(T1330-01-01) Long-Term Ocular Delivery of Griseofulvin for Therapy of Choroidal Neovascularization

Purpose: Current treatments for choroidal neovascularization (CNV), the key pathobiological feature of neovascular “wet” age-related macular degeneration (AMD) based on anti-vascular endothelial growth factor (anti-VEGF) therapy are ineffective in a significant fraction of patients. In the quest of alternative targets, we identified ferrochelatase (FECH) as a promising target. FECH is a heme synthesis enzyme and can be inhibited by a metabolite of griseofulvin (GRF), a microtubule inhibitor. We have developed an injectable and long-acting microparticle (MP) formulation for sustained delivery of GRF.

Methods: GRF was incorporated in poly(lactic-co-glycolic acid) (PLGA) MPs with optional addition of Mg(OH)₂ to control drug release kinetics. GRF and PLGA were added to the organic (O) phase, and Mg(OH)₂ (0-20 wt%) was added to the internal aqueous (W1) phase. The effects of Mg(OH)₂ content on particle size (laser diffraction), drug loading, morphology (scanning electron microscopy), and in vitro release were evaluated. The antiproliferative effect of MPs was evaluated on human retinal endothelial cells (HRECs) by the AlamarBlue assay. In vivo efficacy was evaluated in a mouse model of laser-induced CNV (L-CNV).

Results: The mean particle size (D50) and loading efficiency of MPs were 15-22 µm and 2.7-5.2 wt%, respectively (Figure 1a, 1b). MPs were spherical with sizes consistent with the laser diffraction measurement. MPs prepared without Mg(OH)₂ had a solid surface, whereas those with Mg(OH)₂ showed porous surfaces. The porosity increased with the amount of Mg(OH)₂ added to the MPs (Figure 1c). MPs without Mg(OH)₂ released ~15% of loaded drug in phosphate-buffered saline (pH 7.4) with 0.2% Tween 80 at 37 °C in 1 day and slowly released the remainder (up to ~80%) in 60 days (Figure 1d). The inclusion of Mg(OH)₂ dramatically changed the GRF release profile. MPs with 20 wt% Mg(OH)₂ showed a high burst release of GRF (~70%), followed by a complete release in 5 days. MPs with 10% and 5 wt% Mg(OH)₂ showed >90% and >75% drug release in 5 days. MPs containing 2 wt% Mg(OH)₂ showed a minimal burst release (~18% in 1 day) followed by continuous release (~100% in 38 days) and were considered most suitable for 1 month GRF delivery. The GRF MPs with 2 wt% Mg(OH)₂ inhibited HREC proliferation over 1 month (Figure 1e). In a mouse model of L-CNV, treatments were given by intravitreal injection at -28, -14, or 0 days from the induction of L-CNV (Figure 2a). Fundus imaging, optical coherence tomography, fluorescein angiography, and ex vivo agglutinin staining of the -28 days cohort suggest a reduction in lesion volume by GRF MPs (Figure 2b). Confocal image analysis showed that GRF MPs were effective at all time points (but free GRF was effective only when it was given at the time of laser), indicating that the MPs provided a sustained effect in reducing CNV volume, unlike free GRF (Figure 3).

Conclusion: GRF-loaded PLGA MPs were developed for long-term ocular delivery. The MP formulation provided a sustained release of bioactive GRF over a month and effectively reduced the L-CNV lesions in mice.

References: 1. Chobisa D et al. Investigative Ophthalmology & Visual Science. 62(8). 2021. E-abstract 199.
2. Basavarajappa HD et al. EMBO Molecular Medicine. 2017. 786-801.

Acknowledgments:

Funding: This work was supported by Dr. Reddy’s Laboratories-Purdue University fellowship.
Disclosures: YY, DC, and TWC- PCT/US2021/038792; TWC- 15/009,339; 62/837,506.

Figure 1. (a) Mean particle size of MPs with varying amounts of Mg(OH)2 (n=3). (b) Loading efficiency of MPs with varying amounts of Mg(OH)2 (n=3). (c) SEM images of MPs with varying amounts of Mg(OH)2. (d) In vitro release of MPs with varying amounts of Mg(OH)2 (n=3 per concentration per time). (e) Long-term proliferation assay of HRECs: GRF MPs (previously subjected to drug release in growth media and collected at specific time points) were incubated with HRECs and tested for HREC proliferation (n=3). Data in (e) were compared by two-way ANOVA followed by multiple comparisons; significance level 0.05.



Figure 2. (a) Timeline of laser-induced choroidal neovascularization (L-CNV) mouse study: 0 day, -14 days, and -28 days cohorts stand for the day of treatment relative to the L-CNV induction performed at day 0. Fundus imaging and optical coherence tomography were done at day 7 and 14 in situ. Fluorescein angiography (FA) was done at day 14 on live animals. Choroidal lesions were measured by ex vivo confocal imaging, (b) L-CNV, fundus (day 7 and 14), OCT (day 7 and 14), FA, and confocal lesion images of -28D cohort. Red lines on fundus images indicate OCT planes. Scale bars = 100 µm.



Figure 3. Volume of CNV estimated by confocal images of agglutinin-stained lesions for: (a) -28D cohort, (b) -14D cohort, and (c) 0D cohort. Data in (a), (b), and (c) were compared by one-way ANOVA with multiple comparisons; significance level 0.05.