Predicton of risk classification of duodenal stromal tumor based on computed tomography imaging features
10.3760/cma.j.cn311367-20240708-00267
- VernacularTitle:基于计算机断层扫描特征预测十二指肠间质瘤危险度分级的应用价值
- Author:
Wenjie YAN
1
;
Haiyan YU
;
Chuanfang XU
;
Mengsu ZENG
;
Mingliang WANG
Author Information
1. 复旦大学附属中山医院放射科，上海 200032
- Publication Type:Journal Article
- Keywords:
Tomography, X-ray computed;
Gastrointestinal stromal tumor;
Duodenum
- From:
Chinese Journal of Digestion
2024;44(12):811-817
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To evaluate the role of computed tomography (CT) imaging features in predicting the risk classification of duodenal stromal tumor (DST), so as to assist clinical treatment decision-making.Methods:A total of 109 patients with DST confirmed by surgical pathology at Zhongshan Hospital, Fudan University were retrospectively enrolled from November 1, 2013 to November 30, 2022. According to the National Institutes of Health risk classification criteria, the 109 DST patients were divided into low-risk group (including very low- and low-risk patients, 70 cases) and high-risk group (including moderate- and high-risk patients, 39 cases). The CT features of the 2 groups were collected, including arterial and venous phase CT values, tumor size index, shape (regular or not), edge (clear or not), growth pattern (intracavity, extracavity or both), degree of cystic necrosis (0 degree, Ⅰ degree, Ⅱ degree, Ⅲ degree), and the presence of enlarged feeding arteries and vascular-like enhancement were collected. Univariate and multivariate logistic regression analyses were performed to identify the CT features related to DST risk classification. Receiver operating characteristic curve (ROC) was drawn to assess the predictive value of the features.Results:The result of univariate logistic regression analysis showed that tumor size index ( P<0.001), arterial phase CT value ( P<0.001), venous phase CT value ( P<0.001), shape ( P<0.001), edge ( P=0.004), growth pattern ( Pintracavity with extracavity=0.004), degree of cystic necrosis ( PⅡ degree=0.003, PⅢ degree=0.002), enlarged feeding arteries ( P=0.001), and vascular-like enhancement ( P=0.004) were all related to DST risk classification. The result of multivariate logistic regression analysis revealed that tumor size index was an independent risk factor for predicting DST risk classification ( OR=1.35, 95% confidence interval (95% CI) 1.10 to 1.65, P=0.003). The result of ROC analysis demonstrated that the area under the curve for tumor size index in predicting DST risk classification was 0.909 (95% CI 0.841 to 0.977), with sensitivity of 82.1% and specificity of 92.9%. Conclusion:Tumor size index based on CT imaging has good predictive performance for DST risk classification, and provides valuable assistance for clinical diagnosis and treatment decisions.
