Integrated CT-derived radiomics for predicting ctDNA status and prognosis in resectable non-small cell lung cancer

This study aimed to develop a radiomics signature for predicting ctDNA status and recurrence risk in non-small cell lung cancer (NSCLC) patients. A retrospective analysis of 631 patients was conducted, with 114 in the training set, 349 in internal validation, and 168 in external validation. Preoperative CT/PET-CT images were processed to extract 851 features using PyRadiomics, which were analyzed with LASSO and logistic regression to construct the radiomics signature. Clinical and genomic data, including tumor mutational burden (TMB) and KRAS/EGFR mutations, were integrated, and WGCNA identified molecular pathways linked to the radiomic features. The signature effectively stratified patients into high- and low-risk groups, with high-risk patients showing increased ctDNA positivity, advanced tumor stages, and elevated TMB. KRAS-mutant high-risk patients demonstrated poorer recurrence-free survival, suggesting potential benefit from combined KRAS-targeted and immunotherapies. Radiomics, when integrated with genomic data, offers a promising non-invasive tool for predicting ctDNA status and guiding personalized treatment strategies.