The role of conventional CT examination, CT perfusion and tumor tissue texture analysis in predicting the risk of gastrointestinal stromal tumors

Abstract

As the most common mesenchymal tumors of the digestive tract, gastrointestinal stromal tumors (GIST) can have significant metastatic potential with a high rate of dissemination to the liver and peritoneal cavity. Based on the modified NIH (National Institutes of Health) classification, these tumors are classified into four stages: very low, low, intermediate and high-risk tumors. For low-risk tumors (LRG), minimally invasive surgery with R0 resection is the most common therapeutic option, while for high-risk tumors (HRG), imatinib adjuvant therapy is included post-surgically. Preoperative risk stratification is important for determining the therapeutic treatment and the patient's prognosis. With modern advances in diagnostic modalities, imaging analysis is no longer limited to visual assessment. As a modality of functional multidetector CT diagnostics, CT perfusion enables the quantification of tumor vascularization, while texture analysis provides a quantitative assessment of the degree of tumor heterogeneity, which overall can help in differentiating the metastatic potential of these tumors. The importance of morphological characteristics of tumors obtained by conventional CT examination in risk prediction has already been proven by numerous studies, and it is considered that, together with the advanced CT techniques mentioned above, it can provide a predictive model for the preoperative assessment of GIST risk stratification.