Copyright: ©Author(s) 2026. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution-NonCommercial (CC BY-NC 4.0) license. No commercial re-use. See permissions. Published by Baishideng Publishing Group Inc.
World J Gastroenterol. Jan 21, 2027; 33(3): 123685
Published online Jan 21, 2027. doi: 10.3748/wjg.123685
Computed tomography-based nomogram for predicting recurrence of gastrointestinal stromal tumors
Lian Zhao, Li-Ming Zhao, Xiao-Hua Zheng, Ke Zou, Jun Xie, Xi-Jiao Liu
Lian Zhao, Xiao-Hua Zheng, Department of Radiology, Chengdu Sixth People’s Hospital, Chengdu 610051, Sichuan Province, China
Lian Zhao, Xi-Jiao Liu, Department of Radiology, West China Hospital, Sichuan University, Chengdu 610041, Sichuan Province, China
Li-Ming Zhao, Department of Radiology, Sichuan Provincial People’s Hospital, University of Electronic Science and Technology of China, Chengdu 610072, Sichuan Province, China
Ke Zou, Department of Research and Development, Chengdu Finsiot Technology Co., Ltd, Chengdu 610000, Sichuan Province, China
Jun Xie, Information Technology Center, West China Hospital, Sichuan University, Chengdu 610041, Sichuan Province, China
Author contributions: Zhao L and Liu XJ participated in the conception and design of the study; Zhao L, Zhao LM, and Zheng XH were involved in the acquisition and analysis; Zhao L wrote the manuscript; Zou K and Xie J accessed and verified the study data. All authors critically reviewed and provided final approval of the manuscript, and were responsible for the decision to submit the manuscript for publication.
AI contribution statement: No artificial intelligence tools were used in the writing of this manuscript. The authors take full responsibility for its scientific content.
Supported by Sichuan Science and Technology Project, No. 2025YFHZ0322; Chengdu Science and Technology Project, No. 2026-YF11-00019-HZ; and Clinical Medicine Research Promotion Program, China Medical Foundation, No. 2024CMFA10.
Institutional review board statement: This study was approved by the Institutional Review Board of the West China Hospital, Sichuan University, No. 2022-449.
Informed consent statement: The need for patient consent was waived due to the retrospective nature of the study.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
Data sharing statement: Datasets are available from the corresponding author.
Corresponding author: Xi-Jiao Liu, MD, Department of Radiology, West China Hospital, Sichuan University, No. 37 Guoxue Lane, Wuhou District, Chengdu 610041, Sichuan Province, China.
liuxijiao@wchscu.cn
Received: June 2, 2026
Revised: July 26, 2026
Accepted: September 18, 2026
Published online: January 21, 2027
Processing time: 165 Days and 12.5 Hours
BACKGROUND
Gastrointestinal stromal tumors (GISTs) exhibit a high degree of biological heterogeneity, and accurate postoperative risk stratification remains a key yet challenging issue in clinical practice. Although surgical resection is the primary treatment for localized GISTs, postoperative recurrence and metastasis remain significant clinical challenges. Various methods for predictive risk stratification based on imaging and radiomics have been proposed; however, many of these methods require complex workflows or lack direct clinical applicability. Therefore, there is an unmet need for a practical, reliable, and interpretable model based on routinely available imaging data, in order to optimize risk stratification and achieve personalized treatment plans.
AIM
To develop and validate a computed tomography (CT)-based prognostic model for predicting recurrence-free survival (RFS) in patients with GISTs.
METHODS
This retrospective study included 417 patients with curatively resected, pathologically confirmed primary GISTs. Preoperative contrast-enhanced CT features, clinicopathologic variables and postoperative treatment details were collected. Least absolute shrinkage and selection operator selected key predictors, which were entered into a multivariable Cox baseline model. A random survival forest algorithm generated an additional risk score that was combined with independent predictors to create the final nomogram. Performance was assessed with the concordance index (C-index), time-dependent receiver operating characteristic curves, calibration curves, and decision curve analysis.
RESULTS
Enlarged feeding or draining vessels, peripheral fat infiltration, and the Ki-67 index were identified as independent predictors of RFS. The baseline model achieved a C-index of 0.829 (95% confidence interval: 0.770-0.884) in the training set. The integrated model showed a trend toward superior predictive performance, with a C-index of 0.867 (95% confidence interval: 0.822-0.913). The areas under the receiver operating characteristic curves for predicting 1-, 3-, and 5-year RFS were 0.932, 0.874, and 0.879 in the training set, and 0.741, 0.858, and 0.818 in the validation set, respectively. Calibration curves showed good agreement between predicted and observed outcomes, and decision curve analysis indicated potential clinical benefit.
CONCLUSION
An interpretable, CT-based nomogram integrating conventional imaging features and machine learning provides accurate prediction of recurrence in GIST patients. This model may assist in individualized risk stratification and postoperative management.
Core Tip: This research introduces a new prognostic model based on computed tomography imaging aimed at predicting recurrence-free survival for individuals diagnosed with gastrointestinal stromal tumors. By analyzing a dataset of 417 patients, we identified important imaging and clinical factors using advanced statistical techniques, such as least absolute shrinkage and selection operator regression and random survival forest methods. Our combined model showed an impressive performance with a concordance-index of 0.867, outperforming conventional strategies. This novel method not only improves risk assessment but also facilitates personalized management after surgery, representing a significant step forward in the care of gastrointestinal stromal tumor patients.
