Multi-omic profiling reveals potential biomarkers of hepatocellular carcinoma prognosis and therapy response among...

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Liver hepatocellular carcinoma (LIHC) presents significant challenges in growth, metastasis, and drug resistance.
• Current predictive models for LIHC are insufficient for effective personalized medicine strategies.

Purpose of the Study:

• To develop a novel predictive model for LIHC prognosis and treatment response.
• To leverage mitochondrial cell death (MCD) patterns and machine learning for a robust prognostic index.

Main Methods:

• Multi-omic screening of MCD-related genes and development of a consensus MCD index (MCDI) using a machine learning framework.
• Validation of MCDI across training, validation, and clinical cohorts using multi-omics and experimental techniques.
• Evaluation of risk subgroup responses to immunotherapy and targeted therapy.

Main Results:

• Identification of nine critical differentially expressed MCD-related genes in LIHC.
• The developed MCDI demonstrated high performance in predicting prognosis and clinical outcomes.
• MCDI correlated with immune infiltration, TIDE scores, and sorafenib sensitivity, with PAK1IP1 identified as a key prognostic gene.

Conclusions:

• A novel predictive model, MCDI, has been developed for LIHC.
• Integrating MCDI into predictive, preventive, and personalized medicine (PPPM) frameworks can improve clinical decision-making and treatment individualization for LIHC patients.