Tissue engineered endothelial keratoplasty in rabbit: tips and tricks

Purpose To report a detailed surgical procedure of tissue engineered endothelial keratoplasty (TEEK) in a rabbit model and its postoperative evaluation. Methods TEEKs were prepared 7 days before transplantation by seeding human or rabbit corneal endothelial cells on either femtosecond laser-cut ultrathin human stromal lamellae (fs-UTSL) or femtosecond laser-cut human anterior lens capsule (fs-HALC). Thirty transplantations were performed on aphakic eyes. Recombinant tissue plasminogen activator (rTPA) was used throughout the surgery. The native endothelium was removed by full-surface scraping and central descemetorhexis. The transplantation was performed as a human Descemet's membrane endothelial keratoplasty. Controls included Descemetorhexis only and transplantation of carrier alone. Postoperative follow-up was performed by slit lamp and optical coherence tomography, followed by histology. Results Controls remained oedematous. No fibrin occurred during surgery. All but three TEEKs adhered immediately. One/6 fs-UTSL and 9/16 fs-HALC cleared perfectly (p = 0.161). All failures could be explained by at least one of the following causes intraoperative bleeding, vitreous prolapsus, early partial detachment, postoperative irido corneal synechiea/angle closure. Presumed immune rejection was observed in three rabbits only after 4 weeks. Immunostaining with anti-human CD166 allowed to perfectly differentiate human cells from rabbit cells. In successful TEEK at 3 or 4 weeks, human cells formed a normal endothelium and started migrating outside the carrier. Conclusion Though the transplantation of a TEEK in rabbits is a complex model with many causes of failure, established procedure including use of rTPA allows reliable preclinical study. In addition, we suggest that fs-HALC might be a potential carrier for TEEK.