Development and Evaluation of a Cemdct-Based Preoperative Risk Stratification Model to Predict Disease-Free Survival after Radical Surgery in Patients with Gastric Cancer.

Feng Caizhen,Cheng Jin,Zeng Xiao,Zhang Yinli,Hong Nan,Ye Yingjiang,Wang Yi
DOI: https://doi.org/10.1007/s00261-021-03049-0
IF: 2.4
2021-01-01
Abdominal Radiology
Abstract:To develop and evaluate a preoperative risk stratification model for predicting disease-free survival (DFS) based on contrast-enhanced multidetector computed tomography (ceMDCT) images in patients with gastric cancer (GC) undergoing radical surgery. We retrospectively enrolled patients with GC who underwent ceMDCT followed by radical surgery. A preoperative risk stratification model was constructed (including risk factor selection, risk status scoring, and risk level assignment) using Cox proportional hazard regression and log-rank analyses in the training cohort; the model was tested in the validation cohort. A nomogram was used to compare the preoperative risk stratification model with a postoperative DFS prediction model. A total of 462 patients (training/validation: 271/191) were included. The ceMDCT features of T category (score of 0 or 2), N category (0, 1, 2, or 3), extramural vessel invasion (0 or 2), and tumor location (0 or 1) were selected to construct the preoperative risk stratification model, with 4 risk levels defined based on risk score. There were significant differences in DFS among the risk levels in both cohorts (p < 0.001). The predictive value of the preoperative model was similar to that of the postoperative model, with concordance indices of 0.791 (95% CI, 0.743–0.837) and 0.739 (95% CI, 0.666–0.812), respectively, in the training cohort and 0.762 (95% CI, 0.696–0.828) and 0.738 (95% CI, 0.684–0.792), respectively, in the validation cohort. A preoperative risk stratification model based on ceMDCT images could be used to predict DFS and thus classify GC cases into various risk levels.
What problem does this paper attempt to address?