[The feasibility analysis of multi-slice spiral CT features of gastric stromal tumors in predicting pathological National Institutes of Health risk classification]
S K Wang,M L Wang,J Z Luo,G Y He,M S Si,Mengsu Zeng
DOI: https://doi.org/10.3760/cma.j.cn112137-20210831-01985
2022-04-05
Abstract:Objective: To investigate the feasibility of multi-slice spiral CT(MSCT) imaging features of gastric stromal tumor (GST) in predicting pathological NIH risk classification, providing imaging basis for patients with GST before treatment. Methods: The clinical and CT imaging data of 504 patients(506 GST lesions), 259males and 245 females, aged from 13 to 85(60±11) years, with GST confirmed by surgery and pathology collected in the Zhongshan Hospital Affiliated to Fudan University and the Affiliated TCM Hospital of Southwest Medical University. According to pathological NIH risk classification, 506 lesions were divided into low risk group (very low and low risk degree, 277 lesions) and high risk group (medium and high risk degree, 229 lesions).Clinical data and imaging characteristics were compared between two groups. Multivariate logistic regression analysis was performed to screen out independent risk factors for statistically significant imaging indicators. Receiver operating curve (ROC) was used to evaluate the predictive value of tumor length for risk classification. Resulst: Between low risk group and high risk group,there were significant differences in gender(male/female:131/146 vs 129/100), gastrointestinal bleeding(present/absent:39/238 vs 59/170), morphology(regular/Irregular:218/59 vs 95/134), calcification(present/absent:36/241 vs 53/176), degree of necrosis(0°/Ⅰ°/Ⅱ°/Ⅲ°:197/61/16/3 vs 58/98/32/41), ulceration(present/absent:32/245 vs 94/135), growth pattern(endophytic/exophytic/mixed:102/105/70 vs 44/98/87), tumor location(fundus/cardia/body/angle/antrum:98/7/135/12/25 vs 98/6/114/5/6), feeding artery(present/absent:32/245 vs 104/125), vascular enhancement(present/absent:19/258 vs 88/141), effusion of around the disease(present/absent:0/277 vs 13/216), positive sign of fat around the disease(present/absent:0/277 vs 30/199),maximum long diameter[2.82(2.04,3.80) cm vs 5.93(4.06,8.29) cm] and short diameter [2.31(1.60,2.88) cm vs 4.40(3.21,6.37) cm]of tumor(all P<0.05).The maximum long diameter of tumor (OR=2.08,95%CI:1.35-3.20) and ulceration positive(OR=2.01,95%CI:1.03-3.92) were independent risk factors of risk classification(all P<0.05).Gastric antrum was used as the reference for tumor location, gastric fundus(OR=7.77,95%CI:2.00-30.24) and gastric body(OR=3.93,95%CI:1.03-15.01) were independent risk factors of risk classification(all P<0.05).The area under curve(AUC) of the maximum long diameter of tumor for predicting risk classification was 0.87, and the optimal critical value, sensitivity and specificity were 4.98cm, 62.9% and 95.3% respectively. Conclusions: MSCT image features of GST had certain characteristics. MSCT has certain predictive value for pathological NIH risk classification of GST, which can provide certain imaging basis for patients before treatment.