Liver Electronic Offering Platform with Artificial intelligence-based Devices

Liver transplantation (LT) is a life-saving procedure for end-stage liver diseases, notably decompensated cirrhosis (DC) and hepato-cellular carcinoma (HCC). Its efficacy is yet hampered by the risk of death/drop-out on the Wait List (WL). This risk is driven by organ shortage and is mitigated by organ offering schemes. According to a sickest first policy, offering schemes prioritize LT candidates with the highest risk of dying, as assessed by predictive models of mortality. To drive allocation, Organ Sharing Organizations (OSOs) have adopted a 20-year old model, the MELD, which predict mortality in DC but not in HCC patients. Because of the dramatic increase in the % of HCC candidates (40% against 10% in earl 20ties), MELD-based offering schemes have become increasingly inaccurate, with a persisting 15-18 to 30% mortality in countries with low/medium donation rate. This scenario together with recent advances in the prognosis of DC and HCC LT candidates and statistics modelling, prompts the LT community to look for up-dated predictive models to refine offering schemes and improve patients’ outcome on the WL. In line with the objectives of the HLTH-2022-TOOL-12-01, and to provide a long-term response to MELD limitations, key European LT stakeholders including OSOs, experts in LT, DC and HCC, bio-statisticians, Research Labs and SMEs joined the LEOPARD task force. Building on an innovative, harmonized OSOs common pre LT data file as well as on recent advances in statistics, LEOPARD propose to develop, validate, and implement in real life-settings i) an evidence-based LEOPARD CIRR HCC machine learning-based predictive algorithm able to improve prediction of mortality on the WL, to be proposed to OSOs as an end-user tool to drive organ allocation and ii) Disease specific DC & HCC LEOPARD calculators available for LT professional as end-users tools to assist them in clinical assessment and decision-making processes. We expect from this project to generate computational tools able to improve stratification and outcomes of patients waiting for LT, with more patients transplanted on time. These tools will be tested in the real life setting in a prospective European Leopard cohort, which also serve to collect more granular data and bio- and tissues samples, for validation of 3 existing molecular predictive signatures, as preparatory steps to the development of 3rnd generation predictive models. Adoption of these tools should result in harmonization of the so far heterogeneous prioritization schemes and in a signification reduction in disparities of access to LT across European countries, a major objective pointed out by EC.