Predicting risk of acute rejection in patients with kidney transplants

Create a model to predict the risk of acute rejection of kidney transplant considering variables related to the immunosuppressant agent used, the receiver, the donor and the transplanted organ.

Cohort study in a population of 68 patients with kidney transplants being treated with tacrolimus triple therapy. Predicting the risk of acute rejection was carried out with a logistic regression analysis using age, sex, re-transplant status, number of HLA incompatibilities, cold ischaemia time, acute tubular necrosis, induction with basiliximab or thymoglobulin and treatment type as explanatory variables. The contribution of variables associated with determining the blood concentration of tacrolimus was also evaluated; these variables include the average blood concentration, the number of values below and included in the pre-defined therapeutic interval, and the time during which those values remained within that interval.

The logistic regression analysis indicates that the risk of acute rejection depends on the acute tubular necrosis (OR: 3; CI 95%, 0.7 to 13.2) and on the time that the blood concentrations of tacrolimus remains within the therapeutic interval (OR: 0.8; CI 95%, 0.7 to 0.9). The final model presents an optimal discrimination power (AUCROC: 77%; CI 95%, 62% to 92%). For the selected cut-off point (probability greater than or equal to 0.24) the model shows a sensitivity of 83% (CI 95%, 74 to 90%) and a specificity of 71% (CI 95%, 61 to 80%).

In patients with kidney transplants, the presence of acute tubular necrosis, together with the time the blood concentration of tacrolimus remained within the predetermined therapeutic interval, permitted the identification of patients with a higher probability of having an acute rejection episode during the first two weeks following the transplant.