Prospective trial of a predictive algorithm to transplant cadaver kidneys into highly sensitized patients

The difficulty of transplanting sensitized patients increases proportionally to the panel reactive antibody (PRA) titer. Because of the high likelihood of a positive final crossmatch, these patients are often excluded from a prospective transplant unless there is a 0 HLA-A, -B, -DR mismatch. To address this problem, we developed a computerized algorithm, termed the Kentucky Antibody Testing System (KATS), that predicts class I HLA antigens that would be both "unacceptable" and "acceptable" to the recipient. This report describes the results of a prospective trial among voluntarily participating centers that agreed to share kidneys based on the KATS predictions for patients whose PRA exceeded 40%.

The results of three antibody screens on each patient were compared with the HLA phenotypes of the cells in the panel in 2x2 tables with calculation of chi-square and correlation coefficient statistics. Private, broad, and public antigens were identified and a list of acceptable and unacceptable antigens were entered into the UNOS computer for each patient listed in the KATS sharing algorithm.

Of the total 418 patients meeting the inclusion/exclusion criteria, the largest single group submitted was Black-not-of-Hispanic-origin females. The mean PRA of the patients was 72%. The first transplant via KATS allocation was performed on March 8, 1997. Between that time and the last transplant on July 31, 2000, 145 kidneys were offered to the participating centers and 48 transplants were performed. Of the many reasons listed for not accepting an offer or not transplanting the shared kidney into its intended recipient, only two occurred because of a positive T cell crossmatch and six because of a positive B cell crossmatch. As compared to all other high PRA patients within Southeastern Organ Procurement Foundation who were transplanted during the study period, they were more likely to be non-Caucasian, to be less well matched for private HLA-A, B, and DR antigens, and to have waited for a longer time than the other groups. Although there was a higher incidence of delayed graft function, there was no significant difference in cold ischemia, rejection episodes, or patient or graft survival.

We conclude that KATS, or some other system to prospectively identify a list of acceptable and unacceptable HLA antigens, could improve the access of highly sensitized patients to a successful kidney transplant.