Identifier: TFG:9370
Authors: Sánchez Álvarez, Mario
Abstract:
Metabolic dysfunction-associated steatotic liver disease (MASLD) has a high global prevalence, affecting approximately 30% of the adult population, with a significant proportion progressing to metabolic steatohepatitis (MASH). The current diagnosis relies mainly on invasive techniques such as liver biopsy, which has important limitations, including patient risk and limited generalizability. In this context, the main objective of this study is to identify proteomic biomarkers that can non-invasively differentiate between the histopathological stages of simple steatosis (SS) and MASH. To achieve this goal, a systematic search was conducted, identifying three studies suitable for inclusion in a proteomic meta-analysis, comprising a total of 169 patients from different cohorts, thereby increasing the robustness and generalizability of the obtained results. The resulting differential expression analysis identified six significantly differentially expressed proteins (DEPs) between the SS and MASH groups. Subsequent functional analyses revealed notable alterations in metabolic pathways mainly related to the complement system, as well as to intracellular transport and metabolism. Finally, using machine learning techniques, a predictive model based on 15 proteomic associations was developed, showing promising performance for reliably and non-invasively differentiating between SS and MASH. However, further improvements in the model’s accuracy are needed before its definitive clinical implementation.
Repositori URV