5, in the presence of excipients such as sucrose. Aliquots of these samples were stored at room temperature (RT) and 4°C for a period of 6 weeks, and sampled at time zero and after each subsequent week to measure the percentage of material added by analytical SEC. The effect of pH on TAF443 aggregation can be seen in the table below. noi i ηη / ι znz / R / v % Aggregates Observed in TAF Solution at various pH over Time Tris pH 7.5 His pH 6.0 His pH 5.5 Lac pH 4.5 Time 0 <1 <1 <1 <2 Day 4 <1 <1 <1 ~3 Week 1 <1 <1 <1 ~4 Week 2 <1 <1 ~2 ~6 Week 4 <1 <1 ~3 ~10 Week 6 <1 <1 ~3 ~10 [000374] Example 22.- Transfection of constructs in CHO cells [000375] DNA constructs for TAFwt, TAF443 and TAF347 were transfected into CHOK1 SV SSI:3 mixes / construct. The normal 3 weeks of recovery were observed in most cell lines. However, the TAFwt and TAF347 cell lines lagged approximately 1 week behind the other cell lines. Once the mixes were established, on day 4 for most and on day 3 for TAFwt and TAF347, samples of conditioned medium were run on Octet. Conditioned medium 3 days for TAFwt and TAF347 showed approximately 7 mg / ml by Octec. Conditioned medium 4 days showed approximately 21 mg / ml for TAF443. Small differences were observed between the mixtures and the mixtures were used to make mixtures of mixtures that are carried out for protein generation. [000376] Example 23,- SEC-MALS of Proteins [000377] The PDGFR segment of TAF has 7 putative glycosylation sites. The protein appears to be heavily glycosylated from SEC-MALS measurements: Construction Protein (kDa) Sugar (kDa) Total (kDa) TAFwt 184 63 247 TAF334 182 62 244 TAF443 187 63 250 [000378] The samples run on SEC-MALS were all greater than 98% pure. The measured molecular weights were reasonable. Some high molecular weight material was observed, probably a tri- to pentamer (data not shown). [000379] Example 24,- Thermal Stability of TAF Proteins [000380] Thermal stability profiles were run for TAFwt, TAF443 and TAF347 in PBS, pH 7.2. Each protein has three peaks (data not shown). The relative positions of the peaks are set forth in the table below: noi i ηη / ι znz / R / v Sample Tm1(°C) Tm2(°C) Tm3(°C) TAFwt 58.1 ±0.1 71.9±0.1 83.2±0.1 TAF347 58.2±0.1 71.9±0.1 81.7±0.1 TAF443 58.2±0.1 71.9±0.1 84.4±0.1 The stabilities of the proteins over the temperature range are very similar. However, it is noted that there are some small changes in Tm3. Tm3 also corresponds to the CH3 domain of the antibody domain of all three TAF proteins and the changes reflect the Cys mutations. The low overall stability of the TAF proteins is probably due to unfolding of the PDGFR segment of the proteins. [000381] Example 25 - Forced aggregation of TAF [000382] Percentage of aggregates in a solution of the three TAF proteins as a function of heat was examined (data not shown). Solutions of each of the proteins (TAFwt, TAF347 and TAF443) started to show aggregates starting around approximately 54°C. The percentage of aggregates for each of the proteins increased sharply as the temperature increased. At 64°C, approximately 40% of each of the TAF proteins constituted aggregates. It is noted that aggregation begins at the lowest Tm, apparently corresponding to the unfolding of the PDGFR portion of the protein. [000383] Example 26.- Thermal stability of TAF443 as a Function of pH [000384] The thermal stability of TAF443 was examined at various pH's as set forth in the table below. In buffer without PBS, 4 thermal peaks of denaturation were seen: Buffer Tm1(°C) Tm2(°C) Tm3(°C) Tm4(°C) Tris pH 7.5 57 67 74 85.9 His pH 7.5 53.3 62.9 75.4 84.9 Succinate 55.9 66.7 74.9 85.8 Lucentis Buffer, pH 4.8 53.9 61.8 75. BS pH 7.2 58.2 71 .9 84.4 [000385] As can be seen, there is a weak dependence on pH. Notably, the Tm2 and Tm3 (presumably CH2, Fab) domains overlap in PBS, but not in other buffers. [000386] Example 27.- Affinity of PDGF / VEGF Dual Antagonist Proteins and Conjugates to Targets [000387] Surface plasmon resonance (SPR) was used to characterize the binding kinetics of recombinant human PDGF-BB (PeproTech, 100-14B) to PDGF / VEGF dual antagonist variants TAF-WT, TAF-347, TAF- 443, TAF443-6A250K, and TAF443-3A250K. Initially, an anti-human IgG antibody (GE Healthcare, BR 1008-39) was covalently amine coupled onto all four flow cells of a CM5 carboxymethylated dextran coated sensorchip at a density of approximately 10000 resonance units (RU) following the manufacturer's protocol. Each PDGF / VEGF variant was captured at a level of approximately 150 RU. The running and sample buffer for PDGF analysis was HBS-EP + 300mM NaCI (10mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) pH 7.4, 300mM NaCI, ethylenediaminetetraacetic acid (EDTA) 3mM, 0.05% (v / v) Tween20). A 2-fold serial dilution series of PDGF-BB ranging in concentration from 1nM to 0.125nM was injected at a flow rate of 100 pL / minute for a 110 second association with dissociations ranging from 300 to 2700 seconds. The surface was then regenerated with a 30 second pulse of 23M MgCI, a 30 second pulse of an ion regeneration buffer (0.46M KSCN, 21.83M MgCI, 0.92M urea, and 1.83M guanidine-HCI, pH7.4 , Andersson et al., Analytical Chemistry, 1999) and then equilibrated with a 30 second pulse of HBS-EP+ 300mM NaCI running buffer. [000388] Similarly, SPR was used to determine the binding affinities of recombinant human VEGF121 (PeproTech, 100-20A) against the PDGF / VEGF dual antagonist variants TAF-WT, TAF-347, and TAF-443. The running and sample buffer for VEGF analysis was HBS-EP+ with a final concentration of 150mM NaCI. A 2-fold serial dilution series of VEGF121 ranging in concentration from 100nM to 12.5nM was injected at a flow rate of 50 pL / minute for approximately a 50 second association with dissociations ranging from 300 to 3600 seconds. The surface was then regenerated with a 30 second pulse of 23M MgCI, a 30 second pulse of ion regeneration buffer (0.46M KSCN, 21.83M MgCI, 0.92M urea, and 1.83M guanidine-HCI, pH7.4, Andersson et al., Analytical Chemistry, 1999) and then equilibrated with a 30 second pulse of HBS-EP + 300mM NaCI running buffer. [000389] All SPR assays were performed at 25°C with a 1Hz data collection rate using a Biacore T200 instrument (GE Healthcare). The resulting PDGF and VEGF sensorgrams were double referenced using both buffer and control surface injections. Rate constants were determined by fitting the data to a 1:1 Langmuir model with Biacore T200 v2.0 evaluation software and the equation KD=kd / ka. noi i ηη / ι znz / B / v Biacore affinity to PDGF-BB Analyte Ligand ka(1 / Ms) kd (1 / s) t1 / 2 (min) Rmax (RU) Ch¡2 / Rmax KD (pM)* PDGF-A* TAF-WT 7.97E+ 07 8.01 E-05 144.28 15.16 0.16% 0.01 PDGF-B* TAF-WT 8.01E+07 9.19E-05 125.68 15.15 0.20% 1.15 PDGF-C* TAF-WT 8.65E+07 1.03E-04 1.03E-04 111.39 111.92 151.92% Avg+ / -Standard Dev 1.1± 0.1 PDGF-A* TAF-347C 4.15E+07 8.41E-05 137.33 13.99 0.86% 2.03 PDGF-B* TAF-347C 5.82E+07 7.87E-05 146.79 13.08 1.355% 1.055% Avg+ / -Standard Dev 1.7± 0.3 PDGF-A* TAF-443C 3.22E+07 4.96E-05 233.15 13.15 0.81% 1.54 PDGF-B* TAF-443C 5.62E+07 8.76E-05 131.89 12.19 0.966% Avg+1. -Standard Dev 1.55± 0.01 PDGF-A* R3643-6A (TAF4436A250K) 7.60E+07 9.46E-05 122.09 8.11 0.41% 1.25 PDGF-B* R3643-6A (TAF4436A250K) 5.62E+073- 5.819E 5.815 % 1.04 PDGF-C* R3643-6A (TAF4436A250K) 3.48E+07 5.13E-05 225.41 8.45 0.80% 1.47 Avg+ / -Standard Dev 1.3± 0.2 PDGF-A* R3643-6A (TAF4436A250K) 5.759.2E07 05 195.04 8.15 0.31% 1.03 PDGF-B* R3643-6A (TAF4436A250K) 2.86E+07 4.96E-05 233.05 8.52 0.60% 1.73 PDGF-C* R3643-6A (TAF4436A250K) 4.71E+07 7.52E % 1.60 Avg+ / -Std. Dev 1.5± 0.4 * A, Β and C refer to separate runs or measurements on the same PDGF-BB analyte ηοι ηη / ι ζπζ / β / υ Biacore affinity to VEGF121 Analyte* Ligand ka(1 / Ms) kd (1 / s) t1 / 2 (min) Rmax (RU) Ch¡2 / Rmax KD (pM)* VEGF-A TAF-WT 1.14E+05 2.01 E-05 573.89 29.4 0.17% 176.60 VEGF-B TAF-WT 6.85E+05 5.89E-05 196.00 13.90 0.16% 86.03 PVEG-C TAF-WT 1.40E+05 2.96E-05 390.68 27.90 27.0D0 PVEG20 0.10 -WT 1.55E+05 2.69E-05 429.78 24.92 0.23% 173.00 Mean / Standard Dev 161.96 ± 53.4 6 PVEG-A TAF347 1.37E+05 2.83E-05 301.55 26.57 0.87% 280.30 TAFEG42E5+0.30 -05 446.21 24.56 0.35% 182.40 Average 231.35 ± 48.95 PVEG-A TAF443 1.46E+05 3.20E-05 361.01 25.87 0.75% 219.30 PVEG-B TAF443 1.35E+05 3.10E-05 372.18 25.47 0.31% 229.70 A, B and C refer to different runs of the same analyte (VEGF121). noi i ηη / ι znz / R / v [000390] Example 28,- Determination of TAF443 Before Maleimide Conjugation [000391] The Cysteine ​​residue of TAF443 is typically "capped" or oxidized by chemicals in the cell culture medium and is not available for conjugation. In this regard, purified TAF443 (OG1321) is subjected to a decapping (ie reduction) procedure to remove the cap and allow free cysteine ​​residue (ie those not encompassed by Cys-Cyu disulfide bonds) conjugates to the maleimide functionality of a polymer. The decapping is done by mixing the TAF protein with a 30x molar excess for 1 hour at 25°C of the reducing agent TCEP (3,3',3"-phosphanotri¡ltripropanoic acid). The reduction reaction with TCEP is monitored by SDS-PAGE. Undenatured TAF runs as a single band at approximately 250 kDa (approximately 40 kDa of this weight is carbohydrate). When fully denatured, the single 250 kDa band is converted into bands corresponding to the heavy and light chains. After denaturation, the TAF protein is washed by UFdF using a Pellion XL Ultrafiltration Cassette with 20mM Tris pH 7.5, 150mM NaCI, 0.5mM TCEP buffer to remove the cap. The TCEP reagent was then stirred in the same UFdF environment with 20mM Tris pH 7.5, 150mM NaCl. The reduced TAF was allowed to refold using air (Room Oxygen) which was again followed by SDS-PAGE as a test. [000392] A detailed procedure for the takedown in as follows: 500 mg of OG1321 was thawed at -80°C to 4°C overnight, and warmed in the 25°C water bath prior to mixing with TCEP to a 30x molar excess. The reaction was incubated in the water bath at 25°C for 1 hour. Samples were taken at 15, 30 and 60 mimites of run on SDS-PAGE in order to assess the completion of the reduction. A UfdF cassette with 10kD MWCO was used to carry out the damper exchange. The first buffer exchange step was done with 20mM Tris pH 7.5, 100mM NaCI 0.5mM TCEP for ~100x to completely remove the cap. A second buffer exchange step was performed with 20mM Tris pH 7.5, 100mM NaCI for ~1000x to remove TCEP prior to air refolding. The final concentration of TCEP in the sample was ~0.5μΜ. Samples were drawn from both steps in the buffer exchange for both SDS-PAGE and SEC analyzes to assess protein aggregation and protein reoxidation status. After the second buffer exchange step, the OG1321 was concentrated to ~2mg / ml, filtered at 0.22pm, and allowed to re-oxidize in air at 4°C. Samples were taken for SDs-PAGE and SEC analyzes at different time points to assess the re-oxidation status. The re-oxidized OG1321 was filtered at 0.22 pm and further concentrated. The sample was further concentrated with VIVACELL 100 30k MWCO rotary concentrators to 46mg / ml and the sample was sterile filtered. It was quantified by OD280. [000393] Example 29,- Conjugation of TAF443 to Biopolymer [000394] TAF443 which was also named OG1321 was conjugated to OG802 polymer (see below) after knockdown using 15x excess of polymer in Tris pH 7.5 buffer to produce OG 1448, shown in Figure 26, which shows the chemical structure of OG1448 which is conjugated to TAF443 to the OG1802 biopolymer. TAF443 is on the extreme right hand side of the molecule shown in the figure, conjugated via cysteine ​​residue 443 in the 5-membered ring. Conjugation was monitored by SDS-PAGE and activated near completion. The conjugate was purified by anion exchange chromatography and buffer-exchanged in the formulation buffer by UF / DF. [000395] In general, there were three steps involved in the synthesis of OG1448 from the components OG1802 and OG1321. Step A: OPG1321 must be reduced or uncapped to release the sulfhydryl groups at cysteine ​​position 443. Although the cysteine ​​position at position 443 of the TAF heavy chain is not believed to be involved in the cysteine-cysteine ​​disulfide pairing, this cysteine ​​is typically capped by components of the medium and absent reduction is not available to react with maleimide. Step B: Reduced TAF is then conjugated to OG1802: Step C: Conjugated TAF (OG1448) is then separated from unconjugated TAF and polymer by chromatography. These three general steps are broken down into several smaller steps in the following table: noi i ηη / ι znz / R / v General Step Description IPC Assays Range Target A Step 1: To reduce OG1321 using tris(2-carboxyethyl)phosphine (TCEP). 30x molar TCEP at 25°C for 1 hour SDS-PAGE non-reducing >95% reduction Step 2: To remove TCEP reducing agent and capping groups using UF / DF. First, wash with 0.5 mM TCEP in Tris pH 7.5 by a volume exchange factor of 100 times; followed by a second wash with Tris pH 7.5 buffer per 1000-fold volume exchange factor to remove TCEP, target final TCEP level less than 0.5 μm non-reducing SDS-PAGE Band shift on removal of reducing agent. Step 3: To refold the protein to ensure that the native disulfide pairs are fully oxidized while internal c-cysteine ​​residues remain reduced. Non-reducing SDS-PAGE Band shift on oxidation of native disulfide pairs. UVA / is for protein Final protein concentration at 6-8 mg / ml Β Step 4: To conjugate OG1321 protein to OG1802 biopolymer. Conjugate by mixing the oxidized OG1321 with OG1802. The process requires 15x molar biopolymer to protein stripped and constant mixing. Lower temperature to 2-8°C for 20 hours and blanket reaction with nitrogen gas to minimize oxidation. Non-reducing SDS-PAGE <20% full length band remains. Analytical AE-HPLC <20% unreacted protein at OD 280 nm. C Step 5: for OG1448 conjugate of unreacted OG1321 protein, unreacted OG1802 biopolymer, analytical AE-HPLC for unreacted polymer and unreacted protein <5% unreacted protein at OD 180 nm; <15% polymer without ηοι ηη / ι ζπζ / β / υ protein aggregates and other process contaminants. Purify OG1448 using MacroCap Q (AEX). Chromatography is performed at pH 7.5 in 20mmM Tris buffer and eluted using NaCI gradient. A mixture is made by combining the fractions. react at OD 220 nm non-reducing SDS-PAGE <5% unreacted protein Step 6: To concentrate OG1448 and exchange chromatography buffers for formulation buffers. The pooled portions from the previous step were diafiltered and then concentrated by UF / DF to target OG1448. UV / Vis for protein concentration OG1448a50 mg / ml Step 7: To remove bioburden from the final product and to dispense into storage containers. The final UF / DF mixture is filtered at 0.2 pm into sterile containers, and the pH and conductivity of the final filtrate are established. The drug substance is stored at -20°C. UV / Vis for protein concentration OG1448a50 mg / ml pH, Conductivity pH 7.2-7.5 noi i ηη / ι znz / B / v [000396] Example 30,- Purification of OG1448 by anion exchange (Macrocap Q) After conjugation of TAF443 to OG1802 as described above, OG1448 was purified as follows: After conjugation of TAF443 to OG1802 as described above, OG1448 was purified as follows: 2 x 400ml Macrocap Q columns were packed according to the ~3:1 ratio of resin:conjugate. Columns were rinsed with 5M NaCI and equilibrated with 20mM Tris pH 7.5, 20mM NaCI (equilibrium buffer) by siphoning. The conjugation reaction mixture was diluted with 20mM Tris pH 7.5 and loaded onto the columns. The columns were then flushed with the equilibration buffer and washed with 20mM Tris pH 7.5 50mm NaCI (Wash 1) and then 20mM Tris pH 7.5, 100mM NaCI (Wash 2). Elution was done with 20mM Tris pH 7.5 with a gradual NaCI gradient of 150mM, 200mM, 220mM, 250mM, 300mM, and 500mM. All column flows, washes, and elution were collected in clean bottles for SDS-PAGE and AEX analyses. The elution fractions containing the conjugate were mixed and concentrated using the Pellicon XL TFF cassette with 30kD MWCO and PES membrane. The concentrated mixture was then buffer-exchanged against buffer-versus-buffer xPBS pH 7.4 for ~100x using the same TFF cassette and transferred to -VIVACELL 100 rotary concentrators for further concentration until the target concentration (~30 mg / ml) was achieved. The final conjugate was filtered through a 0.2 pm PED syringe filter for batch release. Example 31.- Reduction of bacterial endotoxin To reduce endotoxin levels in the final protein (OG1321) or conjugate (OG1448), purification procedures for either protein or conjugate that use cation exchanges instead of anion exchanges can be employed. For example, in the above procedure to purify OG1321, TMAE ion exchange resin was used. Instead of TMAE resin, CEX cation exchange resin can be used. However, in order to use the CEX residue, the pH of the solution containing the protein in question must be lowered to below that of the present protein invention. For OG1321, the pH of the protein solution after the protein A column was lowered to pH 3.5. OG1321 binds to the Poros XS column at pH 7.5. Then, Porox XS (CEX) can be used to bind and elute OG1321. [000397] Example 32.- Route 1, Synthesis of OG1802 A first route for the synthesis of OG1802 is as follows. First, the TFA initiator / amine salt (Compound L) having the structure shown in Figure 27 was synthesized as follows. [000398] First, compound K, having the structure shown in Figure 28, was synthesized as follows. In a 200 mL round bottom flask under nitrogen was placed Compound J (OG15623) (1.9 g, 2.67 mmol, 3.3 equiv) noi i ηη / ι znz / R / v and Compound E (0.525 g, 0.81 mmol, 1.0 equiv) (see FIG 38) followed by dimethylformamide (10 mL) then diisopropylethylamine (2.5 mL, 14.6 mmol, 18 equiv). The flask was cooled to 0°C using an ice bath. To this was added propylphosphonic anhydride solution (50 wt% in ethyl acetate, 2.5 mL, 4.04 mmol, 5 equiv) over ~6 minutes. [000399] The reaction was warmed to room temperature and stirred for 15 minutes. The reaction was quenched by adding water (20 mL), saturated aqueous sodium bicarbonate (20 mL), and ethyl acetate (100 mL). The organic layer was separated and the aqueous layer was extracted with ethyl acetate (75 mL). The combined organic layers were washed with saturated aqueous sodium bicarbonate (30 mL), 0.5 M aqueous citric acid (40 mL), water (25 mL), and saturated aqueous sodium chloride (40 mL), then dried (sodium sulfate). sodium), filtered and concentrated under vacuum. The residue that was used without further purification resulted in 2.0 g (0.80 mmol, 99%) of Compound K. 1H NMR (400 MHz DMSO-d6): 0D = 1.36 (s, 9H, OCCH3), 1.90 (s, 54H, CC(CH3)2Br), 2.31 (t, J = 7.2 Hz, 6H, CCH2CH2NH). 2.98 (d, J = 5.6 Hz, 6H, CCH2NH), 3.04 (q, J = 6.0 Hz, 2H, OCH2CH2NH). 3.18 (s, 2H, OCH2C), 3.3-3.37 (m, 8H, CH2), 3.47-3.55 (m, 12H, CH2), 3.58 (s, 6H, OCH2C), 3.87 (s, 6H, O=CCH2O) , 4.27 (s, 18H, CCH2OC=O), 6.74 (br t, 1H, CH2NHC=O), 7.69 (t, J = 6.8 Hz, 3H, CH2NHC=O), 7.84 (t, J = 6.0 Hz, 3H , CH2NHC=O). LC-MS (ES, m / z): [(M+2H-boc) / 2]+Calculated for (C84H136Br9N7O33+2H-Boc) / 2 = 1196.6; Found 1196.6. [000400] Compound L (Figure 27) was then synthesized as follows: In a 100 mL round bottom flask under nitrogen was added Compound K (2.0 g, 0.8 mmol), dichloromethane (10 mL) followed by trifluoroacetic acid ( 5mL). The reaction was stirred at room temperature for 30 minutes. The reaction was concentrated under vacuum. The reaction was diluted using dichloromethane (10 mL) and concentrated in vacuo. The residue was dissolved using acetonitrile (100 mL), filtered through a syringe filter (Acrodisc CR25, PN 4225T) and loaded onto a prep HPLC column and eluted with 60% acetontrile in water (with trifluoroacetic acid). 0.1%) to 98% acetonitrile (with 0.1% trifluoroacetic acid). The tubes containing the product were mixed, concentrated under vacuum, frozen and placed in a lyophilizer. This resulted in 990 mgs (0.4 mmol, 50% over 2 steps) Compound L as a white powder. 1H NMR (400 MHz DMSO-d6): DD= 1.90 (s, 54H, CC(CH3)2Br), 2.31 (t, J = 7.2 Hz, 6H, CCH2CH2NH), 2.97-3.0 (m, 8H, CCH2NH and OCH2CH2NH ), 3.17 (s, 2H, OCH2C), 3.3 (q, 6H, CH2CH2NHC=O), 3.4-3.59 (m, 20H, CH2), 3.87 (s, 6H, O=CCH2O), 4.27 (s, 18H, CCH2OC=O), 7.69-7.84 (m, 9H, both CH2NHC=O and NH3+). LC-MS (ES, m / z): [(M+2H) / 2]+Calculated for (C84H136Br9N7O33+2H) / 2 = 1196.6; Found 1197.4. [000401] Then, compound L was used as a starter to synthesize the MPC polymer. Typically the initiator is prepared as a stock solution in DMF of approximately 100 mg / ml. The initiator and the ligand (2,2'-bip¡r¡d¡lo) were introduced into a Schlenk tube. The resulting solution was cooled to -78°C using a dry ice / acetone mixture, and degassed under vacuum for 10 minutes. The tube was filled under argon and catalyst (CuBr unless otherwise stated), held under argon, was introduced into the Schlenck tube (the molar ratio of bromine atom in initiator / catalyst (CuBr) / ligand remained at 1 / 1 / 2). The solution became dark brown immediately. The Schlenk tube was sealed and immediately purged by applying a short vacuum / argon cycle. Se noi i ηη / ι znz / B / v prepared a HEMA-PC solution by mixing a defined amount of monomer, prepared in a drawer kept under nitrogen, with 200 degree degassed ethanol. The monomer solution was added dropwise into the Schlenk tube (via cannula) (and homogenized by gentle agitation). The temperature was maintained at -78°C. Full vacuum was applied to the reaction mixture for at least 10 to 15 minutes until solution bubbling ceased. The tube was then filled with argon and warmed to room temperature. The solution was stirred, and as polymerization proceeded, the solution became viscous. After 3 to 8 hours or only overnight the reaction was quenched by direct exposure to air akin to oxidize Cu(I) to Cu(II), the mixture became blue-green in color, and was passed through a silica column in order to remove the copper catalyst. The collected solution was concentrated by rotary evaporation and the resulting mixture was either precipitated with tetrahydrofuran or dialyzed against water followed by freeze-drying to produce a free flowing white powder. The table below sets forth the polymer data for the polymer using compound L as an initiator. noi i ηη / ι znz / R / v Theoretical MW (kDa) Polymer ID No. Initiator Mn(kDa) Mp(kDa) PDI 500 130 L 490 530 1.1 750 150 L 645 750 1.1 [000402] The maleimide Mal-PEG4-PFP-ester was then pressed (as set forth in Figure 29) to the 750 kDa polymer referred to above to provide OG1802. Polymer R3707 (750 kDa polymer made using L as initiator, 515 mg) was placed in a 20 mL flask and dissolved using ethanol (4.0 mL) after stirring for 40 minutes. To this was added a 1% solution of 4-methylmorpholine in acetonitrile (22uL). In a separate bottle, Mal-PEG4-PFP (1.97mg) was dissolved in acetonitrile (1.0mL) and this solution was added to the polymer solution over ~2 minutes at room temperature and the resulting solution was stirred overnight. The reaction was diluted with 0.1% aqueous trifluoroacetic acid (2 mL) (pH ~5) followed by water (~12 mL), filtered through a syringe filter (Acrodisc Supor, PN 4612) and placed evenly in 3 Amicon centrifuge membrane dialysis tubes (30,000 mwco). The tubes were diluted and mixed with water (~5 mL each), placed in a centrifuge (rpm 3200) for 25 minutes. The filtrate is removed for analysis while the retentate is eluted and mixed with water (~10 µL / tube). The centrifuge procedure was repeated 5 more times, after which the retentate was removed and placed in a flask. Amicon membrane tubes were rinsed with water (2 x ~2 mL each tube) and this was combined with the retentate. The retentate solution was filtered through a syringe filter (Acrodisc Supor, PN 4612), frozen, and placed in a lyophilizer. This resulted in 485 mgs as a white powder. [000403] Example 33.- TAF Biacore Union Studies (OG1448 and OG1321) [000404] The binding affinity of OG1448 (and OG1321) to their proposed targets was assessed by Biacore assay. Binding studies were performed at 25°C and 37°C using BioRad Proteon XPR36 and Biacore 2000 optical biosensors equipped with GLM chip sensors (Proteon) and CM4 (Biacore) and equilibrated with running buffer (10 mM HEPES, NaCI 150 mM, 0.005% Tween-20, 0.2 mg / ml BSA). OG1448, OG1321, bevacizumab, aflibercept and anti-PDGF were immobilized to the sensor surface by amine coupling. [000405] The binding of the proteins coupled to the ligands was determined by standard methodology. For example, rhVEGFA-165 was tested for binding in a three-fold dilution series starting at 52 nM. rhVEGFA165 was injected through the surface for five minutes and then the dissociation phase was monitored for >1000 seconds as the surfaces were washed with the buffer. The rhVEGFA-165 / OG1448 complex appeared quite stable, as indicated by the apparently flat response during the washout phase (>300 seconds) (data not shown). The dissociation phase for 52 nM rhVEGFA-165 was monitored for more than 2 hours. No decrease in the binding response was observed over time. noi i ηη / ι 7Π7 / Ε / Υ [000406] Similarly, rhPDGF-BB was tested for binding in a three-fold series starting at 11.4 nM. For the rhPDGF-BB / OG1448 interactions, the rate constants were too fast to be reported with confidence due to mass transport effects. The following Kd constants were observed: Kd (pM) OG1321 OG1448 Bevacizumab Aflibercept Anti-PDGF rhVEGFA-165 (25°C) 9.8±0.1 5.1 ±0.1 9.6±0.8 1.56±0.2 rhPDGF-BB (25°C) 14±3 17±2 107±3 Example 34.- TAF.- a Competitive Inhibitor of the Binding of rhVEGFA-165 to rhVEGFR As a measure of its potential potency in anti-VEGF activity, the binding activity of TAF (OG1448 and OG1321) to VEGFA-165 was assessed in a competitive binding assay where TAF, at different concentrations, was competent with immobilized rhVEGFR for binding. rhVEGF binding. rhVEGFA-165 bound by the immobilized VEGFR was determined by ELISA (data not shown). Human VEGFR1 / Fc was coated onto the bottom of 96-well ELISA plates at 1.0 pg / mL Various concentrations of TAF (OG1448 and OG1321), ranging from 0.39 to 200 nM, were incubated with 0.1 nM biotinylated VEGFA-165 for 30 minutes before adding to ELISA plates. Biotinylated rhVEGFA-165 bound to VEGFR1 was detected by streptavidin-HRP and followed by growth with HRP substrates. Ranibizumab (Lucentis) and bevacizumab (Avastin) were similarly tested for inhibition of competitive binding of VEGFA-165 to VEGFR1. OG1321, OG1448, ranibizumab, and bevacizumab showed similar IC50s upon inhibition of VEGF-165 binding to rhVEGFR suggesting similar potential potency in anti-VEGF activity. These results suggest that TAF (both OG1448 and OG1321) may be as potent as the approved agents ranibizumab and bevacizumab, thus suitable for treating neovascular (ie, wet) AMD. ICso(nM) OG1321 OG1448 Ranibizumab Bevacizumab Competitive Union to rhVEGFA-165 (vs VEGFR) 12.5±1.2* 8.5±1.Γ 12.5±1.2* 10.7±0.9* *Mean and SD of at least three trials. noi i ηη / ι znz / B / v [000407] Example 35.- OG1448.- a Competitive Inhibitor of the Binding of rhVEGFA-165 to rhVEGFR in the Presence of rhPDGF-BB [000408] To assess whether OG1448 can bind to rhVEGFA-165 in the presence of rhPDGF-BB, ie, either rhPDGF-BB binding or TAF receptor decoy inhibits the ability of TAF to bind rhVEGFA-165, a binding study similar to Example 27 was carried out but in the presence of various concentrations of rhPDGF-BB. [000409] Human VEGFR1 / Fc was coated onto the bottom of 96-well ELISA plates at 1.0 pg / mL. Various concentrations of OG1448 were incubated with 0.1 nM rhVEGFA-165 plus rhPDGF-BB at 0.4, 1.2, and 2.0 nM, respectively, for 30 min before adding to ELISA plates. Binding of rhVEGFA-165 to rhVEGFRI was detected by biotinylated anti-VEGFA antibody, 0.4 pg / mL, followed with streptavidin-HRP and HRP substrate. OG1448 was found to have an IC50 (nM) of 10.1. This is quite comparable to the IC50 observed without rhPDGFBB of Example 28. The value of OG1321 was not determined in this assay but is expected to be similar to OG1448. [000410] Example 36.- TAF.- a Competitive Inhibitor of the Binding of rhPDGF-BB to rhPDGFR [000411] As a measure of its potential potency of anti-PDGF activity, the binding activity of TAF (OG1448 and OG1321) to rhPDGF-BB was assessed in a competitive binding assay where TAF, at different concentrations, was competing with PDGFR immobilized for rhPDGFBB binding. rhPDGF-BB bound to immobilized PDGFR was determined by ELISA assay. [000412] Human PDGFR / Fc was coated onto the bottom of 96-well ELISA plates at 0.4 pg / mL. Various concentrations of OG1448 and OG1321, ranging from 1 pM to 20 nM, were incubated with 0.2 nM rhPDGF-BB for 30 min before adding to ELISA plates. rhPDGF-BB bound to rhPDGFR was detected by biotinylated anti-PDGFBB antibody, 0.4 pg / mL, followed with streptavidin-HRP and HRP substrate. [000413] OG1448, OG1321 and a control anti-PDGF antibody showed similar IC50s when inhibiting the binding of rhPDGFBB to PDGFR, as shown in the table below, suggesting highly potent anti-PDGF activity. IC50 (pM) Competitive Binding to rhPDGFBB (vs rhPDGFR) OG1321 OG1448 Anti-PDGFBB 46±21* 54±21 66 *Mean and SD of three assays [000414] Example 37.- OF1448.- a Competitive Inhibitor of the Binding of rhPDGF-BB to rhPDGFR in the Presence of rhVEGFA-165 [000415] To assess whether OG1448 can bind to rhPDGF-BB in the presence of rhVEGFA-165, a similar competitive inhibition of PDGF binding assay (as in Example 29) was performed in the presence and absence of rhVEGFA-165. [000416] Human PDGFRb / Fc was coated onto the bottom of 96-well ELISA plates at 0.4 pg / mL. Various concentrations of OG1448 were incubated with 0.2 nM PDGFBB and with 0.2 nM PDGFRb plus rhVEGFA-165 at 0.2 nM, 0.6 nM, and 1.0 nM, respectively, for 30 min before adding to ELISA plates. PDGF-BB binding to PDGFRb was detected by biotinylated anti-PEGFBB antibody, 0.4 pg / mL, followed by streptavidin HRP and substrate HRP. The IC50 (pM) in the presence of rhVEGFA-165 (25) was comparable to the figure derived in Example 29. The figure for OG1321 in the presence of rhVEGFA-165 was not determined but is expected to be similar. [000417] Example 38. Inhibition of VEGF-induced Proliferation of Human Primary Retinal Microvascular Endothelial Cells (HRMVEC) [000418] Endothelial cell proliferation is a crucial step in angiogenesis and thus in the pathogenesis of neovascular AMD. The ability of OG1448 to antagonize the proliferative action of VEGF on primary human retinal microvascular endothelial cells may be a measure of its bioactivity in treating neovascular AMD. noi i ηη / ι 7Π7 / Β / Υ [000419] HRMVEC were stimulated with 1.3 nM rhVEGF165-A for 3 days in the presence of various concentrations of TAF (OG1448 and OG1321) and control drugs. Cell proliferation was measured by WST-1 cell proliferation detection reagent. The results are shown in the table below: IC50 (nM) OG1321 OG1448 Ranibizumab Bevacizumab Aflibercept Inhibition of VEGF-induced proliferation of HRMVEC 0.43±0.05* 0.49±0.05* 0.98±1.21* 0.81 ±0.32* 0.55±0.08* *Mean and SD of at least 3 assays [000420] OG1448 and OG1321 demonstrated an IC50 in this trial comparable to other anti-VEGF therapies. These data show that TAF (both OG1448 and OG1321) have at least comparable potency to inhibit VEGF-mediated retinal microvascular endothelial cell proliferation activity as ranibizumab, bevacizumab and aflibercept. [000421] Example.- Inhibition and Proliferation Induced by PDGF of Human Cerebral Vascular Pericytes Primary (HBVP) [000422] The migration and proliferation of pericytes are crucial events in angiogenesis and therefore play important roles in the pathogenesis of neovascular AMD. The ability of TAF (OG1448 and OG1321) to antagonize the proliferative action of PDGF in human brain pericytes may be a measure of its effectiveness in treating neovascular AMD. [000423] HBVP were stimulated with 2.0 nM PDGFVBB for 3 days in the presence of various concentrations of TAF (OG1449 and OG1321) and a reference anti-PDGF-BB antibody (R&D Systems, Catalog # AB-220-NA). Cell proliferation was measured by WST-1 cell proliferation detection reagent, noi i ηη / ι znz / B / v IC50 (nM) OG1321 OG1448 Anti-PDGF Inhibition of PDGF-induced proliferation of HPVP 5.0±2.0* 2.9±1.4 5.4 *Mean and SD of at least 3 assays From the several previous experiments comparing OG1321 (TAF443) to OG1448 (TAF443 polymer conjugate), it can be seen that conjugation to HEMA-PC biopolymer has no negative impact on protein activity. [000424] OG1448 and OG1321 show comparable IC50 to anti-PDGF antibody. [000425] Example 40,- Inhibition of Sprout in Co-culture of Retinal Microvascular Endothelial Cells Human (HRMVEC) and Human Mesenchymal Pericytes (HMP) [000426] To mimic the in vivo conditions where endothelial cells and pericytes coexist in blood vessels and co-proliferate and migrate during angiogenesis, crucial events in neovascular AMD, a three-dimensional co-culture of HRMVEC and HMP was established with the aim of evaluating the ability of OG1448 to inhibit angiogenesis in this complex model. [000427] Vehicle, Avastatin, an anti-PDGF-BB antibody (same as above), Avastatin is a combination with anti-PDGF-BB antibody, and OG1448 were added to the co-cultures on day 7. On day 14 , immunohistochemical staining of CD31 (endothelial cells) and aSMA (pericytes) was used to quantify the lengths of shoots emanating from established endothelial cell spheroids compared across experimental groups. [000428] OG1448 was more effective at inhibiting pericyte / endothelial sprouting in HRMVEC-HMP co-culture than Avastin alone or anti-PDGF alone at two different concentrations. In addition, OG1448 was also more effective at inhibiting flare than a combination of Avastin and the anti-PDGF-BB antibody. This demonstrates that OG1448 is synergistic with respect to Avastin and an anti-PDGF-BB antibody. The results are shown in the table below and in Figure 40. Drug Mean Total Shoot Length (pix) S.D. (pix) Relative Angiogenesis, % S.D. % Vehicle 6999 1266 100 18 Avastin-5nM 4700 722 67 10 Avastin-25nM 3763 909 54 13 Anti-PDGF-5nM 4924 884 70 13 Anti-PDGF-25nM 4461 1051 64 15 Avastin + anti-MPDGF-514 7 4 97 Avastin + antiPDGF-25nM 4287 822 61 12 OG1448-5nM 3584 478 51 7 OG1448-25nM 2933 360 42 5 noi i ηη / ι znz / B / v [000429] Example 41.- Efficiency of OG1448 in the Inhibition of Laser-Induced Colloidal Neovascularization in Cynomolgus Monkeys [000430] The in vivo efficiency of OG1448 was assessed using the laser-induced colloidal neovascularization (CNV) model in cynomolgus monkeys, a well-recognized primate model of CNV. See, for example, Nork TM, Dubielzig RR, Christian BJ, et al. 2011. Prevention of experimental choroidal neovascularization and resolution of active lesions by VEGF trap in nonhuman primates. Arch Ophthalmol. 129: 1042-1052; Uoyd RL, Harris J, Wadhwa S, Chambers W. 2008. Food and Drug Administration approval process for ophthalmic drugs in the U.S. Curr Opin Ophthalmol. 19:190-194, both of which are thus incorporated by reference. In this model, laser lesions in the chororetinal complex are placed in the macula of the monkey eye with evidence of Bruch's membrane rupture. Colloidal neovascularization developed in two to three weeks. At various time points, fluorescein angiography was used to evaluate clinically relevant lesions (Grade IV) showing fluorescein leakage beyond the primary lesion. This CNV model has been used extensively for the study of CNV lesions and was used as a benchmark for all currently approved treatment for neovascular AMD. In this model, all anti-VEGF agents approved for neovascular AMD are effective in inhibiting leakage from clinically relevant Grade IV lesions. The study was conducted at Covance, Madison, Wl. [000431] In summary, a dose-related response to a single intravitreal injection of OG1448 at 0.5 to 2.4 mg / eye (calculated based on protein content) was observed in animals in which CNV lesions were allowed to settle. developed over 14 days prior to treatment and assessed at subsequent time points using the fluorescein angiography approach in clinically relevant Grade IV lesions in the retina / choroid. At 0.5 mg / eye, the beneficial effect on Grade IV lesions was discernible (p=0.019; generalized estimating equation [GEE] model; 0.5 mg treatment Group 7 versus PBS injected with placebo Group 5). At 2.4 mg / eye OG1448, one dose (in molar equivalence) within the therapeutic dose of bevacizumab or aflibercept, was highly effective (75% reduction in CNV Grade IV lesions at day 43 from day 15 versus 27 % reduction in PBS-treated group) (p=0.007; GEE model; 2.4 mg Treatment Group 9 vs. placebo-injected PBS Group 5) in ameliorating leakage from Grade IV - CNV lesions. [000432] OG1448 shows effectiveness in inhibiting clinically relevant Grade IV lesion leakage in this brand CNV model. [000433] The groups and study design are shown in the following table. The study included tolerability groups (Groups 1 through 4) however for purposes of this patent application, only groups for pharmacological activity and a control group treated with phosphate buffered saline (PBS) injection are shown. saline (PBS) injection are shown. noi i ηη / ι znz / R / v Group Number of Females Dose Route Dose Level Dose concentration (mg / ml) mg / left eye / dose mg / right eye / dose mg / kg / dose 5 6 Intravitreaa 0 0 NA 0 6 6 Intravitreaa 0.24 0.24 NA 5.9 7 6 lntravitreab 0.51 0.51 NA 10.2 9 6 lntravitreab 2.40 2.40 NA 26.6 NA = not applicable to = on days 1, 15, and 29 (a total of 3 doses); laser on day 8 of the dosing phase. b = once; laser treatment in 15 days before injection. [000434] Two treatment regimens were evaluated. In the prevention regimen, OG1448 was given intravitreously three times and laterally at 0.24 mg / eye / dose (dose content was based on protein content; Group 6) or PBS (Group 5) on days 1, 15, and 29 with laser treatment on day 8 of the dosing phase. Fluorescein angiograms on days 15, 21, 30, 37, and 43 of laser treatment (days 22, 28, 37, 44, and 50 of the dosing week) were used for evaluation of clinically relevant Grade IV lesions. [000435] In the treatment regimen (Groups 7 [0.5 mg], 8 [0.5 mg] and 9 [2.4 mg], OG1448 was administered intravitreously to both eyes of 6 animals at doses of 0.5 mg (Groups 7 and 9) or 2.4 mg / eye (Group 9) 15 days after laser induction when CNV lesions were established Fluorescein angiograms obtained on days 15, 21, 30, 37 and 43 of laser treatment were used for evaluation of the clinically relevant Grade IV injuries. [000436] Using the Generalized Estimating Equation (Gee) models (Halekoh, U & Yan J (2006) The R Package geepack for Generalized Estimating Equations Journal of Statistical Software 15, 2, pp1-11), a related answer to the dose to OG1448 was observed in the intervention regimen. At 0.5 mg / eye, the effect was remarkable as shown by the difference in % change in Grade IV lesions compared to vehicle control (0.5 mg Treatment Group 7 versus PBS injected with placebo Group 5; p=0.019 , GEE). At 2.4 mg / eye, OG1448 (one molar equivalency dose within the therapeutic dose of bevacizumab or aflibercept) a 75% reduction in percent change in Grade IV lesions (2.4 mg Treatment Group 9 versus placebo-injected PBS Group 5; p=0.0007, GEE) was observed at day 43 compared to a 27% reduction in CNV in the PBS control group. Data from the various experiments in the monkey CNV model are shown in Figure 41. [000437] OG1448 shows dose-dependent effectiveness in inhibiting clinically relevant Grade IV lesion leakage in this CNV model. These results are consistent with previously described studies showing activity of OG1448 against VEGF-mediated angiogenic activities. [000438] Example 42,- Pharmacokinetics and Tissue Distribution [000439] A pharmacokinetic and tissue distribution study was carried out using 125I-OG1448 using New Zealand Red White F1 Cross pigmented rabbits. In summary, this study showed a vitreal half-life of 16.1 days for OG1448 in rabbits, approximately three times that reported for aflibercept (4.5 days) and 5 times that of ranibizumab (2.9 days) (Bakri SJ, Snyder MR, Reid JM et al. 2007. Pharmacokinetics of Intravitreal Ranibizumab [Lucentis], Ophthalmology 114:2179-2182) with low plasma exposure (approximately 0.2% of that of vitreous exposure) and a plasma half-life of 6.5 days (aflibercept reported 6.5 days) ( Struble C, Koehler-Stec E, Zimmer E, and Tu W. 2008. Pharmacokinetics and ocular tissue penetration of VEGF Trap after intravitreal injections in rabbits. EVER; Portorz, Slovenia). [000440] The purpose of this study was to evaluate the ocular distribution and pharmacokinetics of non-radiolabeled test articles and radiolabeled test articles following intravitreal or intravenous dosing to male New Zealand Red White F1 rabbits. Treatment groups and study design are shown in the table below noi i ηη / ι znz / R / v Groups and study design (Covance study) Group # of males Dose Route Test Item Dose (mg) Sample Collected 1 14 IVT 125I-OG1448 0.25 / eye (OU) Blood, ocular tissues 2 2 IV 125I-OG1448 0.25 / animal Blood 3 6 IVT OG1448 0.25 (OD) Blood, whole eyes for histology 4 6 IVT OG1448 0.25 (OD) Blood, vitreous IVT: intravitreal; IV: intravenous; OU: Both eyes; RE: right eye [000441] Two PK parameters were obtained based on radioassays. The clearance profiles of the vitreous, retina, and choroid were similar to each other. This pattern is consistent with other established CNV treatments such as ranibizumab or aflibercept. Set forth in the table below are the pharmacokinetic parameters in different ocular tissues after single bilateral intravitreal injection of 0.12 mg of 125I-OG1448. noi i ηη / ι znz / R / v Matrix Cmax (NG Eq. / G) Tl / 2 (day) AUCo-m (Day*NG EQ. / G) Exposure as % of vitreous exposure Plasma 494 6.48 3,790 0.189 Aqueous humor 5,250 11.6 68,800 3,423 Choroid-RPE 4,170 32.8 134,000 6,667 Iris-ciliary body 12,100 42.6 235,000 11,692 Retina 13,500 30.4 309,000 15,373 Vitreous humor 112,000 16.1 2,010,000 100.00 [000442] The ocular tissue half-life of various VEGF inhibitors is compared to OG1448 in the table below and in Figure 42, suggesting that OG1448 may be above a pharmaceutically active minimum inhibitory concentration of 0.1 pg / ml for more than 90 days, as opposed to 30 days for Lucentis and 50 days for Eylea: Ocular Tissue Elimination Half-Life (Days) Humor Vitreous Retina Choroid Pegaptinib1 3.5 - - Ranibizumab1 2.9 2.9 Aflibercept1 4.5 5.5 4.8 OG14482 16.1 30.5 32.9 1Based on publicly available data from 28-day rabbit studies: Drolet DW, Nelson J, Tucker CE, et al. 2000. Pharmacokinetics and safety of an anti-vascular endothelial growth factor Aptamer (NX 1828) following injection into the vitreous humor of rhesus monkeys. Pharm Res. 17:1503-1510; Gaudreault J, Fei D, Bojor JC et al. 2007. Pharmacokinetics and retinal distribution of ranibizumab, a humanized antibody fragment directed against VEGF-A, following intravitreal administration in rabbits. Retina 27:859-870; Bakri (2007), above; Struble 2008, supra. 2Based on 250 pg intravitreal injection into rabbit eye [000443] The study showed a vitreal half-life of 16.1 days for OG1448 in rabbits, approximately three times the vitreal half-life of 4.5 days reported for aflibercept and five times the vitreal half-life of ranibizumab (2.9 days) (Bakri 2007, supra ) with low plasma exposure (approximately 0.2% of that of vitreous exposure); plasma exposure is consistent with that of aflibercept (Sinapis Cl, Routsias JG, Sinapis Al, et al. 2011. Pharmacokinetics of intravitreal bevacizumab [AvastinMR] in rabbits. Clinical Ophthalmology 5:697-704). Similar to the data reported for ranibizumab and aflibercept, the vitreous, retinal, and choroidal clearance profiles are similar to each other. [000444] Example 43,- Toxicology [000445] Two non-GLP single dose systemic and ocular tolerability pilot studies of OG1448 were conducted at Covance: (i) a single dose 57-day intravenous or intravitreal tolerability study in pigmented rabbits and (i) a Single dose tolerability study after intravitreal (58-day study) or intravenous (28-day study) administration in cynomolgus monkeys. [000446] In summary, single dose intravitreal injection of 0.25 mg OG1448 / dose / eye in rabbits was initially well tolerated but was associated with persistent anterior segment inflammation (mild to moderate conjunctival hyperemia, mild to moderate watery redness, and cells) and posterior (mild to moderate vitreous white cells, mild to moderate vitreous opacity and presence of vitreous floaters, and multifocal gray-white subretinal inflammatory foci) that developed approximately 2 weeks post-dose (or later). This inflammatory response was ameliorated with immunosuppressive and anti-inflammatory therapy. Post-dose onset time and treatment response are consistent with a typical immune-mediated response for intraocularly administered humanized biopharmaceuticals in animals. [000447] In contrast, a single intravitreal dose at 0.24 or 1.4 mg OG1448 / dose / eye was well tolerated in cynomolgus monkeys without adverse finding or evidence of immune reactions ophthalmologically, clinically, and histopathologically. [000448] In the efficiency study (discussed above), intravitreal doses of 0.24 mg / eye / dose three times 14 days apart or a single injection of 0.5 mg / eye / dose were well tolerated with at least 40 days of follow up as shown in eye exams. No immune related reactions were seen in the eyes of the treated animals. [000449] These studies demonstrate that OG1448 is well tolerated when administered intravitreal or intravenously at the doses tested. [000450] Example 44.- Individual Dose Tolerance in Cynomolgus Monkeys [000451] The purpose of this part of the study was to evaluate the tolerability of OG1448 after intravitreal or intravenous administration in cynomolgus monkeys. [000452] The systemic and ocular tolerability groups and study design are shown in the table below: noi i ηη / ι znz / R / v Group and Study Design Group No. of females Dose Route Dose Levelb Dose Concentration (mg / ml) pg / left eye / Dose mg / right eye / Dose mg / Kg / Dose 1 3 IVT 0 0.236 NA 5.9 2 3 IVT 0 1.36 NA 27.2 3 2 IVT NA NA 0.235 9.4 4 2 IV NA NA 1.41 9.4 IVT=intravitreal; IV=intravenous; NA=not applicable a. The right eye in the animals in Groups 1 and 2 received the test article by intravitreal injection. Animals in Groups 3 and 4 received the test article by intravenous injection. b. The left eye of animals in Groups 1 and received vehicle control only (phosphate buffered saline, pH 7.4). noi i ηη / ι znz / R / v [000453] Eye examinations by board-certified veterinary ophthalmologists were performed across all four groups at pre-dose and (i) for intravitreal groups: on days 3, 8, 15, 29, 43, and 57 , and (ii) for the intravenous groups: on days 3, 8, 15, and 29. Animals were followed up with clinical observations and clinical pathology on days 3, 8, 15, 29, 43, and 57 when applicable. Anatomical pathology, macroscopic observation during necropsy of all animals and microscopic evaluations for ocular tissues for groups 1 and 2 (57 days) and for a normal list of systemic organs for groups 3 and 4 (29 days) were also performed. ). [000454] There were no adverse or toxicologically significant findings in either group. There were no findings in clinical observations or body weight in any group. There were no OG1448-related macroscopic or microscopic findings of anatomic pathology for any group (ocular tissues for intravitreally injected groups and normal list of organs / tissues for intravenously injected groups). [000455] Ophthalmologic findings for the intravitreal administration groups were limited to injection-related events such as mild to moderate and momentary presence of aqueous and / or vitreous cells and scarring at the aqueous humor sampling site. [000456] Example 45.- Synthesis of Polymer OG1786 [000457] OG1786 is the nine-arm primer for polymer synthesis used as a precursor in the synthesis of OG1802. Each arm is terminated with a 2-bromoisobutyrate that is capable of initiating polymerization under ATRP. OG1786 is a salt of trifluoroacetic acid (TFA) as shown in Figure 30. OG1786 is prepared as follows. First, OG1550 is reacted with TFA (trifluoroacetic acid) to produce OG1546 as shown in Figure 31. [000458] In a 1 liter round bottom flask equipped with a magnetic stir bar and addition funnel was added OG1550 (14.8 g), methyl tert-butyl ether (MTBE) (350 ml) and water (30 ml ). The mixture was stirred to dissolve the OG1550, then cooled in an ice bath. To this mixture was added a solution of trifluoroacetic acid (4.9 ml) in water (90 ml) dropwise over 90 minutes. After the addition was complete, the mixture was stirred an additional 15 minutes, then removed from the ice bath and allowed to warm to room temperature. The mixture was stirred (after removal from the ice bath) for an additional 4-5 hours, until TLC showed ~5% starting material remaining and the pH of the aqueous humor was between 3 and 4 (pH paper). [000459] The mixture was divided. The MTBE layer was washed with water (30 ml). The aqueous layers are combined then the aqueous is extracted with MTBE (150 mL). This second phase of MTBE was washed with water (30 ml). The combined aqueous layers were washed with a third portion of MTBE (100 mL). The third phase of MTBE was washed with water (25 ml). The aqueous layers were combined again (-250 mL, pH ~4, per pH paper). [000460] The product was collected by lyophilization. 11.5 g of white solid were obtained. This material is extremely hygroscopic, best handled under nitrogen. The product was confirmed by LCMS. [000461] The OG1546 prepared was then reacted with OG1563 to produce OG1784 (as depicted in Figure 32). [000462] In a 250 ml flask under nitrogen equipped with a stir bar was added OG1546 (hygroscopic, 9.0 g), followed by Ν,Ν-dimethylformamide (110 ml). The mixture was stirred at room temperature until all the OG1546 had dissolved (approximately 15 minutes), then OG1563 (29.9 g) was added and the mixture stirred an additional 3 minutes until the OG1563 had also dissolved. The resulting solution was cooled in an ice bath, and Ν,Ν-diisopropylethylamine (37.6 mL) was added over 3 minutes, followed by propylphosphonic anhydride (T3P), 50% in ethyl acetate (34.5 mL) dropwise over 5 minutes (addition of T3P is exothermic). After the T3P addition was complete, the flask was removed from the cooling bath and allowed to come to room temperature. Samples were then taken at 5 minute intervals for LCMS analysis. The reaction showed a very light yellow / brown color. [000463] After 20 minutes, the reaction was cooled again in an ice bath and 5 ml of water was added. The mixture was then removed from the cooling bath and an additional 50 ml portion of water was added, followed by 50 ml of 0.5 M citric acid then isopropylacetate (300 ml). The mixture was divided. The aqueous phase (-300 ml) was extracted with additional isopropyl acetate (150 ml). The aqueous phase was AQ1 for HPLC testing. The combined organic portions were washed with aqueous citric acid (115 mL, 65 mm, which was the mixture of 15 mL 0.5 M citric acid plus 100 mL water), and the aqueous phase was AQ2 (pH -2). The organic phase was washed with water / saturated sodium chloride (100 ml / 25 ml), and the aqueous phase was AQ3 (pH -3). The organic phase was finally washed with saturated sodium chloride (100 ml) and the aqueous phase was AQ4. None of the AQ fractions contained any significant product (data not provided). The organic phase confirmed the product by LCMS. The product was dried over sodium sulphate (80g), filtered and rinsed with isopropyl acetate (75ml) and concentrated on a rotary evaporator to a light brown oil (33.2g). The crude product was stored overnight under nitrogen. [000464] The next day, the crude product was allowed to come to room temperature, then dissolved in acetonitrile / water (46 mL / 12 mL) and filtered using an HPLC filter disc (Cole-Parmer PTFE 0.2 pm, No. noi i ηη / ι znz / R / v product 02915-20). The filtrate was divided into three equal portions and purified in three runs. [000465] Loaded onto a RediSep Rf Gold C18 column (275 g, SN 69-2203-339, Lot# 24126-611 Y) equilibrated with 50% acetonitrile / water. Material was eluted at 100 ml / min using the following gradient (solvent A: water, solvent B: acetonitrile). All relevant fractions were verified by HPLC. Fractions judged to be sufficiently pure were pooled (from all three runs) and concentrated (bath temp maintained at ca. 20°C) on the rotoevaporator, then partitioned between dichloromethane (100 mL) and water (5 mL). ml) / saturated sodium chloride (25 ml). The aqueous part was extracted two more times with dichloromethane (2 x 30 ml). The combined organics were dried over sodium sulfate (35 g), filtered, rinsed with DCM (30 mL), and concentrated. The product and purity were confirmed by LCMS methods. noi i ηη / ι 7Π7 / Ε / Υ OG1784 lot R5172 R5228 OG1546 used 5.3 g 9.0 g OG1563 used 17.6 g 29.9 g Yield isolated 53% 58% Purity (w / w 210 nm) 99.3% 100.0% [000466] OG1405 was then prepared from OG1784 as depicted in Figure 33. In a 500 mL round bottom flask equipped with a magnetic stir bar was added OG1784 (20.9 g), followed by dichloromethane (50 mL). then trifluoroacetic acid (20 ml). The mixture was stirred at room temperature and HPLC analysis showed complete deprotection in 23 minutes. The mixture was concentrated on a rotary evaporator, redissolved in dichloromethane (25ml) and re-concentrated, then redissolved in acetonitrile (25ml) and re-concentrated. The product was confirmed by LCMS. The material from the above (OG1405, 34.5 g, assume 21.0 g as quantitative yield) was used as a crude oil in the next step. No purification needed. OG1405 was then reacted with OG1402 to prepare OG1785 as set forth in Figure 34. In a 500 mL flask under nitrogen equipped with a stir bar was placed OG1402 (5.5 g), followed by acetonitrile (70 mL), then Ν ,Ν-diisopropylethylamine (26.3) and T3P solution (see above) (7.9 ml). The solution was stirred at room temperature for 30 minutes, then cooled in an ice-water bath and a solution of OG1405 (crude oil from above, 34.5 g) in acetonitrile (70 mL) was added. The mixture was warmed to room temperature. After 20 minutes, the reaction was cooled in an ice-water bath and quenched with water (5 mL). The mixture was then concentrated under vacuum under a rotary evaporator to half volume. Samples were taken for LCMS. [000467] More water (50 ml) was added, followed by 0.5 M citric acid (75 ml) and isopropyl acetate (175 ml). The mixture was divided in 5 minutes. The aqueous part was extracted with added isopropyl acetate (50 ml). The combined organic parts were washed with aqueous citric acid (0.13 M, 30 mL, consisting of 10 mL 0.5 M citric acid and 20 mL water). The organic parts were then washed with the mixture of saturated sodium chloride (25 ml) and water (25 ml), then finally washed with saturated sodium chloride (25 ml). They were then dried over sodium sulphate (124g), filtered and rinsed with isopropyl acetate (30ml) and concentrated under rotary evaporator to a light brown oil (27.3g). Samples were taken for analysis by LCMS. [000468] The oil was dissolved in acetonitrile / water (3:1, 15 mL / 5 mL), filtered through an HPLC filter disk (Cole-Parmer 0.2 pm PTFE membrane, product number 02915-20) and was divided into three equal portions, each of which was further purified as follows. [000469] Portions were loaded onto a Redi-Sep Gold C18 column (275 g, SN-69-2203-339, Lot 241234611W) equilibrated in 50% solvent B (acetonitrile) / 50% solvent (water). The material was then purified by reverse phase HPLC with a solvent A: water / solvent B: acetonitrile gradient. The appropriate fractions were pooled and partitioned between dichloromethane (150 ml) and water (5 ml) / saturated sodium chloride (25 ml). The aqueous part was extracted twice with dichloromethane (2 x 50 ml). The combined organics were dried over sodium sulphate (60g), filtered, rinsed with dichloromethane (40ml) and concentrated. The structure and purity was confirmed by various analyzes including LCMS: OG1785 was isolated as a foamy solid (R5329, 19.0 g, 83% yield, 95.1% purity (w / w 210 nm), stored under nitrogen at 4°C. [000470] Then, the tert-butyloxycarbonyl protecting group on OG1785 was removed using trifluoroacetic acid (TFA) to produce OG1786 as depicted in Figure 35. [000471] Example 46.- Synthesis of Polymer 1801 [000472] Compound OG1802 was conjugated to a dulphydryl group of TAF443 to produce OG1448. The OG1801 polymer was first produced from the OG1786 initiator. OG1801 has an amine functionality, which is more stable (than maleimide) during polymer synthesis. To synthesize the OG1801 polymer, a modified version of ATRP was used in which copper (Cu(l)) species are generated in situ by adding metallic copper to Cu(lI). The starting materials and reagents needed in the reaction were calculated based on the batch input of the monomer (HEMAPC) OG47, as well as the target molecular weight (MW). [000473] 50 g of OG47 monomer was weighed into the drawer and 200 ml of degassed EtOH was added to dissolve the monomer at room temperature; shown for the monomer concentration test. Cu(II), Byp, Cu(0) is weighed in a 500 ml flask; purging with argon while adding the monomer solution to the flask; the flask is sealed with a stopper and placed under vacuum for 5 minutes until there are no bubbles. The reaction gradually changed color from light green to dark green, then to light brown; -200 mg of OG1786 initiator is weighed into the drawer, and dissolved in -2000 uL of DMF under room temperature to produce 100 mg / mL of stock solution; shown for initiator concentration and purity test; Add the initiator solution to the flask under argon. The reaction solution becomes dark brown and thickening begins with the passage of time; the system is sealed and the reaction is allowed to occur for 2 days. [000474] OG1801 is then prepared for the addition of the maleimide and catalyst (copper) is removed as follows: A pre-packed Red¡SepMRRf normal phase silica column is used to remove the catalyst. The column size is chosen based on the amount of copper in the reaction mixture. For example, a 300 g column (Cat. # 69-2203-330, Column Size 330 g, CV=443 mL) was used for a 50 g batch of OG1801. Teflon tubing was used for the entire connection since EtOH is the elution solvent. noi i ηη / ι znz / B / v [000475] After removal of copper, all reactions were transferred to a round bottom flask in batches, and EtOH was evaporated by rotary evaporator at 45-50°C under reduced pressure to dryness. In this step, the volume of EtOH collected from the condensation was monitored to ensure that EtOH removal was >90%. The polymer was dissolved in 250 ml of WFI and filtered using a 0.2 um filter. This resulted in a clear to light yellow polymer solution at 150 mg / ml. The solution can be stored at 2-8°C for up to 3 months before use. [000476] Example 47,- Synthesis of polymer OG1802 [000477] Starting materials and reagents needed in the reaction were calculated based on the batch input of OG1801. The linker is 3-maleimidopropionic acid, NHS ester. 30 ml of 0.5 M sodium phosphate (in WFI, pH 8) are added to 50 g of polymer solution (-150 mg / ml). Allow to stir for 1 minute; the pH was 8.0 by pH paper. 204.8 mg of linker is weighed out and dissolved in 4.1 ml DMF to give the concentrated solution of 502 mg / ml. The linker solution is added dropwise 815 uL per minute to the polymer solution with vigorous stirring. It takes 5 minutes to add 4095 uL of linker solution. It is reacted at room temperature for 30 minutes. The reaction is quenched with 20 mL of 5% acetic acid to achieve a final pH of 5. The solution is filtered using a 1 L vacuum filter (0.02 um). [000478] OG1802 is then filtered as follows: Milipore crossflow cassettes were used for purification of polymer in aqueous system. Start with concentration of the polymer solution to 250 ml (-200 mg / ml). Fresh WFI from the reservoir is added and the flow rate of the fresh WFI feed is adjusted to the same as the perneate (-2 ml / min). The UF / DF was adjusted at 2-8°C overnight. Typically, 2.5 L of WFI (10x volume ratio to polymer solution) was used. A sample of the retentate was collected for purity testing. The purity sought was >98%. The polymer solution is filtered through a 1L 0.2 μΜ filter bottle. the polymer solution can be stored at 2-8°C for up to 3 months prior to conjugation. [000479] Example 48.- Formulations of OG1448; injectability 27.2 mg / ml and 44.5 mg / ml solutions of OG1448 were prepared using 1.7 mM KH2PO4, 5 mM Na2HPO4, 150 mM NaCI in sterile water for injection. The OG1448 conjugate was concentrated by a Millipore Pellicon XL TFF cartridge (catalog# PXB030A50, EMD Millipore), 30kD MWCO or VIVACELL 100 rotary concentrator (catalog# VC1022, Sartorius), 30kD MWCO, depending on volume. The 27.2 mg / ml TAF solution was injected intravitreally into the monkeys for the efficiency experiments described above through a 1 / 21 inch 30 gauge (G) needle. Excessive pressure was not required to push the OG1448 through the needle. The 44.5 mg / mL solution was tested for injectability in the laboratory and was then able to be pushed through the needle without undue pressure by a female operator. [000480] Example 49.- Storage Stability [000481] A current stability study was carried out using OG1448 reference lot R5606 at 44.5 mg / ml in PBS pH 7.34 (as described above). Three samples were chosen for the study: room temperature (RT), 4°C and -20°C. The sampling frequency is at 0,14,28, 91,181 and 362 days. Samples were evaluated by SDS-PAGE and analytical AE-HPLC for unreacted and sequestered protein and potential aggregates. Se noi 1 ηη / ι znz / R / v observed (data not shown) that OG1448 demonstrates greater than 5% protein impurity by AE-HPLC at all three temperatures up to six months, which is similar to the level at time 0. This study is currently being conducted. [000482] Example 50.- Alternative Phosphorylcholine Polymers [000483] A HEA-PC polymer was synthesized as described below. HEA-PC (2-(acryloyloxy)ethyl-2(tnmethylammon¡o)ethylphosphate), which is an acrylate as opposed to the HEMA-PC methacrylate described above, has the following structure: noi i ηη / ι znz / R / v HEA-PC HEA-PC was polymerized to the initiator shown in Figure 23 as compound L Reagent Name Amount MW Initiator Compound L (see above) 1.65 mg 2505.5 HEA-PC Monomer 0.461 g 281.24 Cu(I) Bromide Catalyst 1.2 mg 143.45 Ligand Tris[2-(dimethylamino)eth¡l]amine (Me6TREN) 2.73 mg 230.39 Solvent A Ν,Ν-Dimethylformamide (DMF) 21.85 μΙ 73.09 Solvent B Water 0.7 ml 18.02 Solvent C Methanol 0.7 ml 32.04 [000484] A stock 200 mg / ml primer solution is prepared by dissolving 2.2 mg primer in 11 µΙ dry DMF and a 200 mg / ml ligand solution is prepared by dissolving 4.6 mg Me6TREN in 23 pL dry DMF. 8.25 μΙ of the stock solution of initiator and 13.6 μΙ of ligand are dispensed into a tube. Degas at -78°C for 5 min then fill with argon and add 1.2 mg of CuBr. It is degassed and filled with argon. A concentrated solution of HEA-PC in methanol (weigh 0.461 g HEA-PC and dissolve in 0.5 mL methanol) is added to the solution inside the reactor at -78°C. The flask is rinsed with 200 μΙ of methanol and added into the reactor at -78°C and then 0.5 mL of distilled water then another 200 μΙ of water. Degas completely until no bubbling is seen and all heterogeneity is gone (solid particles dissolve or disappear). Fill with 4 psi of argon and allow the reaction to proceed at room temperature for 1 hour. The reaction was already viscous. The reaction was allowed to proceed for approximately 1 hour. A solution of bipyridine in methanol (5mg in 0.5uL) was added. Another 2-3 mL of metal was added and the catalyst was allowed to oxidize overnight at 4°C. The conversion was determined by 1H NMR and was estimated to be 94%. [000485] The next day the polymer was dialyzed and subjected to SEC / MALS analysis using a Shodex SB806M_HQ column (7.8x300mm) in 1x PBS pH 7.4 at 1 mL / min, giving a PDI of 1.157, Mn of 723.5 kDa, Mp of 820.4 kDa and Mw of 837.2 kDa (before dialysis the PDI is 1.12, Mn = 695 kDa, Mp = 778 kDa). Then, a maleimide functionality was added to the polymer so that a protein, including TAF443, can be conjugated. [000486] The maleimide ester Mal-PEG4-PFP (see Example 23 above) was then pressed onto the HEA-PC polymer as shown in Example 23. The resulting maleimide-functional HE-PC polymer was then can be conjugated to a sulfhydryl group as discussed herein for HEMA-PC polymer. [000487] An acrylamide-PC polymer was also produced using the monomer 2-(acrylamyl)ethyl-2(trimethylammonium)ethylphosphate (Am-PC), which has the following structure noi i ηη / ι znz / B / v Am-PC [000488] Am-PC was used for polymerization using a 3-arm initiator (a TFA salt) having the structure: ηαι [000489] The synthesis of Am-PC polymer was carried out as follows: Reagent Name / Identity Quantity MW Primer 3-arm primer (see above) 2.2 mg 885.35 Am-PC monomer 0.5 g 280.26 Catalyst (I) Copper (I) Bromide 1 mg 143.45 Catalyst (II) Copper (II) Bromide 0.2 mg 223.35 Ligand T r¡s[2-(d¡methylamino)eth¡l]amine (Me6TREN) 3.94 mg 230.39 Solvent A Ν,Ν-Dimethylformamide (DMF) 31.7 pl 73.09 Solvent B Water 1 ml 18.02 Solvent C Methanol 1 mi 32.04 [000490] A stock solution of ligand at 200 mg / mL was prepared by dissolving 9 mg of Me6TREN in 45 uL of dry DMF. 19.7uL of the concentrated solution are added to a reaction vessel. A 200mg / ml stock stock solution is prepared by dissolving 6.5mg of material in 32.5uL of DMF. They are added to 11uL of the concentrated initiator solution to the ligand of the above. Degas for 5 min. 1mg of CuBr is added. A stock solution of 200mg / ml CuBr2a is prepared by dissolving 4mg of CuBr2 in 20 pL of DMF. 0.5g of monomer (AmPC) is added to 1 mL of methanol (slow dissolve / viscous solution), followed by 1uL of the concentrated CuBr2 solution. Add the monomer solution dropwise to the above reaction mixture, rinse with 1 mL of water. The reaction mixture is degassed completely (freeze-thaw). The reaction is allowed to proceed for 24 hours. [000491] The Am-PC polymer can then be dialyzed. The molecular weight of the above polymer was determined by SEC / MALS: Mn is 215kDa, Mp: 250kDa, PDI is 1.17. The conversion was estimated by 1H NMR to be 94%. A maleimide functionality can be added to the Am-PC polymer as discussed above for HEMA-PC and HEA-PC. The maleimide-functional Am-PC polymer can be conjugated to a protein, such as TAF443, as described above. [000492] Example 51.- Ellman's Inverted Test to Calculate Free Maleimide in a Compound [000493] Following the addition of maleimide functionality to the OG1801 polymer to form OG1802 (see above), an Ellman's assay was used to determine the amount of functional (ie, conjugable) maleimide in a sample. The thiol converts Ellman's reagent (DTNB) to TNB- then TNB2- in neutral and alkaline pH water, which gives a yellow color (measured at 412nm). A standard curve is established with cysteine. Since maleimide reacts with thiol, this assay actually measures the thiol (cysteine) left behind. Inhibition is calculated as the (original thiol left after addition of maleimide polymer) / (original thiol) and is expressed as a percentage. [000494] Reagents Used in Assay: A standard curve was prepared using the cysteine ​​from 62.5 μΜ to 2 μΜ. Stock polymer solutions were prepared by dissolving the powder in 1xPBS pH7.4 (reaction buffer) and mixing thoroughly. Polymer solutions of cysteines of equal molarity were mixed and allowed to react at 27°C for 30 minutes. The 150μΜ DTNB solution was added to the cysteine ​​standards and polymer / cysteine ​​reactions and the color was developed at 27°C for 5 minutes. The OD at 412nm was read in the Spectramax plate reader and the percent inhibition was calculated with the Softmax Pro Software and the standard curve was cysteine. [000495] All patent filings, websites, other publications, accession numbers and the like cited above or below are incorporated by reference in their entirety for all purposes to the same degree as if each individual point were specifically indicated individually to be incorporated as a reference. If different versions of a sequence are associated with an accession number at different times, this means the version associated with the accession number on the effective date of filing this application. The effective filing date means the earlier mass of the actual filing date or filing date of a priority application with reference to the accession number if applicable. Similarly, if different versions of a publication, website, or the like are published at different times, the most recently published version on the effective date of filing is proposed unless otherwise indicated. Any feature, step, element, mode, or aspect of the invention may be used in combination with others unless specifically indicated otherwise. Although the present invention has been described in some detail by way of illustration and example for purposes of clarity and understanding, it will be apparent that certain changes and modifications may be practiced within the scope of the appended claims. noi i ηη / ι 7Π7 / Β / Υ

Claims

1. A fusion protein comprising a vascular endothelial growth factor antagonist (hereinafter 'VEGFj') linked to a platelet-derived growth factor antagonist (hereinafter 'PDGFj'), wherein the VEGF antagonist is an anti-VEGF antibody, and the PDGF antagonist is an extracellular trap segment of the PDGF receptor (hereinafter 'PDGFRj'), wherein the PDGFR extracellular trap segment comprises the D1-D3 domains of PDGFR-β.

2. The fusion protein of claim 1, wherein the anti-VEGF antibody comprises a heavy chain and a light chain, and the heavy chain is fused via a linker to the C-terminal of the PDGFR extracellular trap segment.

3. The fusion protein of claim 1, wherein the anti-VEGF antibody is an IgG.

4. The fusion protein of claim 1, wherein the PDGFR extracellular trap segment comprises an amino acid sequence as set out in amino acids 33-314 of SEQ ID NO:

11.

5. The fusion protein of claim 1, wherein the anti-VEGF antibody is an anti-VEGF-A antibody.

6. The fusion protein of claim 2, wherein the linker has an amino acid sequence GG or has the amino acid sequence as set out in SEQ ID NO: 40 or SEQ ID NO:

41.

7. The fusion protein of claim 6, wherein the linker has the amino acid sequence as set out in SEQ ID NO:

40.

8. The fusion protein of claim 7, wherein the anti-VEGF antibody heavy chain comprises three complementarity-determining regions (CDRs) having the amino acid sequences as set forth in SEQ ID NO: 42, SEQ ID NO: 43 and SEQ ID NO: 44, respectively, and wherein the anti-VEGF antibody light chain comprises three CDRs having the amino acid sequences as set forth in SEQ ID NO: 45, SEQ ID NO: 46 and SEQ ID NO: 47, respectively.

9. The fusion protein of claim 8, wherein the heavy chain is IgG1, and the light chain is kappa.

10. The fusion protein of claim 9, wherein the heavy chain constant domain has the amino acid sequence as stated in SEQ ID NO: 17, and the light chain constant domain has the amino acid sequence as stated in SEQ ID NO:

18.

11. The fusion protein of claim 10, wherein the heavy chain constant domain has the amino acid sequence as set out in SEQ ID NO: 17 except for the presence of one or more mutations to reduce effector function.

12. The fusion protein of claim 11, wherein the mutations are for one or more of the following amino acid positions with reference to SEQ ID NO: 17: E116, L117, L118, G119, G120, A210, A213 and P214.

13. The fusion protein of claim 12, wherein the mutations are selected from the group consisting of E116P, L117V, L117A, L118A, G120A, A210G, A213S and P214S.

14. The fusion protein of claim 13, wherein the mutations are: L117A, L118A and G120A.

15. The fusion protein of claim 9, wherein the heavy chain constant domain has the amino acid sequence as set out in SEQ ID NO: 17 except for the presence of the L117A, L118A and G120A mutations, and a cisterna residue added at another position of the heavy chain constant domain by recombinant DNA technology.

16. The fusion protein of claim 15, wherein the added cysteine ​​residue is selected from the group consisting of Q230C and L326C.

17. Use of the fusion protein of claim 1 for the treatment or prophylaxis of an eye disease with a neovascular component, in a patient in need thereof.

18. Use of the fusion protein of claim 1 to inhibit VEGF-induced proliferation of primary human retinal microvascular endothelial cells (HRMVECs) in a patient in need thereof.

19. Use of the fusion protein of claim 1 to inhibit PDGF-induced proliferation of human cerebral vascular primary pericytes (HBVP) in a patient in need thereof.

20. Use of the fusion protein of claim 1 to inhibit sprouting in a co-culture of human retinal microvascular endothelial cells (HRMVECs) and human mesenchymal pericytes (HMPs) in a patient in need thereof.

21. Use of the fusion protein of claim 1 to inhibit laser-induced choroidal neovascularization in a patient in need thereof.