CT-based Radiomics Nomogram for Differentiating Gastric Hepatoid Adenocarcinoma from Gastric Adenocarcinoma: a Multicentre Study.

Jing Wang,Bing Kang,Cong Sun,Fengying Du,Jianxian Lin,Fanghui Ding,Zhengjun Dai,Yifei Zhang,Chenggang Yang,Liang Shang,Leping Li,Qingqi Hong,Changming Huang,Guangbin Wang

DOI: https://doi.org/10.1080/17474124.2023.2166490

2023-01-01

Expert Review of Gastroenterology & Hepatology

Abstract:BackgroundTo develop a CT-based radiomics nomogram for the high-precision preoperative differentiation of gastric hepatoid adenocarcinoma (GHAC) patients from gastric adenocarcinoma (GAC) patients.Research design and methods108 patients with GHAC from 6 centers and 108 GAC patients matched by age, sex and T stage undergoing pathological examination were retrospectively reviewed. Patients from 5 centers were divided into two cohorts (training and internal validation) at a 7:3 ratio, the remaining patients were external test cohort. Venous-phase CT images were retrieved for tumor segmentation and feature extraction. A radiomics model was developed by the least absolute shrinkage and selection operator method. The nomogram was developed by clinical factors and the radiomics score.Results1409 features were extracted and a radiomics model consisting of 19 features was developed, which showed a favorable performance in discriminating GHAC from GAC (AUC(training cohort) = 0.998, AUC(internal validation set) = 0.942, AUC(external test cohort) = 0.731). The radiomics nomogram, including the radiomics score, AFP, and CA72_4, achieved good calibration and discrimination (AUC(training cohort) = 0.998, AUC(internal validation set) = 0.954, AUC(external test cohort) = 0.909).ConclusionsThe noninvasive CT-based nomogram, including radiomics score, AFP, and CA72_4, showed favorable predictive efficacy for differentiating GHAC from GAC and might be useful for clinical decision-making.