Background: Preimplantation scores assist with correct kidney graft allocation, but macroscopic graft features have never been evaluated in this scenario.
Methods: We designed a graft appraisal questionnaire, assessed its reproducibility by comparing the senior and junior surgeon responses and evaluated which features can predict transplant outcomes in 202 patients transplanted from 144 donors at a tertiary center. We created new prediction models in combination with validated preimplantation scores. The primary outcome was graft loss or eGFR<30 mL/min/1.73 m2 at six months and secondary outcomes were delayed graft function, early graft loss and graft function at six months.
Results: Interrater correlation was very good for adherent perinephric fat (kappa=0.91) and acceptable for cortical surface roughness (kappa=0.51) and cortical color (kappa=0.47). Adherent perirenal fat (Odds ratio=4.77; 95% CI: 2.10-10.85) and surface roughness (OR=2.11, 95% CI: 1.25-3.58) were independent predictors of the primary outcome, improving the kidney donor risk index efficacy model (AUC 0.71 vs. 0.82, P≤0.001), while cortical color and adherent fat improved the Irish risk model for delayed graft function (AUC 0.76 vs. 0.82, P=0.03). We created nomograms to visually assess the risk of both endpoints.
Conclusions: Kidney graft macroscopic appraisal is reproducible between surgeons and can improve the accuracy of clinical preimplantational prediction scores.