Optimizing Noninvasive assessMent Of DysmEtabolic Compensated Advanced Liver Disease

Optimizing Noninvasive assessMent Of DysmEtabolic Compensated Advanced Liver Disease by Integration of Artificial Intelligence Model and omicS Data

Non-alcoholic fatty liver disease (NAFLD) is responsible for a significant proportion of liver-related deaths and healthcare costs in the United States, accounting for approximately 36% of liver-related deaths and over one billion dollars in annual healthcare expenses. [PMID: 34863359] A recent analysis of healthcare costs in Italy showed that out of the 9,729 NAFLD/NASH patients who were hospitalized and analyzed, the vast majority (97%) did not have advanced liver disease, while 1.3% had compensated advanced liver disease (cACLD), 3.1% had decompensated cirrhosis, 0.8% had hepatocellular carcinoma, and 0.1% underwent liver transplantation.

The burden of comorbidities was high across all patient cohorts, and patients with cACLD required a greater number of inpatient services, outpatient visits, and the pharmacy fills compared to those without advanced liver disease. As disease severity increased, mean total annual costs also increased primarily due to higher inpatient services costs. In Italy, as in other EU countries, most of the healthcare costs for patients were attributed to NAFLD/NASH-related liver complications. Thus, the optimization of the non-invasive diagnosis of cACLD represents an urgent need in dysmetabolic liver disease. These advancements will play a crucial role in early detection, risk stratification, and effective management of highly prevalent liver diseases such as NAFLD/NASH and their progression.

The study aims to significantly enhance diagnostic innovation and contribute to the existing literature on the stratification of cACLD caused by metabolic-dysfunction liver disease, a major factor leading to cirrhosis, liver cancer, and liver transplant in individuals with non-communicable diseases. By integrating radiomics, digital pathology, non-invasive scores, and omics the results are expected to provide novel evidence for diagnostic advancements.

The incorporation of AI is anticipated to lead to more efficient diagnostic management, effectively addressing the impact of cACLD on healthcare systems. The outcomes of this research will yield a substantial database and intellectual content, both of which will be made available to the scientific community and multiple stakeholders, including patient associations, policymakers, healthcare providers, and industry players.

The primary goal is to foster innovation in diagnostics and mitigate the impact of cACLD on national health systems. By accurately predicting individuals at higher risk of liver or extra-hepatic complications, this study aims to revolutionize diagnostic methods, ultimately leading to improved patient outcomes and resource optimization in healthcare settings.

Inclusion Criteria:

• age>=18; sex (M,F);
• dysmetabolic liver disease according new nomenclature definition;
• suspicion of cACLD by LSM>=10 with VCTE;
• routine esogastroduodenoscopy report within 12 months of VCTE for identification of high-risk varices (HRV).

Exclusion Criteria:

• portal vein thrombosis,
• infiltrative liver neoplasms, and conditions are known for their potential influence on the LSM results (congestive liver disease, extrahepatic biliary obstruction, ALT > 5x upper normal limit).