Contrast-enhanced CT texture analysis for the prediction of delayed graft function following kidney transplantation from cadaveric donors.

Abstract

Delayed graft function (DGF) is a common complication after transplantation of deceased donor kidneys. The aim of this study was to investigate the feasibility of using computed tomography texture analysis (CT-TA) of the donor kidney to predict delayed graft function (DGF) following kidney transplantation from cadaveric donors. We made a retrospective review of all consecutive DBD and DCD kidney donors admitted to our institution and their corresponding KTRs between December 2014 and January 2019. We extracted 15 image features from unenhanced CT and contrast-enhanced CT corresponding to first order and second order Haralick textural features. Predictors of DGF were evaluated by univariable and multivariable analysis. Receiver operating characteristic (ROC) analysis was performed and the area under the ROC curve (AUC) to predict DGF was calculated for the predictors. A total of 115 patients were included in the study. DGF occurred in 15 patients (13%). Recipient body mass index (BMI) (P=0.003) and Skewness (P=0.05) represented independent predictors in the multivariate model. The combination of both clinical and textural features in a bivariate model reached a ROC-AUC of 0.79 (95% CI: 0.64-0.94) in predicting the probability of DGF. Results from this preliminary study suggest that CT texture analysis might be a promising quantitative imaging tool to help physician predict DFG after kidney transplantation from cadaveric donors. 4/5.