Comparative Analysis of Intravitreal Diffusion Patterns Across Ex Vivo Human and In Vivo/Ex Vivo Animal Models

To investigate vitreous barrier function by examining ocular distribution patterns of anionic fluorescein dextran molecules with varying molecular weights (MWs) following intravitreal injection and to assess the influence of injection site and plasmin pretreatment on dye distribution.

We studied dye distribution across five MWs (3 kDa, 40 kDa, 70 kDa, 500 kDa, and 2 MDa) in an ex vivo pig eye model, with temporal or nasal injections. For experiments involving ex vivo pig eyes pretreated with plasmin, in vivo pig eyes, and ex vivo human eyes, we utilized only 40-kDa and 2-MDa dyes injected temporally.

No significant differences were observed between nasal and temporal injections. In ex vivo pig eyes, 3-kDa fluorescein showed near-complete vitreous diffusion within 24 hours. The 40-kDa and 70-kDa dyes continued to spread beyond 24 hours, whereas the 500-kDa and 2-MDa dyes demonstrated restricted distribution with minimal change up to 48 hours. In vivo pig eyes showed limited distribution of 2-MDa fluorescein. Plasmin pretreatment did not affect 2-MDa distribution but enhanced the spread of 40-kDa fluorescein. In ex vivo human eyes, 40-kDa fluorescein distributed widely, whereas 2-MDa fluorescein remained highly localized and delineated a vitreous bursa structure that the dye did not penetrate.

We observed a size-dependent relationship between dye molecules and their spread throughout the vitreous cavity, highlighting the role of vitreous as a barrier that can influence the distribution and effectiveness of treatments, including gene therapies. Future research should focus on improving the mobilization of larger substances within the vitreous while recognizing its potential as a natural reservoir for slow drug release when injected into the vitreous bursae.