Intravitreal Metformin Protects Against Choroidal Neovascularization and Light-Induced Retinal Degeneration

Abstract

Neovascular age-related macular degeneration (nAMD), a leading cause of blindness in older adults, presents a challenging pathophysiology involving choroidal neovascularization (CNV) and retinal degeneration. Current treatments relying on intravitreal (IVT) administration of anti-angiogenic agents are costly and of moderate effectiveness. Metformin, the common anti-diabetic drug, has been associated with decreased odds of developing AMD. Studies have shown that metformin can mitigate cellular aging, neoangiogenesis, and inflammation across multiple diseases. This preclinical study assessed metformin's impact on vessel growth using choroidal explants before exploring IVT metformin's effects on laser-induced CNV and light-induced retinal degeneration in C57BL/6J and BALB/cJ mice, respectively. Metformin reduced new vessel growth in choroidal explants in a dose-dependent relationship. Following laser induction, IVT metformin suppressed CNV and decreased peripheral infiltration of IBA1⁺ macrophages/microglia. Furthermore, IVT metformin protected against retinal thinning in response to light-induced degeneration. IVT metformin downregulated genes in the choroid and retinal pigment epithelium which are associated with angiogenesis and inflammation, two key processes that drive nAMD progression. These findings underscore metformin's capacity as an anti-angiogenic and neuroprotective agent, demonstrating this drug's potential as an accessible option to help manage nAMD.

Keywords: age-related macular degeneration; choroidal neovascularization; intravitreal injection; metformin; retinal degeneration.

Figure 1

Metformin treatment significantly reduces choroidal sprouting area. Representative reverse phase brightfield images of control (A), 1 mg/mL (B), and 5 mg/mL metformin (C) treatment on Day 4; scale bar = 1 mm. Metformin reduced choroidal sprouting angiogenesis in a dose-dependent manner (D); n = 6–14 per group. Comparisons were made using Brown-Forsythe and Welch ANOVA followed by Dunnett’s T3 multiple comparisons test.

Figure 2

Effect of intravitreal metformin treatment on retinal morphology and thickness. 12-week-old, male C57Bl/6 mice received 0.5 μL intravitreal (IVT) administration of either vehicle control (PBS) or 20 μg/μL metformin (10 μg total) (n = 5 per group). After 7 days, eyes were enucleated, cross-sectioned, and stained with hematoxylin and eosin. Representative 20×-magnified images of vertical cross-sections are shown (A). The inner nuclear layer (INL) (B), outer nuclear layer (ONL) (C), and total retinal thickness (TRT) (D) were measured at four locations around the optic nerve: 100, 200, 300, and 500 µm in opposite directions. Eyes receiving IVT metformin had normal histology and similar retinal thicknesses without evidence of toxicity compared to eyes receiving vehicle control. Data are presented as mean ± SEM. Statistical testing was performed using Student’s t-test for comparison.

Figure 3

Intravitreal metformin suppresses laser-induced choroidal neovascularization (CNV) lesion size and IBA1⁺ microglia infiltration. Twelve-week-old male C57BL/6 mice underwent laser induction of CNV and immediate intravitreal (IVT) injection of vehicle control or metformin. Choroid/RPE flatmounts were made 7 days after the procedure and stained for Isolectin B4 (red) and IBA1 (green). Representative 20× confocal images of mice given IVT 0.5 μL sterile vehicle control (A,C), IVT 0.5 μL of 10 μg/μL metformin solution (5 μg total) (B), and IVT 0.5 μL of 20 μg/μL metformin solution (10 μg total) (D); scale bar = 80 μm. Lesion area (E), average IBA1 intensity (F), and number of IBA1⁺ macrophages/microglia (G) were compared between IVT metformin 5 μg and IVT vehicle. Lesion area (H), average IBA1 intensity (I), and number of IBA1⁺ macrophages/microglia (J) were compared between IVT metformin 10 μg and IVT vehicle. Experimental groups consisted of n = 16 mice per group; n = 4 lesions per eye. Data are presented as mean ± SEM. Statistical testing was performed using Student’s t-test for comparison; * p < 0.05, *** p < 0.001, n.s. = not significant.

Figure 4

Intravitreal metformin treatment alters chorioretinal gene expression. Twelve-week-old male C57BL/6 mice received laser induction of choroidal neovascularization (CNV) and immediate intravitreal (IVT) injection of either 0.5 μL vehicle control or 20 μg/μL metformin solution equal to 10 μg total (n = 4 mice per group). Non-lasered mice (n = 6) were used as negative controls. Gene expression of Ampka and Ampkb (A), as well as Dll4, Iba1, Il6, Nos3, Tnf, and Vegfa (B) was assessed using qRT-PCR, with expression levels normalized to Gapdh. Data are presented as mean ± SEM. Statistical testing was performed using Student’s t-test for comparison; * p < 0.05.

Figure 5

Intravitreal metformin treatment protects against light-induced retinal degeneration. Twelve-week-old, male BALB/cJ mice received 0.5 μL intravitreal (IVT) administration of either vehicle control (PBS) or metformin 20 μg/μL (total 10 μg) 6 h prior to dark adaptation. After 16 h of dark adaptation, mice were exposed to 2 h of light at an intensity of 7500 lux/hour (n = 5 per group). After 7 days, eyes were enucleated, cross-sectioned, and stained with hematoxylin and eosin. Representative 20×-magnified images of vertical cross-sections are shown (A). The inner nuclear layer (INL) (B), outer nuclear layer (ONL) (C), and total retinal thickness (TRT) (D) were measured at four locations around the optic nerve: 100, 200, 300, and 500 µm in opposite directions. Eyes receiving IVT metformin demonstrated thicker ONL and TRT compared to eyes receiving vehicle control. Data are presented as mean ± SEM. Statistical testing was performed using Student’s t-test for comparison; * p < 0.05.