Abstract
Purpose:
The role of HLA matching in corneal transplantation remains controversial. We designed an observational, prospective, longitudinal clinical trial (ISRCTN25094892) to determine the influence of HLA class II matching on the risk of allograft rejection in high-risk, HLA class I matched penetrating keratoplasties (PK).
Methods:
Patients at increased risk of rejecting corneal allografts were registered for the study after giving informed consent and placed on a waiting list held by NHS Blood & Transplant (NHSBT). Tissue typing used DNA methodology to avoid the errors inherent in serological typing. All corneas were stored by organ culture for up to 4 weeks and had a minimum endothelial cell density of 2200 cells/mm². When corneas from a tissue-typed donor became available, the HLA type was compared with those of patients on the waiting list. Patients matched at HLA class I (≤2 mismatches) were identified and then the corneas allocated by cohort minimization to patients in this group with 0, 1 or 2 HLA class II mismatches. Patients were followed for 5 years. The primary outcome measure was time to first rejection episode. Data were analyzed both by univariate and multiple regression methods (Cox proportional hazards). The level of significance was set at p<0.05. Survival estimates and relative risks (RR) are quoted with 95% confidence intervals (95% CI).
Results:
Recruitment closed with 1137 transplants. The overall rejection-free survival at 5 years was 60% (95% CI 56, 63; n=1072). Univariate Kaplan-Meier rejection-free survivals for 0, 1 and 2 HLA class II mismatches were, respectively, 60% (95% CI 51, 67; n=180), 63% (95% CI 57, 67; n=480), and 57% (95% CI 51, 62; n=412) (p=0.4). This lack of influence of HLA class II was confirmed in the Cox regression model (p=0.2). Recipient age had a major influence on rejection with recipients ≤40 years having a 3-fold greater risk of rejection than those over 60 years (RR 2.9, 95% CI 1.8, 4.7; p<0.0001). The next most influential factor was the number of preoperative risk factors. Recipients with ≥3 preoperative risk factors were at >2-fold higher risk of rejection than those with no risk factors (RR 2.3, 95% CI 1.5, 3.7; p=0.0004).
Conclusions:
Matching for HLA class II did not reduce the risk of rejection in high risk PK matched for HLA class I.