Biomarker Prediction of Delayed Graft Function and Prognosis Post-Kidney Transplantation.

Abstract

Introduction: Aminoacylase-1 (ACY1) and additional biomarkers were evaluated for prediction of delayed graft function (DGF) and prognosis following kidney transplantation. Method(s): Serum biomarkers were measured (days 1-2 posttransplant for DGF prediction and 1-3 for prognosis) in 237 patients transplanted in Leeds (2003-2011) in the discovery phase, and 319 patients from 7 UK transplant centers (2012-2016) in the validation phase. Median follow-up was 13.28 years (interquartile range [IQR]: 12.4-13.8) and 9.03 years (IQR: 5.19-10.16), respectively. Result(s): DGF occurred in 29.5% of discovery cohort patients and 18.2% of validation cohort patients. A DGF linear predictor combining ACY1, soluble tumor necrosis factor receptor-1 (sTNFR1) and cystatin C (CysC) demonstrated an area under the receiver operating characteristic (AUROC) of 0.93, decreasing to 0.83 during validation. Comparable values for the individual components were ACY1 (0.79 vs. 0.65), sTNFR1 (0.88 vs. 0.89), CysC (0.89 vs. 0.82), and 0.75 vs. 0.81 for serum creatinine (Cr) as the gold standard. The linear predictor variables for death-censored graft survival (DCGS) were CysC, ACY1, midkine, and recipient age at transplant, but k-statistic values of 0.55 in all transplants and 0.52 in deceased donor kidney transplants (DDKT) precluded validation. Individually, sTNFR1, CysC, and Cr were significantly associated with DCGS in both the discovery and validation phases. Preliminary findings indicated a consistent association of ACY1 and CysC with DCGS in DDKTs affected by DGF. Impacts of clinical practice changes and biomarker performance across the 2 phases are described. Conclusion(s): Several biomarkers show potential as predictors of DGF or outcome and should be explored further.