Background: Early graft loss after liver transplantation (LT) remains a major challenge. Most predictive models rely on isolated factors or post-transplant variables, limiting their utility in pre-transplant decision-making. We developed a novel score integrating donor, recipient, and procedural variables to estimate 12-month graft loss based on the balance between risk and mitigation factors. Methods: In this retrospective study, 268 adult primary LT recipients from two European centers were analyzed. The primary endpoint was 12-month graft loss (death or retransplantation). Independent predictors were identified using multivariable logistic regression based on pre-transplant variables. A population-centered approach quantified individual deviations from average risk, classifying variables as risk exposure or mitigation factors. These were combined into a Risk-Mitigation Balance score. Model performance was evaluated using AUC and Brier score, and patients were stratified into three risk groups. Results: Twelve-month graft loss occurred in 17.9% of patients. Acute liver failure and warm ischemia time were the strongest predictors. The Risk-Mitigation Balance score showed good discrimination (AUC = 0.74) and calibration (Brier score 0.116), outperforming early post-transplant models (AUC = 0.65). Stratification identified three groups with significantly different graft loss rates: 7.4% (mitigation), 14.8% (intermediate), and 44.2% (risk) (p < 0.001), with clear separation on survival analysis. Conclusions: This novel score enables pre-transplant estimation of graft loss by integrating risk and modifiable factors, supporting more informed and personalized allocation decisions.
Lai et al. (Thu,) studied this question.