A, A modified Boyden chamber migration assay was used to assess the effect of r84, bevacizumab (bev) on VEGF-induced endothelial cell (EC) migration. 20,000 HDMEC, PAE-KDR, PAE-Flt-1 cells were plated on 8.0 µm cell culture inserts and allowed to migrate overnight towards SFM or human VEGF (100 ng/mL) +/− 500-fold molar excess antibody (bev, r84, control IgG). r84, bev block the VEGF-induced migration of VEGFR2-expressing ECs (A, HDMEC, PAE-KDR). Bev blocks VEGF-induced migration of endothelial cells expressing VEGR1, but r84 does not (A, PAE-Flt-1). Western blots of VEGF-induced signaling in PAE-KDR (B), HDMEC (C), and PAE-Flt-1 (D) lysates following stimulation of cells with 50 ng/mL human VEGF +/− 500-fold molar excess antibody (bev, r84, control IgG). r84 and bev block p-VEGFR2 and downstream phosphorylation (p38, PLCγ, ERK1/2) (B, C), but only bev blocks VEGF-induced VEGFR1 phosphorylation in PAE-Flt-1 stimulated cells (D). *p<0.05, **p<0.01, ***p<0.001, statistical differences in A compared to VEGF alone, unless otherwise indicated.

Frozen sections of A549, H460, H1299 tumors treated with control IgG (Ctrl), r84, or bevacizumab (bev) were analyzed by immunofluorescence for endothelial, pericyte (PC), and lymphatic markers, as well as for VEGFR2 expression. Number of positive-staining entities per high powered field was evaluated using Nikon Elements software. A, r84, bev treatment significantly decreases tumor MVD, shown by a reduction in tumor endomucin positive endothelial cells (red). r84, bev treatment induced a trend towards increased NG2 positive (green) PC coverage of vessels as compared to Ctrl. B, r84, bev treatment significantly reduces the number of VEGFR2 positive cells in H460, H1299 tumors as shown by RAFL-2 staining (red). Only r84 treatment significantly reduced VEGFR2 staining in A549 tumors. C, r84, bev treatment significantly decreased H460, H1299 tumor LVD, as indicated by a reduction in lyve1 positive cells (red). Only r84 treatment significantly reduced A549 tumor LVD. *p<0.05, **p<0.01, ***p<0.001, statistical differences compared to Ctrl, unless otherwise indicated.

A, Recombinant human VEGF coated at 0.5 µg/mL was detected with a titration of fully human monoclonal antibody r84. r84 bound to VEGF was detected with an anti-human Fc HRP-conjugated antibody, demonstrating r84 binds both human and mouse VEGF-A (open squares and circles, respectively). B, Recombinant human and mouse VEGF-A, and human VEGF-B, -C, -D, and PlGF coated at 0.5 µg/mL was detected with r84 at 1 µg/mL. Binding of r84 to VEGF family member was detected as in A, demonstrating r84 binds only human and mouse VEGF-A and not other VEGF family members. VEGFR2 (C) and VEGFR1 (D, left panel) coated at 1 µg/mL were incubated with 4.76 nM or 2.38 nM biotinylated VEGF, respectively, +/− the indicated fold excesses of antibody (Control IgG, 2C3, r84). r84, 2C3 specifically block biotinylated-VEGF binding to VEGFR2 (C), but not VEGFR1 (D, left panel). In contrast, a 500-fold molar excess bevacizumab (bev) reduces biotinylated-VEGF binding to VEGFR1, compared to biotinylated-VEGF alone or plus r84 (D, right panel).

2.5 million human NSCLC cells were injected subcutaneously into the right flank of NOD/SCID mice. Therapy began one day post tumor cell injection (TCI), and continued for 4–8 weeks. Tumor volumes were measured twice/week and final tumor weights were recorded at sacrifice. r84 and bevacizumab (bev) similarly control tumor growth and final weight compared to control IgG treatment in H460, H1299 models (A, B). C, In A549 NSCLC tumor bearing animals, r84's control of tumor growth was significantly different from bev (p<0.05). D, Titration of antibody dosing in A549 tumor xenografts showed no change in tumor growth and final tumor weight with increasing doses of bev, however there was increasing control of tumor growth with increasing doses of r84, with doses of r84 controlling growth better than bev (D). A–C, n = 8 mice per group; D, n = 6 mice per group. *p<0.05, **p<0.01, ***p<0.001, statistical differences compared to Ctrl treatment, unless otherwise indicated.

5 million PANC-1 tumor cells were injected subcutaneously into NOD/SCID mice. Tumor bearing (TB) and non tumor bearing (NTB) mice received long-term 12-week therapy with 50 mg/kg/week r84 or a control IgG. Following 12-weeks of therapy, animals were sacrificed and organs and blood were collected for toxicity analysis. n = 5 animals per group. A, Hematoxylin and eosin staining of formalin-fixed, paraffin-embedded kidney and liver sections demonstrated that control of tumor growth is achieved without induction of kidney or liver histopathologic changes. B, Blood chemistry analysis of serum samples collected from mice at sacrifice indicated that r84 treatment does not induce changes in serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), or blood urea nitrogen (BUN) as compared to control IgG therapy and to age-matched Naïve animals that did not have tumor and never received antibody therapy. Immunocompetent Kras/INK4a mice that spontaneously develop pancreatic cancer were treated for eight weeks with saline, 25 mg/kg/week mcr84, or 50 mg/kg/week sunitinib. After 2 weeks of therapy, mcr84 and sunitinib significantly increased mean systolic blood pressure (C, left panel, ***p<0.001), but this effect was lost by week 7 of continuous therapy (C, right panel). D, Urine samples collected during week 6 of therapy and assayed for total levels of urine protein and creatine (Upro/Creatine ratio displayed) showed no significant difference between treated animals as compared to control, indicating that extended therapy with mcr84 and sunitinib did not induce kidney damage in this model.