Data Collection to Design and Validate LEOPARD Predictive Models of Delisting in Liver Transplant Candidates

What this Trial Is About

Researchers are working on the LEOPARD project funded by the European Commission to improve how liver transplant candidates are prioritized. The study focuses on creating and validating new AI-based predictive models to better estimate mortality and dropout risks for patients with decompensated cirrhosis, end-stage chronic liver diseases, or Hepato-cellular carcinoma (HCC) who are on liver transplant waiting lists across Europe. Current MELD-based systems have become less accurate, and mortality rates vary widely by country and condition. This observational study will gather health data prospectively from 4,500 liver transplant candidates across 50 centers in seven countries. The data is collected in two phases: a training set of 3,000 patients and a validation set of 1,500 patients. Participants include those with decompensated cirrhosis, other end-stage liver diseases listed under MELD allocation, and HCC patients with or without cirrhosis. The study does not involve experimental treatments but collects detailed clinical information to develop predictive models. Participants will be monitored for outcomes including death or removal from the transplant list due to health decline, up to 12 months after listing. Researchers will track causes of death or dropout, time to transplantation, and post-transplant survival at 6 and 12 months. Data collection includes longitudinal health records and imaging results, with participants followed according to standard care. The goal is to improve fairness and timing in liver transplant allocation based on more accurate risk predictions.

LEOPARD Training and Validation Data Collection Study