Simultaneous Ang-2/VEGF-A inhibition prevents subretinal fibrosis progression in preclinical mouse models of choroidal neovascularization (CNV)

Abstract

Purpose : Faricimab, a bispecific antibody (ab) currently in phase (Ph) 3 trials, targets both angiopoietin-2 (Ang-2) and VEGF-A, key drivers of vascular instability, and demonstrated improvement in BCVA and sustained efficacy vs anti-VEGF monotherapy in Ph2 clinical trials of diabetic macular edema (DME) and neovascular AMD (nAMD). We present new preclinical data from 2 CNV mouse models showing that dual Ang-2/VEGF-A blockade significantly reduces fibrosis vs anti-VEGF monotherapy.

Methods : 7-week (W)-old JR5558 mice developing bilateral spontaneous neovascular fibrotic lesions were treated intraperitoneally (ip) with mouse cross-reactive tool abs against Ang-2, VEGF-A, or both (bispecific anti–VEGF-A/Ang-2 ab), and IgG (control). Fibronectin (FN) immunostaining was performed on retinal pigment epithelium (RPE)/choroid flat mounts to assess fibrosis at 1 (PT1), 3 (PT2), and 5 (PT3) W post treatment. Fluorophore-labeled collagen hybridizing peptides (CHPs) were used to detect collagen remodeling in active fibrotic lesions. Effect of Ang-2/VEGF-A blockade on subretinal fibrosis was analyzed in a laser-induced CNV mouse model by treating 10–12W-old wild-type mice ip with tool abs directly after and at 1W post laser injury. FN- and CHP-positive (+) areas were analyzed 3W post laser injury.

Results : There was significant reduction in FN+ area in the RPE/choroid of JR5558 mice with the bispecific ab (38%, P<0.01) and anti–Ang-2 (41%, P<0.001) vs IgG at PT1; effect of anti–VEGF-A alone was not significant. Only dual Ang-2/VEGF-A inhibition maintained significant reduction at PT2 (47%, P<0.01) and PT3 (54%, P<0.05). Dual inhibition significantly prevented collagen remodeling, as shown by reduced CHP in RPE/choroid lesions at PT2 (66%, P<0.01). Anti–VEGF-A or anti–Ang-2 alone showed no significant effect. In the laser-induced CNV mouse model, dual inhibition significantly reduced both FN+ (47%, P<0.05) and CHP+ (39%, P<0.01) areas at 3W; blocking VEGF-A or Ang-2 alone had no significant effect.

Conclusions : We present preclinical data from 2 independent CNV mouse models suggesting sustained prevention of scarring with dual Ang-2/VEGF-A pathway inhibition. Our findings support the hypotheses that Ang-2 and VEGF-A drive subretinal fibrosis. Future studies are underway to determine how dual Ang-2/VEGF-A inhibition limits subretinal scarring.

This is a 2021 ARVO Annual Meeting abstract.