MPCD Index for Hepatocellular Carcinoma Patients Based on Mitochondrial Function and Cell Death Patterns

Area of Science:

• Oncology
• Mitochondrial Biology
• Computational Biology

Background:

• Hepatocellular carcinoma (HCC) is a heterogeneous cancer with poor prognosis.
• Mitochondria play a crucial role in cancer cell death by regulating metabolism, but their specific link to cell death patterns in HCC is not well understood.

Purpose of the Study:

• To develop a machine learning-based index (MPCDI) to quantify mitochondrial function and programmed cell death (PCD) in HCC.
• To investigate the association of MPCDI with multi-omics features, immune microenvironment, and drug sensitivity in HCC patients.

Main Methods:

• Utilized a machine learning framework on scRNA-seq and RNA-seq data from TCGA, GEO, and ICGC datasets.
• Constructed the MPCDI signature using identified genes and classified HCC patients into high- and low-MPCDI groups.
• Analyzed immune status, TIDE scores, immunotherapy response, and drug sensitivity data from multiple databases (CTRP, GDSC, PRISM).

Main Results:

• Developed the MPCDI signature comprising four genes (S100A9, FYN, LGALS3, HMOX1) as an independent prognostic factor for HCC.
• Identified significant differences in immune microenvironment between high- and low-MPCDI groups.
• Found that high-MPCDI patients exhibited higher TIDE scores and poorer immunotherapy response, suggesting limited suitability for immunotherapy.
• Indicated staurosporine as a potential therapeutic agent for high-MPCDI HCC patients based on drug sensitivity analysis.

Conclusions:

• The MPCDI signature offers novel insights into mitochondrial function and PCD patterns in HCC.
• MPCDI can stratify HCC patients based on immune characteristics and predict immunotherapy response.
• This index provides a foundation for personalized diagnosis and treatment strategies in HCC, including potential drug sensitivities.