BACKGROUND: Magnifying narrow-band imaging has enabled observation of the mucosal and vascular patterns of gastrointestinal lesions. This study investigated the potential value of magnifying endoscopy with narrow-band imaging for the classification of gastric intraepithelial neoplasia. METHODS: Seventy-six patients with gastric intraepithelial neoplasia (82 lesions) at People's Liberation Army General Hospital from December 2009 to November 2010 were analyzed. All patients underwent magnifying endoscopy with narrow-band imaging, and their lesions were differentiated into probable low-grade intraepithelial neoplasia or possible high-grade intraepithelial neoplasia on the basis of the imaging features. Pathologic proof was subsequently obtained by endoscopic submucosal dissection in every case. The validity of magnifying endoscopy with narrow-band imaging was calculated, considering histopathology to be the gold standard. RESULTS: Magnifying endoscopy with narrow-band imaging showed 22 low-grade intraepithelial neoplastic lesions and 60 high-grade intraepithelial neoplastic lesions. Of the 22 low-grade intraepithelial neoplastic lesions, 16 showed the same results on both imaging and pathology. Of the 60 high-grade intraepithelial neoplastic lesions, 53 showed the same results on both imaging and pathology. Thus, the sensitivity of magnifying endoscopy with narrow-band imaging for high-grade intraepithelial neoplasia was 89.83%, which was higher than that for low-grade intraepithelial neoplasia (69.57%). However, the specificity for high-grade intraepithelial neoplasia (69.57%) was lower than that for low-grade intraepithelial neoplasia (89.83%). The overall accuracy of magnifying endoscopy with narrow-band imaging was 84.15%. CONCLUSIONS: Magnifying endoscopy with narrow-band imaging can distinguish between gastric low- and high-grade intraepithelial neoplasia. It may be a convenient and effective method for the classification of gastric intraepithelial neoplasia.
BACKGROUND: Magnifying narrow-band imaging has enabled observation of the mucosal and vascular patterns of gastrointestinal lesions. This study investigated the potential value of magnifying endoscopy with narrow-band imaging for the classification of gastric intraepithelial neoplasia. METHODS: Seventy-six patients with gastric intraepithelial neoplasia (82 lesions) at People's Liberation Army General Hospital from December 2009 to November 2010 were analyzed. All patients underwent magnifying endoscopy with narrow-band imaging, and their lesions were differentiated into probable low-grade intraepithelial neoplasia or possible high-grade intraepithelial neoplasia on the basis of the imaging features. Pathologic proof was subsequently obtained by endoscopic submucosal dissection in every case. The validity of magnifying endoscopy with narrow-band imaging was calculated, considering histopathology to be the gold standard. RESULTS: Magnifying endoscopy with narrow-band imaging showed 22 low-grade intraepithelial neoplastic lesions and 60 high-grade intraepithelial neoplastic lesions. Of the 22 low-grade intraepithelial neoplastic lesions, 16 showed the same results on both imaging and pathology. Of the 60 high-grade intraepithelial neoplastic lesions, 53 showed the same results on both imaging and pathology. Thus, the sensitivity of magnifying endoscopy with narrow-band imaging for high-grade intraepithelial neoplasia was 89.83%, which was higher than that for low-grade intraepithelial neoplasia (69.57%). However, the specificity for high-grade intraepithelial neoplasia (69.57%) was lower than that for low-grade intraepithelial neoplasia (89.83%). The overall accuracy of magnifying endoscopy with narrow-band imaging was 84.15%. CONCLUSIONS: Magnifying endoscopy with narrow-band imaging can distinguish between gastric low- and high-grade intraepithelial neoplasia. It may be a convenient and effective method for the classification of gastric intraepithelial neoplasia.