Immunosuppression with a subconjunctival implant releasing dexamethasone in a rabbit model of penetrating keratoplasty

Purpose Rejection occurs in 20–30% of cases after penetrating keratoplasty (PK). Most often, its prevention relies only on steroid eyedrops but topical treatments expose to limited therapeutic adherence. Aim: to assess efficacy of continuous immunosuppression using a subconjunctival implant releasing dexamethasone (SCIRD) in a rabbit model of PK. Methods 7.5 mm PK were performed in NZW rabbits. After randomisation they received either a SCIRD (Ozurdex, Allergan) (n = 8), dexamethasone eyedrops (1 mg/ml, Dexafree, Thea) 3 times a day (n = 6), or 0.9% NaCl (n = 6). The running suture was left in place to stimulate angiogenesis. Weekly follow‐up by digital slit lamp and anterior segment OCT. Images were analysed blind to the treatment for transparency, neovessels, and central corneal thickness (CCT). After 5 or 6 weeks, animals were killed and corneas were processed for standard histology. Results Placebo group: constant early neovascularization growing to neovessels penetrating the graft on >270° after 5 weeks + rejection (opaque graft, 360° neovessels, CCT > 500 μm) in 50% of cases. Eye drops and implants groups: similar evolution, without rejection after 6 weeks and normal CCT. Moderate neovascularization occurred in 5/6 rabbits treated with eyedrops and 6/8 treated with the implant. Histology confirmed clinical diagnosis in all cases. The implants disappeared after 3–5 weeks. No adverse effect. Conclusions Despite a severe rejection model, a SCIRD was not less efficient than eye drops to prevent rejection during the first 6 weeks. These findings highlight the potential benefit of subconjunctival implants of steroids after corneal graft.