Clinical Classification of MAFLD Based Liver Biopsy

Metabolic Dysfunction-associated Steatohepatitis (MASH), Metabolic Dysfunction Associated Steatotic Liver Disease

MAFLD, MASLD, liver-biopsy, classification, prediction model

Metabolic dysfunction-associated Fatty Liver Disease (MAFLD), also known as Non-Alcoholic Fatty Liver Disease (NAFLD), is the most common chronic progressive liver disease in China. It is closely related to the high incidence of cardiovascular-renal-metabolic syndrome and both liver and non-liver malignancies, posing a serious threat to public health. However, the diagnostic criteria for MAFLD are not unified globally, and the classification and staging still rely on liver biopsy for pathological assessment. The characteristics, mechanisms, and predictive indicators of liver and extrahepatic disease outcomes in MAFLD patients are not yet clear. The severe form of MAFLD, metabolic dysfunction-associated steatohepatitis (MASH), has been a hot and challenging area of research for non-invasive tests (NITs). However, serum markers, imaging examinations, and novel markers under development cannot replace liver biopsy for the diagnosis of MASH. Clinically, the disease outcomes of MAFLD mainly depend on metabolic cardiovascular risk factors and fibrosis staging. Both liver biopsy and NIT-diagnosed advanced fibrosis and cirrhosis can predict liver-related events and all-cause mortality risks in MAFLD patients. Artificial intelligence and machine learning methods can improve the consistency of pathologists in diagnosing MASH and fibrosis. The Agile score, which combines gender, T2DM status, AST/ALT ratio, platelet count, and liver stiffness measurement (LSM), can improve the diagnostic efficacy of advanced fibrosis and cirrhosis in MAFLD patients and the efficiency of predicting liver-related events. However, the predictive effect of fibrosis staging and its changes on liver cancer needs to be improved. There is a lack of high-quality research on early warning indicators for the incidence of CVD, chronic kidney disease, and non-liver malignancies in MAFLD patients. It is necessary to explore the role of conventional indicators such as low-density lipoprotein cholesterol, lipoprotein(a), uric acid, and high-sensitivity C-reactive protein, as well as multi-omics parameters, in the classification, staging, and risk prediction of MAFLD. MAFLD is an increasingly serious public health issue associated with a higher risk of liver-related events, cardiovascular-renal-metabolic syndrome, and malignancies. The prevalence of MAFLD in China is high, but the rate of standardized management is low. Even patients with the same classification and staging often have different clinical characteristics and outcomes. There is currently a lack of a clinical classification and early warning system for MAFLD that combines metabolic cardiovascular risk factors and NITs for different outcome risks.

1. Recruitment and Data Collection: On the basis of an existing cohort of 1,500 liver biopsy cases, recruit an additional 500 cases from a national multicenter liver biopsy follow-up cohort (totaling 2,000 cases). Collect demographic, anthropometric, laboratory, imaging, and liver biopsy results for these patients. Concurrently, biological samples, including blood, urine, feces, and liver biopsy tissues, will be collected. Utilize these samples to perform quantitative metabolite information based on database matching. Employ techniques such as genomics, epigenomics, proteomics, metabolomics, immunomics, and microbiome metagenomics to screen for differential biomarkers across different subgroups. Combine these findings with clinical and imaging parameters of MAFLD patients to analyze and explore key parameters and molecules at different stages and outcomes of MAFLD disease progression. 2. Development and Validation of a Diagnostic and Prognostic System: Based on key molecules identified through multi-omics, in conjunction with characteristic parameters from clinical and imaging data of MAFLD patients, use machine learning methods (such as random forests neural networks) combined with logistic regression to establish a novel non-invasive diagnostic and prognostic assessment system for adverse outcomes (cardiovascular events, non-liver malignancies, and liver-related events). Validate this new assessment system to ensure its reliability and accuracy in predicting disease outcomes.

No related papers found yet.

None listed

China