[Limbal allografting from related live donors for limbal stem cell deficiency secondary to ocular chemical burn]

To investigate the efficacy and complications of limbal allografting from related live donors (RL-LAT) for limbal stem cell deficiency secondary to ocular chemical burn.

Retrospective comparative case series. Limbal allografting from related live human leukocyte antigen (HLA)-matched (n = 10) and-unmatched donors (n = 2) was performed in 12 consecutive patients (12 eyes) with limbal stem cell deficiency secondary to ocular chemical burn, including 9 alkali burn and 3 acid burn; whereas allolimbal cadaver transplant was performed in another 15 patients in Zhongshan Ophthalmic Center between April 2006 to December 2007. Systemic cyclosporine A was administered to later eight recipients in living related donor group and all patients in cadaver donor group. Main outcome measures were restoration of corneal epithelium, reduction in corneal vascularization, improved corneal clarity, and visual improvement. Mean follow-up was (11.4 ± 3.6) months (from 12 to 27).

In living related donor group, ten patients (83.3%) achieved initially reepithelialize, and epithelial healing time was (9.6 ± 3.4) d. Nine patients (75.0%) achieved stable ocular surface (stable corneal epithelium, decreased vascularization of corneal surface, and reduced corneal opacification). However, in seven eyes (77.8%) achieved stable ocular surface, gradual recurrence of peripheral corneal vascularization occurred after 5 to 8 months postoperatively. Whereas in cadaver donor group, twelve patients (80.0%) achieved initially reepithelialize, and epithelial healing time was (11.2 ± 4.6) d. Seven patients (46.7%) achieved stable ocular surface. However, in all seven eyes achieved stable ocular surface, gradual recurrence of peripheral corneal vascularization occurred after 4 to 10 months postoperatively. The differences of both the rate of initially reepithelialize and epithelial healing time between these two groups showed no statistical significance (χ(2) = 1.764, P = 0.117; t = 0.116, P = 0.133, respectively). However, the differences of both the successful rate of ocular surface reconstruction and the recurrent rate of corneal vascularization during the follow-up period between these two groups were statistically significant (χ(2) = 4.158, P = 0.015; χ(2) = 3.463, P = 0.022, respectively). In living related donor group, features of graft rejection developed in two patients with postoperative stable ocular surface, Both rejected recipients had one among HLA-matched and one among-unmatched respectively. Whereas in cadaver donor group, features of graft rejection developed in three patients with postoperative stable ocular surface. None of the donor eyes had any complications.

Limbal allografting from related live donors is more effective to reconstruct the ocular surface of limbal stem cell deficiency secondary to ocular chemical burn than allolimbal cadaver transplant. HLA-matched limbal allografting could reduce postoperative graft rejection and systemic cyclosporine A is useful in ensuring long-term survival.