Related Experiment Video
Updated: Jul 2, 2026
07:47
Multiomics Analysis of TMEM200A as a Pan-Cancer Biomarker
Published on: September 15, 2023
Multi-omics Signature Predicts Anti-EGFR Therapy Benefit in Colorectal Cancer Liver Metastases: A Multi-center Cohort
Yu Liu1,2,3, Yuan Chen4, Guofeng Zhou5
1Department of Colorectal Surgery, Zhongshan Hospital, Fudan University, Shanghai, 200032, China.
Current Cancer Drug Targets
|July 1, 2026
Summary
A new multi-omics deep learning model accurately predicts which patients with RAS wild-type colorectal cancer liver metastases (CRLM) will respond to anti-EGFR therapy, improving treatment selection. This approach integrates radiomic and genetic data for enhanced predictive capability.
Area of Science:
- Oncology
- Bioinformatics
- Medical Imaging
Background:
- Colorectal cancer liver metastases (CRLM) management is challenging due to primary resistance to anti-EGFR therapy in RAS wild-type patients.
- Predicting therapeutic response in this cohort requires advanced analytical tools.
Purpose of the Study:
- To develop and validate a multi-omics deep learning framework for predicting anti-EGFR therapy response in RAS wild-type CRLM.
- To integrate radiomic and genetic data for improved therapeutic stratification.
Main Methods:
- A multi-omics deep learning framework was developed using PyTorch.
- Individual radiomic and genetic signatures were created from CT images and genetic data.
- A combined model integrated these signatures for final classification probability.
Main Results:
- The combined multi-omics signature achieved an AUC of 0.86 for predicting cetuximab sensitivity.
- The signature outperformed existing biomarkers in identifying treatment-sensitive cases.
- A significant correlation with progression-free survival (PFS) was observed.
Conclusions:
- The multi-omics signature shows promise in identifying RAS wild-type CRLM patients likely to benefit from anti-EGFR therapy.
- This tool may aid in early treatment stratification and personalized decision-making.
- Further external validation is warranted.
