Related Experiment Video
Updated: Jun 29, 2025
Utilizing 18F-FDG PET/CT Imaging and Quantitative Histology to Measure Dynamic Changes in the Glucose Metabolism in Mouse Models of Lung Cancer
Published on: July 21, 2018
Distinguishing EGFR mutation molecular subtypes based on MRI radiomics features of lung adenocarcinoma brain
Jiali Xu1, Yuqiong Yang2, Zhizhen Gao3
1Department of Radiology, The First Affiliated Hospital of Bengbu Medical University, Bengbu, Anhui 233004, China; Department of Medical Imaging Diagnosis, School of Medical Imaging, Bengbu Medical University, Bengbu, Anhui, China.
This study shows that radiomics features from MRI of lung adenocarcinoma brain metastases can identify epidermal growth factor receptor (EGFR) mutation subtypes. This non-invasive method aids in predicting EGFR 19Del and 21L858R mutations in the primary tumor.
Area of Science:
- Radiology and Imaging
- Oncology
- Molecular Diagnostics
Background:
- Lung adenocarcinoma (LUAD) is a major cause of cancer-related deaths.
- Epidermal growth factor receptor (EGFR) mutations are key drivers in LUAD, influencing treatment decisions.
- Distinguishing between common EGFR mutations (19Del and 21L858R) is crucial for targeted therapy selection.
Purpose of the Study:
- To assess the feasibility of using radiomics from brain metastases MRI to predict EGFR mutation subtypes in the primary LUAD lesion.
- To develop and validate a radiomics-based model for differentiating between EGFR exon 19 deletion (19Del) and exon 21 L858R point mutation (21L858R).
Main Methods:
- Retrospective analysis of 86 LUAD patients with brain metastases and known EGFR mutation status.
- Extraction of 3D radiomics features from contrast-enhanced T1-weighted MRI scans of brain metastases.
- Development of a radiomics model using principal component analysis, Relief feature selection, and adaptive boosting classification, validated with cross-validation and an independent test set.
Main Results:
- The radiomics model achieved an Area Under the Curve (AUC) of 0.895 in the training set and 0.759 in the testing set for distinguishing between 19Del and 21L858R mutations.
- Patient age was identified as an independent predictor.
- A combined model of age and radiomics showed a slightly higher AUC (0.888) than radiomics alone (0.866), with both models demonstrating clinical utility.
Conclusions:
- Radiomics analysis of lung adenocarcinoma brain metastases on MRI can successfully predict EGFR 19Del and 21L858R mutations in the primary tumor.
- This non-invasive imaging approach offers a potential method for molecular subtyping, aiding in personalized treatment strategies for LUAD patients.
