To extend radiomics-based patient stratification to individuals with stage IIIB–IV metastatic lung adenocarcinoma (MLUAD) using unsupervised clustering of CT-derived radiomic features, which may enhance predictive accuracy for treatment outcomes.
Key Findings:
Radiomic features were associated with specific oncogenic alterations in LUAD, suggesting potential for targeted therapies.
Unsupervised clustering revealed distinct imaging patterns linked to treatment response and overall survival, indicating the utility of radiomics in clinical decision-making.
Smoking history influenced the molecular heterogeneity and radiomic profiles observed, highlighting the need for personalized approaches.
Interpretation:
The study suggests that CT-derived radiomic features can serve as non-invasive biomarkers for predicting oncogenic alterations and clinical outcomes in metastatic lung adenocarcinoma, potentially guiding treatment strategies.
Limitations:
Single-center study may limit generalizability and diversity of the patient population.
Retrospective design may introduce selection bias.
Potential variability in imaging acquisition and analysis across different centers.
Conclusion:
CT-based radiogenomics can enhance patient stratification and treatment personalization in metastatic lung adenocarcinoma, warranting further validation in larger, multi-center studies.
by Amandine Crombé, Lou Andrea Sitruk, Cécile Masson-Grehaigne, Mathilde Lafon, Jean Palussiere, Benjamin Bonhomme, Sophie Cousin, Nathalie Lassau, Antoine Italiano
