Abstract
BACKGROUND: Preoperative distinguishing between benign and malignant gastrointestinal stromal tumors (GISTs) poses a challenge. Ultrasound elastography has emerged as a promising diagnostic tool; however, further investigation is needed to assess its diagnostic accuracy in evaluating GISTs. AIM: To evaluate the accuracy of ultrasound elastography for differentiating between benign and malignant GISTs. METHODS: This prospective study included 110 patients with 103 histopathologically confirmed GISTs between January 2021 and December 2023. All tumors underwent conventional ultrasound examination, strain elastography (SE), and shear-wave elastography (SWE) before surgical resection. The study evaluated elastographic parameters such as strain ratio, elastographic patterns, mean elastic modulus, and heterogeneity index. Diagnostic performance was evaluated using receiver operating characteristic curve analysis, with histopathological diagnosis as the reference standard. RESULTS: Of the 103 GISTs, 45 (43.7%) were benign and 58 (56.3%) were malignant based on modified National Institutes of Health criteria. Malignant GISTs exhibited significantly higher strain ratios (4.82 ± 1.73 vs 2.31 ± 0.89; P < 0.001) and mean elastic modulus values (45.6 ± 15.8 kPa vs 21.3 ± 8.4 kPa; P < 0.001) than benign tumors. The optimal cutoff values were 3.45 for the strain ratio (sensitivity: 84.5%, specificity: 86.7%) and 32.5 kPa for the mean elastic modulus (sensitivity: 87.9%, specificity: 88.9%). The areas under the curve were 0.892 and 0.918, respectively. Interobserver agreement was excellent for both SE [intraclass correlation coefficient (ICC) = 0.88] and SWE (ICC range: 0.85-0.93) measurements. CONCLUSION: Ultrasound elastography shows high diagnostic accuracy in distinguishing between benign and malignant GISTs. Combining SE and SWE provides complementary parameters for preoperative risk stratification.