Clinicopathological characteristics of gastrointestinal mesenchymal tumors and diagnostic value of endoscopic ultrasonography

Abstract

Gastrointestinal mesenchymal tumors (GIMTs) are tumors which arise from mesenchymal cells other than lymphocytes or epithelial cells and include gastrointestinal stromal tumors (GISTs), smooth muscle tumors, neurogenic tumors, fibroblast tumors, and liparomphalus. Here, we studied the clinicopathologic characteristics of GIMTs and determined the diagnostic value of endoscopic ultrasonography (EUS) in GIMTs. The morphological characteristics of 415 GIMT cases were observed using light microscopy. The expression of CD117, CD34, smooth muscle actin (SMA), S-100, and Ki-67 were detected using immunohistochemistry. EUS results were retrospectively analyzed in 76 cases. Among the 415 GIMT cases, there were 229 GIST cases, 178 smooth muscle tumor cases, and eight neurogenic tumor cases. Diffuse strong expression of CD117 was observed in 202 (88.2%) GISTs and strong expression of CD34 was observed in 135 (59.0%) GISTs. SMA and S-100 expression was detected in smooth muscle tumors (leiomyomas/leiomyosarcoma) and neurogenic tumors, respectively. Of the 12 cases with GISTs, nine cases with positive ki-67 expression and accompanied with more mitosis were pathologically diagnosed as malignant GISTs. Most of the borderline and malignant tumors were found in male patients. EUS was 98.7% accurate in determining GIMT locations, 82.9% accurate in diagnosing GIMTs, and 80.3% accurate in differentiating benign from malignant tumors. GISTs, rather than leiomyomas, are the most common GIMTs. Immunohistochemical markers, such as CD117, CD34, SMA, and S-100, can differentiate GISTs from smooth muscle and neurogenic tumors. The expression of Ki-67 in patients with significant karyokinesis supports a malignant tumor diagnosis. EUS is an accurate method for detecting GIMTs and differentiating between benign and malignant tumors. Use of endoscopic ultrasound-guided fine-needle aspiration biopsies may enable clinicians to make more accurate diagnoses than currently used methods.