Utility of preoperative computed tomography features in predicting the Ki-67 labeling index of gastric gastrointestinal stromal tumors. Issue 142 (September 2021)

• Record Type: Journal Article
• Authors: Chen, Xiao-Shan; Shan, Ying-Chan; Dong, San-Yuan; Wang, Wen-Tao; Yang, Yu-Tao; Liu, Li-Heng; Xu, Zhi-Han; Zeng, Meng-Su; Rao, Sheng-Xiang
• Is Part Of: European journal of radiology. Issue 142(2021)
• Publication Date: 2021-09
• Subjects: Computed tomography -- Gastrointestinal stromal tumors -- Gastric GISTs -- Ki-67 labeling index; CT computed tomography -- Ki-67LI Ki-67 labeling index -- GISTs gastrointestinal stromal tumors -- NIH National Institutes of Health -- ROI the region of interest -- OR odds ratio -- CI confidence interval -- 3D three-dimensional -- ICC Inter- and intraclass correlation coefficients; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2021.109840
• Languages: English
• ISSNs: 0720-048X

Abstract

Abstract: Purpose: To evaluate the value of preoperative computed tomography (CT) features including morphologic and quantitative features for predicting the Ki-67 labeling index (Ki-67LI) of gastric gastrointestinal stromal tumors (GISTs). Methods: We retrospectively included 167 patients with gastric GISTs who underwent preoperative contrast-enhanced CT. We assessed the morphologic features of preoperative CT images and the quantitative features including the maximum diameter of tumor, total tumor volume, mean total tumor CT value, necrosis volume, necrosis volume ratio, enhanced tissue volume, and mean CT value of enhanced tissue. Potential predictive parameters to distinguish the high-level Ki-67LI group (>4%, n = 125) from the low-level Ki-67LI group (≤4%, n = 42) were compared and subsequently determined in multivariable logistic regression analysis. Results: Growth pattern (p = 0.036), shape (p = 0.000), maximum diameter (p = 0.018), total tumor volume (p = 0.021), mean total tumor CT value (p = 0.009), necrosis volume (p = 0.006), necrosis volume ratio (p = 0.000), enhanced tissue volume (p = 0.027), and mean CT value of enhanced tissue (p = 0.004) were significantly different between the two groups. Multivariate logistic regression analysis indicated that lobulated/irregular shape (odds ratio [OR] = 3.817; p = 0.000) and high necrosis volume ratio (OR = 1.935; p = 0.024) were independent factors of high-level Ki-67LI. Conclusions: Higher necrosis volume ratio in combination with lobulated/irregular shape could potentially predict high expression of Ki-67LI for gastric GISTs.