Objective: To develop and evaluate a nomogram integrating radiomic features from contrast-enhanced CT with clinical variables for pre-treatment predictions of the response to neoadjuvant therapy (NAT) in locally advanced gastric cancer (LAGC). Methods: In this retrospective multicenter study, 183 LAGC patients from the First Affiliated Hospital of Nanjing Medical University (2014–2023) were included. Radiomic features were extracted from manually delineated pre-treatment CT regions of interest. A machine learning-based predictive model combining radiomic scores and clinical data was constructed. Model performance was assessed using receiver operating characteristic (ROC) analysis, calibration curves, and decision curve analysis (DCA). Results: Multivariate analysis identified the radiomic score, preoperative N stage, and neoadjuvant regimen as independent predictors of NAT responses (all p < 0.05). The integrated nomogram achieved an area under the ROC curve of 0.807 and showed a moderate net benefit in DCA compared with the radiomics-only model. Conclusions: The radiomics–clinical nomogram demonstrates moderate predictive performance for pre-treatment stratification of NAT responses in LAGC. These findings are exploratory and hypothesis-generating, and further validation in independent cohorts is required before clinical application.
Zhou et al. (Mon,) studied this question.