Related Experiment Video
Updated: Mar 14, 2026
04:04
Endobronchial Ultrasound-guided Intratumoral Injection of Cisplatin for the Treatment of Isolated Mediastinal Recurrence of Lung Cancer
Published on: February 12, 2017
11.1K
Defining a Radiomic Response Phenotype: A Pilot Study using targeted therapy in NSCLC.
Hugo J W L Aerts1,2,3, Patrick Grossmann1,3, Yongqiang Tan4
1Departments of Radiation Oncology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA.
Scientific Reports
|September 21, 2016
Summary
Radiomics analysis of CT scans can predict EGFR mutation status and gefitinib response in lung cancer patients before treatment. Changes in radiomic features after treatment further enhance prediction accuracy and stability.
Area of Science:
- Oncology
- Medical Imaging
- Radiomics
- Drug Development
Background:
- Medical imaging is crucial in oncology for non-invasive tumor visualization.
- Radiomics quantifies tumor phenotype using image-analysis algorithms, offering insights beyond size.
- Predicting treatment response is key for personalized oncology and drug development.
Purpose of the Study:
- To investigate if radiomics can identify a gefitinib response-phenotype in early-stage non-small cell lung cancer (NSCLC).
- To assess the predictive value of radiomic features for EGFR mutation status and gefitinib response.
- To evaluate the stability and test-retest reliability of radiomic features.
Main Methods:
- Analysis of high-resolution computed-tomography (CT) scans from 47 NSCLC patients.
- Radiomic feature extraction from baseline and post-treatment (3 weeks) scans.
- Statistical analysis to determine feature predictability for EGFR mutation status and gefitinib response.
- Technical validation for feature stability using test-retest analysis.
Main Results:
- Baseline radiomic feature Laws-Energy significantly predicted EGFR mutation status (AUC=0.67, p=0.03).
- No single feature predicted response on post-treatment scans, but changes in features were highly predictive (AUC range=0.74-0.91).
- Radiomic features demonstrated high test-retest stability (ICC=0.96 ± 0.06).
Conclusions:
- Pre-treatment radiomic data can non-invasively predict EGFR mutation status and gefitinib response in NSCLC.
- Radiomics-based phenotyping shows potential for improving patient stratification based on TKI sensitivity.
- This pilot study highlights radiomics as a valuable tool for response assessment in targeted cancer therapy.
