Multivariate analysis of risk factors in cadaver donor kidney transplantation

Data collected prospectively on 3811 kidney transplants performed between June 1977 and July 1982 with follow-up to July 1984 by the 42 member institutions of the South-Eastern Organ Procurement foundation were analyzed to identify factors associated with graft and patient outcome in patients not receiving cyclosporine. Multivariate Cox regression analysis was used to examine the association and relative risk of 24 variables with three actuarial outcomes: overall graft failure, irreversible rejection, and patient death. Factors having no suggested association with any outcome included: recipient sex, history of pregnancy, blood group, and time on dialysis; organ preservation method, time and source; donor race; crossmatch test sensitivity; and annual center transplant rate. In decreasing order of relative risk, the factors most significantly associated with irreversible rejection were: loss of two or more prior grafts, low HLA-A,B match, lack of pretransplant blood transfusion, high (greater than 60%) pretransplant sensitization to leukocyte (HLA) antigens, and delayed graft function. Splenectomy, insulin-dependent diabetes, and antilymphocyte serum therapy provided the greatest risk of patient death. Factors such as recipient age, race, and native nephrectomy had suggested associations with outcome. By adding each center as a separate covariate in the analysis, other center-dependent factors were quantitated and found in some cases to have a highly significant association with graft and patient outcome. These results provide a basis for evaluating the potential risk of graft loss or patient death for those prospective cadaver kidney transplant recipients not being considered for cyclosporine therapy.