Related Experiment Video
Updated: Aug 25, 2025
Author Spotlight: Advancements in Molecular Biomarker Testing for Non-Squamous Non-Small Cell Lung Cancer
Published on: September 8, 2023
Radiogenomic System for Non-Invasive Identification of Multiple Actionable Mutations and PD-L1 Expression in
Jun Shao1, Jiechao Ma2, Shu Zhang2
1Department of Respiratory and Critical Care Medicine, Med-X Center for Manufacturing, West China Hospital, West China School of Medicine, Sichuan University, No. 37 GuoXue Alley, Chengdu 610041, China.
Purpose:
Personalized treatments such as targeted therapy and immunotherapy have revolutionized the predominantly therapeutic paradigm for non-small cell lung cancer (NSCLC). However, these treatment decisions require the determination of targetable genomic and molecular alterations through invasive genetic or immunohistochemistry (IHC) tests. Numerous previous studies have demonstrated that artificial intelligence can accurately predict the single-gene status of tumors based on radiologic imaging, but few studies have achieved the simultaneous evaluation of multiple genes to reflect more realistic clinical scenarios.
Methods:
We proposed a multi-label multi-task deep learning (MMDL) system for non-invasively predicting actionable NSCLC mutations and PD-L1 expression utilizing routinely acquired computed tomography (CT) images. This radiogenomic system integrated transformer-based deep learning features and radiomic features of CT volumes from 1096 NSCLC patients based on next-generation sequencing (NGS) and IHC tests.
Results:
For each task cohort, we randomly split the corresponding dataset into training (80%), validation (10%), and testing (10%) subsets. The area under the receiver operating characteristic curves (AUCs) of the MMDL system achieved 0.862 (95% confidence interval (CI), 0.758-0.969) for discrimination of a panel of 8 mutated genes, including EGFR, ALK, ERBB2, BRAF, MET, ROS1, RET and KRAS, 0.856 (95% CI, 0.663-0.948) for identification of a 10-molecular status panel (previous 8 genes plus TP53 and PD-L1); and 0.868 (95% CI, 0.641-0.972) for classifying EGFR / PD-L1 subtype, respectively.
Conclusions:
To the best of our knowledge, this study is the first deep learning system to simultaneously analyze 10 molecular expressions, which might be utilized as an assistive tool in conjunction with or in lieu of ancillary testing to support precision treatment options.
Insights
This study introduces a deep learning system to predict multiple non-small cell lung cancer (NSCLC) mutations and PD-L1 expression from CT scans. This AI tool aids in non-invasive, simultaneous molecular analysis for personalized NSCLC treatments.
Area of Science:
- Radiology and Oncology
- Artificial Intelligence in Medicine
- Genomics and Molecular Diagnostics
Background:
- Personalized treatments like targeted therapy and immunotherapy have transformed non-small cell lung cancer (NSCLC) care.
- Current treatment decisions rely on invasive genetic or immunohistochemistry (IHC) tests to identify targetable alterations.
- While AI can predict single-gene status from imaging, simultaneous multi-gene evaluation is less explored.
Purpose of the Study:
- To develop and validate a multi-label multi-task deep learning (MMDL) system for non-invasively predicting actionable NSCLC mutations and PD-L1 expression.
- To assess the system's ability to simultaneously analyze multiple molecular markers using routine computed tomography (CT) images.
- To provide a potential alternative or supplement to invasive testing for guiding precision medicine in NSCLC.
Main Methods:
- A radiogenomic MMDL system was proposed, integrating transformer-based deep learning and radiomic features from CT volumes.
- The system was trained and validated on data from 1096 NSCLC patients with known next-generation sequencing (NGS) and IHC results.
- Datasets were randomly split into training (80%), validation (10%), and testing (10%) subsets for each prediction task.
Main Results:
- The MMDL system achieved an AUC of 0.862 for predicting 8 common mutated genes (e.g., EGFR, ALK, KRAS).
- It achieved an AUC of 0.856 for identifying a 10-molecular status panel, including TP53 and PD-L1.
- The system demonstrated strong performance in classifying EGFR/PD-L1 subtypes with an AUC of 0.868.
Conclusions:
- This study presents the first deep learning system capable of simultaneously analyzing 10 molecular expressions in NSCLC.
- The MMDL system shows promise as an assistive tool for precision treatment selection in NSCLC.
- This non-invasive approach may reduce the need for or complement traditional ancillary testing methods.
