Patients who underwent kidney transplantation (KT) at Samsung Medical Center between January 2005 and December 2020 were the subjects of a recent investigation. The study aimed to determine the factors associated with kidney allograft rejection and develop a prediction model for acute kidney rejection based on these factors. The study excluded pediatric patients and those who received spontaneous solid organ transplantation or dual or en-bloc KT.
The researchers collected recipient and donor data from medical records, including sex, body mass index (BMI), underlying disease, pre-dialysis information, blood type, serum creatinine, donor type, previous transplantation history, and induction agent. They also investigated data on cold ischemic time, warm ischemic time, graft weight, and performed pathologic reports to determine the presence of borderline rejection. The immunologic risks of patients were classified into three groups: high, intermediate, and low, based on various conditions such as ABO incompatibility and cross-match positivity. Human Leukocyte Antigen (HLA) mismatch evaluations were also conducted.
Desensitization procedures were performed before transplantation depending on the immunologic risk. The maintenance protocol involved a triple immunosuppressive regimen for all patients. The researchers used various machine learning methods and logistic regression to train a prediction model for acute kidney rejection. The model’s performance was evaluated through hold-out validation and area under the curve (AUC) measurements. Fisher’s exact test, chi-square test, and logistic regression were used for statistical analysis. The study was approved by the Institutional Review Board of Samsung Medical Center, and informed consent was waived due to the retrospective nature of the study.
