The development of a predictive model of graft function in uncontrolled donors after circulatory death: validity of a pulsatile renal preservation machine cut-off value for kidney acceptance.

Background The criteria for kidney suitability in uncontrolled donors after circulatory death (uDCD) procured after regional normothermic perfusion are based on macroscopic appearance and renal haemodynamic values with final renal resistance (FRR). However, these criteria have not been analysed to predict the future graft function. This study presents a model to predict the outcome in uDCD kidneys and define the predictive FRR value. Methods All uDCD kidney transplants performed in our hospital from 2004 to 2016 were included. Donors and recipients and pre-transplantation data are described. The endpoint was glomerular filtration rate (GFR) 30mL/min at 6months after transplantation. Results A total of 194 recipients were included. FRR in donors 60years old was (meanSD) 0.27 +/- 0.11 versus 0.22 +/- 0.09mmHg/mL/min in donors <60years (P=0.042). Kidney survival was 88.2% versus 84% at 12months and 60.7% versus 30.8% at 120months (P=0.067). For the group of recipients from donors 60years, the FRR was 0.37 +/- 0.08mmHg/mL/min in the GFR<30mL/min group versus 0.18 +/- 0.06mmHg/mL/min in the GFR 30mL/min group (P<0.001). The value FRR 0.3mmHg/mL/min predicts 59-79% of GFR <30mL/min [odds ratio = 2.16, 95% confidence interval (CI) 1.80-6.40; P<0.001]. The predictive accuracy of FRR for GFR by ROC curve was 0.968 (95% CI). The best cut-off for FRR was 0.3mmHg/mL/min to predict GFR at 6months with a sensitivity of 67%, specificity of 100%, positive predictive value of 83% and negative predictive value of 92%. Conclusions Our results suggest that in uDCD donors the combination of donor age 60years together with FRR 0.3mmHg/mL/min could predict poor outcome at 6months after transplantation in low immunological risk recipients.