Time dependency of factors affecting renal allograft survival

The function of renal transplants can deteriorate at any time posttransplan t, but the risks and mechanisms may differ at different times posttransplan t. Survival of 522 consecutive cadaveric renal transplant recipients follow ed for at least 6 mo were analyzed, with patient death censored. The overal l risk factors in univariate analysis were acute rejection requiring antibo dy therapy (AR), delayed graft function, elevated serum creatinine at 6 mo, high panel-reactive antibodies, and donor age greater than or equal to 55 yr, with borderline effects of recipient age and female fender. These risks were studied in each of three intervals posttransplantation: less than or equal to 6 mo, 6 mo to 5 yr, and >5 yr. Of the 135 graft failures, 53 occur red less than or equal to 6 mo, 61 between 6 mo and 5 yr, and 21 beyond 5 y r. By multivariate analysis, the risks for graft failure in interval less t han or equal to 6 mo were AR (hazard ratio (HR) = 4.86, P < 0.001); delayed graft function (HR = 1.47, P = 0.06); and high panel-reactive antibodies ( HR = 2.04, P = 0.03). Between 6 mo and 5 yr, the risks for graft loss were AR (HR = 2.87, P < 0.001) and serum creatinine at 6 mo greater than or equa l to 150 mu mol/L (HR = 3.69, P < 0.001). Beyond 5 yr the risk factors were donor age greater than or equal to 55 yr (HR = 5.87, P = 0.002), with a bo rderline effect of kidneys from female donors (HR = 2.28, P = 0.07). HLA-A, -B, and -DR matching and presensitization had most of their effect through early AR and impaired function. The results indicate that risks for graft loss are time-dependent: early losses correlate with injury and rejection, but late events correlate with donor age and possibly workload.