LEOPARD: Liver Electronic Offering Platform with ARtificial Intelligence-based Devices

A European commission-funded project

The LEOPARD project stands as a pioneering effort in the field of liver transplantation (LT), uniting stakeholders across Europe to revolutionise organ allocation strategies for individuals with decompensated cirrhosis (DC) and hepatocellular carcinoma (HCC).

Among patients currently listed for liver transplantation in Europe, mortality/drop-out on the waitlist averages 15-20% with large disparities across European countries. In recognition of the critical need to prioritise LT candidates based on mortality risk, particularly in the context of organ shortages, and the limitations of existing predictive models such as the Model for End-Stage Liver Disease (MELD) in accurately assessing this risk, there is a growing urgency for updated algorithms to refine organ offering schemes. 

The LEOPARD project seeks to enhance liver transplantation outcomes by creating and validating an AI-based predictive algorithm, considering recently identified predictors, that surpasses current models in stratifying both DC and HCC patients by mortality/dropout risks on the waitlist. Additionally, the project will develop calculators for DC and HCC candidates to aid in patient prioritisation, as well as integrate predictive signatures from OMICs/radiomics to improve risk assessment accuracy. 

At its core, the LEOPARD project aims to improve patient outcomes by ensuring timely transplantations, harmonising European prioritisation schemes, and advocating for equitable access to LT to significantly reduce mortality on the waitlist. 

A message from the LEOPARD project co-ordinator,
Christophe Duvoux:

The global adoption of MELD-based offering schemes in the early 2000s was a major step forward in improving liver graft allocation and reducing the mortality of liver transplant candidates on the waitlist. 

Yet, 20 years later, considerable changes in the epidemiology of liver transplant candidates have been associated with increasing limitations of MELD-based systems. 

The changes encompass a notable rise in the number of patients listed for hepatocellular carcinoma (HCC), reaching up to 40% in many programs. This increase involves those who may not be adequately assessed by the MELD scoring system, as well as the listing of older patients with multiple comorbidities and sicker individuals with advanced cirrhosis, often presenting with higher MELD scores and acute-on-chronic liver failure (ACLF). Consequently, these changes have resulted in increasing limitations of MELD–based allocation, translating in daily practice to a 30% rate of DC patients listed under MELD exceptions rules and a persistent 15 to 30% risk of dropout due to mortality whilst on the waitlist. This demonstrates the need for an urgent investigation to revise MELD-based allocation methods.

The LEOPARD project will build on 3 major opportunities to tackle this issue: 

  1. Recent advances in knowledge, including the identification of new predictors of mortality in DC patients, independent of MELD, and new predictors of dropout in HCC patients

  2. Increasing evidence that the accuracy of AI machine learning-based predictive models can outperform the precision of conventional statistical predictive models, opening the door to machine learning algorithms to better stratify liver transplant candidates

  3. Design new computational models for new stratification strategies, a strategic opportunity offered by the European Commission to unlock the potential of AI (Horizon Europe 2022 tender, Call 12 01)

The funding of the LEOPARD project by the European Commission offers the first consortium of key European LT stakeholders. The consortium includes organ-sharing organisations (OSOs), including ABM, CNT, ONT, EuroTransplant, as well as 50 liver transplantation centres, major key opinion leaders in the field of hepatology, liver oncology, liver surgery and liver transplantation, statisticians, research labs, small and medium-sized enterprises (SMEs) and patient organisations.

The project will explore and prospectively validate advanced, second and next-generation predictive algorithms in liver transplant candidates which will be offered to participating OSOs to improve outcomes, reduce mortality on the waitlist, and harmonise offering schemes across Europe. 

“As LEOPARD Coordinator, I am proud and honoured to serve this unique initiative.” 

Christophe Duvoux