JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Consensus artificial intelligence-driven prognostic signature for predicting the prognosis of hepatocellular carcinoma: a multi-center and large-scale study

  • 0Cardiovascular Medicine, Affiliated Haikou Hospital of Xiangya Medical College, Central South University, Haikou, 570208, China.
Npj Precision Oncology +

|

July 2, 2025

|

July 2, 2025

View abstract on PubMed

Summary

This summary is machine-generated.

A new artificial intelligence model, the consensus artificial intelligence-derived prognostic signature (CAIPS), accurately predicts outcomes for hepatocellular carcinoma (HCC) patients. This tool aids in personalized treatment strategies and identifies potential drug therapies.

Area Of Science

  • Oncology
  • Bioinformatics
  • Artificial Intelligence

Background

  • Hepatocellular carcinoma (HCC) is a major global cancer with high mortality.
  • Molecular stratification is crucial for advancing precision oncology in HCC.
  • Existing prognostic signatures often lack broad applicability.

Purpose Of The Study

  • To develop a robust, AI-derived prognostic signature for hepatocellular carcinoma (HCC).
  • To validate the signature's accuracy and clinical utility.
  • To identify therapeutic targets and optimize personalized treatment strategies for HCC.

Main Methods

  • Integrated ten machine learning algorithms (101 methods) across six multi-center HCC cohorts (n=1110).
  • Developed a seven-gene consensus artificial intelligence-derived prognostic signature (CAIPS) using StepCox and GBM.
  • Performed multi-omics profiling, drug screening (CTPR, PRISM, Connectivity Map), and functional validation (PITX1 knockdown).

Main Results

  • The seven-gene CAIPS demonstrated superior prognostic accuracy compared to clinical parameters and existing signatures.
  • High CAIPS scores correlated with metabolic dysregulation and genomic instability.
  • Low CAIPS scores predicted enhanced response to transcatheter arterial chemoembolization, targeted therapies, and immunotherapy.
  • Irinotecan and BI-2536 were identified as potential therapeutics for high-CAIPS HCC.
  • PITX1 knockdown suppressed HCC progression via Wnt/β-catenin signaling inhibition.

Conclusions

  • CAIPS is a powerful multi-dimensional biomarker for HCC risk stratification and personalized management.
  • Irinotecan and BI-2536 show promise as targeted agents for high-risk HCC patients.
  • The study provides actionable frameworks for precision oncology in HCC treatment.