Search

We determined the human leucocyte antigen (HLA)-A, -B and -DR allele frequencies in recipients and donors of 115 cornea transplants, for recipients who developed graft rejection and those who did not. No difference in HLA allele frequencies of the recipients was found. The frequencies of the HLA-A26, -B35 and -B44 alleles in cornea donors were increased in recipients who developed graft failure. The detrimental effect on corneal graft survival of these alleles was significant (p < 0.001). No such effect was observed in renal transplantation. Corneal graft survival was similar when one or two A26, B35 or B44 alleles were present on the donor cornea. The negative effect was similar in magnitude to the previously reported negative effect of an HLA-B locus match between donor and recipient. When both a B-locus match and an A26, B35 or B44 allele were present, the negative effect on graft survival was twice as strong, indicating that different immune mechanisms are responsible for these phenomena.

The results of tissue typing on 115 recipient/donor pairs prior to corneal grafting were analyzed with the proportional hazard regression model for the incidence of the first rejection episode and for graft failure from rejection. Like other investigators, we found that a previously failed corneal graft and the degree of recipient corneal vascularization were significant risk factor for graft rejection. ABO blood group matching had no effect. The absence of mismatches in both the HLA-A and HLA-DR loci decreased the incidence of rejection. However, no difference was observed for the presence of one versus two mismatches. Regression results for the HLA-A and DR loci were not significant. Surprisingly, matching for one or both HLA-B alleles resulted in a significantly higher incidence of graft rejection episodes (P < 0.005) and of graft failure (P < 0.052). This adverse matching effect for the B locus was proportional to the number of mismatches.

This paper reviews the clinical post-operative management of keratoplasty and the management of corneal graft rejection. In both instances corticosteroids remain the mainstay of treatment; however, the literature shows a wide range for both route and frequency of administration. Grafts at 'high risk' require more immunosuppressive therapy, but no universally accepted definition of high risk exists and consequently different treatment regimens are difficult to compare and evaluate. Studies using univariate and multivariate survival analysis suggest that recipient corneas can be divided into low, medium and high risk depending on the number of quadrants of vascularisation (avascular, 1-2 quadrants and 3+ quadrants respectively). This wider classification would make the devising and comparing of treatment regimens more consistent. In high-risk cases, corticosteroids alone provide insufficient immunosuppression and systemic cyclosporine is needed in exceptional cases. When managing rejection episodes, a severe reaction involving the endothelium often does not respond to topical steroids alone, and systemic corticosteroids are required. Instead of oral steroids, we now prefer to use an intravenous 'pulse' of 500 mg methylprednisolone: this is at least as effective, avoids prolonged medication, and may confer some long-term benefit.

Systemic cyclosporine (CSA) was given to 43 high-risk keratoplasty patients (vascularisation in three or four quadrants and > 16 stromal vessels): 14 received CSA for 12 months (long CSA group) and 29 for a 4-6 month period (short CSA group). A group of 37 similar high-risk keratoplasty patients received no CSA (no CSA group). In the no CSA group 27 (73%) grafts had rejection episodes compared with 21 (48.8%) in the combined CSA group (p = 0.025). Rejection was reversed in only 23.3% of the no CSA group, compared with 50% and 87.5% in the short (p = 0.06) and long (p = 0.002) CSA groups respectively. Compared with the no CSA group, overall graft survival was better in both the short (p = 0.019) and long (p = 0.0056) CSA groups. This improvement resulted from both a reduction in the incidence, and an increase in the reversal rate, of rejection episodes. The improvement continued after stopping CSA, suggesting that some immunological privilege had been re-established. Acuities of 20/40 or better were achieved by 44.8% and 50% of eyes in the short and long CSA groups respectively, compared with 13.5% of eyes in the no CSA group.

In a previous study, we reported that a single intravenous pulse of 500 mg of methylprednisolone is more effective than oral prednisone in reversing rejection in patients who present early with severe endothelial rejection. In addition, the grafts that survived were less likely to undergo a further rejection episode. This study investigates whether a second pulse given at 24 h (n = 31) or at 48 h (n = 29) has any advantage over a single pulse (n = 30). All three treatment regimens were equally effective in reversing rejection (74.2%, 79.3%, and 83.3%, respectively), and there was no statistical difference in graft survival. An additional pulse therefore appears to offer no advantage over single pulse therapy. In high-risk grafts, current therapy has been reported to be less effective in reversing graft rejection. This study showed no significant difference between high- and low-risk grafts, indicating that pulse therapy may have a particular role in these difficult cases.

In corneal graft rejection, rapid reversal of the rejection process is necessary to minimise endothelial cell loss. Ten consecutive patients with acute endothelial rejection were treated with a single 500 mg pulse of methylprednisolone intravenously and topical prednisolone 1% drops hourly. The rejection episode was successfully reversed in eight (80%) of the 10 grafts. This preliminary trial indicates that cortico-steroid pulse therapy may be beneficial in the management of severe corneal graft rejection with the advantage of avoiding prolonged oral corticosteroid therapy.

Four patients (six eyes) underwent successful corneal transplantations while receiving systemic cyclosporin. The six eyes had marked corneal vascularization preoperatively and in four, previous grafts had rejected. These eyes were therefore at high risk with respect to the likelihood of rejection and subsequent graft failure. All six grafts have remained clear (mean follow up twelve months) and only one minor rejection episode has occurred: this was easily reversed. The patients were monitored by repeated serum creatinine and trough serum cyclosporin levels. No significant side effects were encountered.

An animal model for corneal graft rejection is described in which corneal neovascularization is induced prior to keratoplasty. This model closely resembles high-risk human corneal grafting, and relies on rejection being initiated solely by the transplanted corneal tissue. All eleven allografts performed using this model rejected; median time to rejection was 17 days. Two treatment regimens were assessed using this rejection model. Topical cyclosporine produced a significant improvement in graft survival (P = 0.0025). When systemic cyclosporine was administered there was a highly significant improvement in graft survival (P = 0.0001).

Twenty-three patients with bacterial keratitis had penetrating keratoplasties performed for deep indolent ulceration or descemetoceles, during the acute period. The period in hospital (17.6 days) was significantly lower than for a control group (35.4 days) who were treated medically and had subsequent grafts. The number of grafts remaining clear was similar, 70% and 72% respectively. Eleven patients (48%) of those who had an acute graft achieved corrected visual acuities of 6/12 or better. No cases of reinfection occurred and no eyes were lost.

In a study of three groups of high-risk keratoplasty patients, Group A (n = 19) received topical corticosteroids only; two grafts (10.5%) survived. Group B (n = 20) received both topical and systemic corticosteroids; one graft (5%) survived. There was no significant difference between Group A and Group B (P = .8095). Group C (n = 18) received systemic cyclosporine in addition to topical and systemic corticosteroids; 16 grafts (88.9%) survived, which was a significant improvement compared to results in the other two groups (P = .0001). No patient suffered any permanent side effects related to the use of cyclosporine.