Radiomics model based on intratumoral and peritumoral features for predicting major pathological response in non-small cell lung cancer receiving neoadjuvant immunochemotherapy

• DOI: 10.3389/fonc.2024.1348678
• Publication Date: 2024-03-20T05:18:20Z
• Authors: Dingpin Huang; Chen Lin; Yangyang Jiang; Enhui Xin; Fangyi Xu; Yi Gan; Rui Xu; Fang Wang; Haiping Zhang; Kaihua Lou; Lei Shi; Hongjie Hu
• Keywords: Neoadjuvant Therapy

Objective To establish a radiomics model based on intratumoral and peritumoral features extracted from pre-treatment CT to predict the major pathological response (MPR) in patients with non-small cell lung cancer (NSCLC) receiving neoadjuvant immunochemotherapy. Methods A total of 148 NSCLC who underwent immunochemotherapy two centers (SRRSH ZCH) were retrospectively included. The SRRSH dataset (n=105) was used as training internal validation cohort. Radiomics (T) regions (P1 = 0-5mm, P2 5-10mm, P3 10-15mm) CT. Intra- inter- class correlation coefficients least absolute shrinkage selection operator feature selection. Four single ROI models mentioned above combined (CR: T+P1+P2+P3) established by using machine learning algorithms. Clinical factors selected construct radiomics-clinical (CRC) model, which validated external center ZCH (n=43). performance assessed DeLong test, calibration curve decision analysis. Results Histopathological type only independent clinical risk factor. CR eight demonstrated good predictive (AUC=0.810) significantly improved than T (AUC=0.810 vs 0.619, p<0.05). CRC yielded best capability (AUC=0.814) obtained satisfactory test set (AUC=0.768, 95% CI: 0.62-0.91). Conclusion We that incorporates histopathological type, providing an effective approach for selecting suitable