LEOPARD, an acronym for Liver electronic offering platform with artificial intelligence-based devices, is a project funded by the EU’s Horizon Europe research and innovation program with the aim to address critical challenges in organ allocation for patients with decompensated cirrhosis and hepatocellular carcinoma.

The LEOPARD kick-off meeting took place on 25–26 January 2024 in Paris, France, gathering project partners, stakeholders, and experts from across Europe. The event marked the official start of the project, outlining its objectives, structure, and the collaborative efforts needed to achieve its goals.

Christophe Duvoux, Coordinator of the LEOPARD project from Assistance Publique–Hôpitaux de Paris (AP–HP), provided an overview of the project’s objectives and the pressing need for improved liver transplantation allocation strategies.  He highlighted key aspects of the project, including the development and validation of AI-based algorithms to better stratify patients by mortality and dropout risks, and the integration of predictive signatures from omics and radiomics to improve risk assessment accuracy.

The meeting featured detailed presentations by the leaders of each work package, outlining specific tasks and timelines:

• Designing a common European pre-liver transplantation data file across participating organ sharing organizations
  Wojtek Polak and Giacomo Germani, representatives of the European Society for Organ Transplantation (ESOT), presented the structure and organization of this work package, focusing on the selection of predictive variables and the logistics of data collection across participating countries.
• Design, implementation, and monitoring of the LEOPARD prospective cohort
  Christophe Duvoux (AP–HP) together with Candy Estevez and Biljana Zafirova from the European Clinical Research Infrastructure Network (ECRIN) discussed the design of the longitudinal cohort study, addressing regulatory and ethical aspects, and data sharing agreements.
• Multimodal data management and processing
  Venkata Satagopam from ELIXIR-LU, the Luxembourgish node of the European Infrastructure for Life Science Information, presented an outline of the data management plan, including database development and the integration of diverse data sources.
• Modeling and statistical analyses
  Etienne Audureau (AP–HP) detailed the methods and timeline for developing predictive models, emphasizing the need for accurate and robust statistical analyses.
• Testing and validation of LEOPARD models
  Etienne Audureau also presented the validation strategy for the predictive models, highlighting the importance of testing across different cohorts and integrating omics signatures. Among these, EF CLIF will test the predictive value of the three prognostic metabolites strongly associated with short-term mortality in the pre-liver transplantation setting of the LEOPARD observational cohort study.
• Communication, dissemination and valorization
  Devi Mey (ESOT) shared the communication and dissemination plan, including strategies for engaging stakeholders and promoting project findings.
• Management and coordination
  Maud Le Graet (FINOVATIS) and Essuih Koffi (AP–HP) presented the project management organizational structure, and discussed funding and coordination activities to ensure a smooth project execution.

The second day featured parallel working sessions focused on the practical organization of each work package, addressing key questions and planning next steps. 

The meeting concluded with summaries from each work package group, presented by their respective leaders, followed by closing remarks from Christophe Duvoux.