The issue

The LEOPARD project addresses a critical issue in liver transplantation: the high mortality and dropout rates among patients on the waitlist, particularly those with decompensated cirrhosis and hepatocellular carcinoma. Currently, mortality/dropout rates average 15–20% among listed patients, with significant disparities across Europe. Existing predictive models, such as the model for end-stage liver disease (MELD), are increasingly inadequate due to changes in the epidemiology of liver transplant candidates. These changes include a growing number of patients with hepatocellular carcinoma, older patients with multiple comorbidities, and those with advanced cirrhosis, often presenting higher MELD scores and acute-on-chronic liver failure (ACLF). As a result, the MELD-based system fails to accurately prioritize patients, leading to inefficiencies and inequities in organ allocation. Approximately 30% of patients with decompensated cirrhosis are listed under MELD exception rules, and the risk of dropout remains persistently high. These limitations highlight the urgent need for improved algorithms that can more accurately assess mortality and dropout risks, ensuring timely and equitable access to liver transplantation and ultimately reducing waitlist mortality.

The approach

The LEOPARD project employs a comprehensive and innovative approach to address the challenges in liver transplantation. Central to this effort is the development and validation of AI-based algorithms designed to better stratify patients by mortality and dropout risks, incorporating newly identified predictors that surpass the MELD score. The project will integrate predictive signatures from multi-omics and radiomics (imaging-based data) to improve risk assessment accuracy. A unified pre-liver transplantation data file will be created for organ sharing organizations (OSOs) across Europe, standardizing data collection and facilitating the harmonization of liver allocation strategies. LEOPARD will also launch a longitudinal cohort study to rigorously validate these AI-based models. The project will conduct extensive testing and validation using diverse cohorts, including available data from OSOs and the prospective cohort, to ensure the reliability of the algorithms in real-world settings. By combining cutting-edge AI technology, comprehensive data integration, and rigorous validation processes, the LEOPARD project aims to significantly improve liver transplant outcomes, reduce waitlist mortality, and harmonize organ allocation practices across Europe.