Factors affecting outcome following transplantation of ex vivo expanded limbal epithelium on amniotic membrane for total limbal deficiency in rabbits.

PURPOSE To determine factors affecting the outcome of corneal surface reconstruction in rabbits with total limbal stem cell deficiency (LSCD), by using autologous limbal epithelial stem cells (LSC) ex vivo, expanded on rabbit amniotic membrane (AM). METHODS Left eyes of 52 rabbits were rendered totally limbal stem cell deficient by n-heptanol debridement of the entire corneal epithelium followed by surgical removal of 360 degrees of limbal rim. After cytologic verification of LSCD, the fibrovascular pannus of each cornea was removed. Group I (n = 10) received a rabbit AM transplant, whereas groups II, III, and IV (n = 42) underwent transplantation of LSCs cultured on rabbit AM (LSC-AM graft) derived from a small limbal biopsy specimen from the right eye. Clinical outcome was graded as a success if a smooth, avascular corneal surface was restored, a partial success if more than two quadrants of corneal surface were smooth, or a failure if the corneal surface was revascularized and irregular. RESULTS A long-term follow-up of more than 1 year was achieved. Compared with the 100% failure rate in group I, inclusion of expanded LSCs resulted in variable success rates in groups II, III, and IV (all P < 0.001). Kaplan-Meier survival analysis showed that different suturing techniques, subconjunctival injection of long-acting steroid, and tarsorrhaphy used in groups II (n = 17) and III (n = 13) did not significantly alter the outcome (P = 0.89). However, the use of a larger graft and human AM as a temporary patch with the explant retained for 1 week in group IV (n = 12) significantly improved the success rate to 83% (P = 0.002). Among eyes showing clinical failure, there was a significant correlation between the logarithm of the first day when an epithelial defect was noted and the time of graft failure (r(2) = 0.60, P < 0.001). Furthermore, the presence of severe lid deformity was borderline significant when correlated with failure cases in all four groups (P = 0.069). CONCLUSIONS Ex vivo expansion of LSCs can be achieved by using rabbit AM culture. Such expanded LSCs can successfully reconstruct corneal surfaces affected by total LSCD. This animal model is useful to investigate culturing variables affecting epithelial stemness so that surgical reconstruction of corneas with total LSCD can be successfully performed. Furthermore, this model can be used to test the feasibility of gene therapies targeting LSCD in the future.