Background: Curative treatment of patients with gastric cancer requires reliable detection of early gastric cancer. Magnifying endoscopy with narrow-band imaging (M-NBI) is useful for the accurate preoperative diagnosis of early gastric cancer. However, the role of M-NBI in screening endoscopy has not been established. The aims of this study were to determine the feasibility and limitations of M-NBI in screening endoscopy.
Methods: We conducted a multicenter prospective uncontrolled trial of patients undergoing routine screening endoscopy patients. We determined the diagnostic accuracy, sensitivity and specificity of M-NBI according to the degree of certainty and need for biopsy, as assessed using the VS (vessel plus surface) classification system. We analyzed the endoscopic and histopathological characteristics of both false negative and false positive high confidence M-NBI diagnoses. We then developed a provisional diagnostic strategy based on the diagnostic performance and limitations identified in this study.
Results: A total of 1097 patients were enrolled in the study. We analyzed 371 detected lesions (20 cancers and 351 non-cancers). The accuracy, sensitivity and specificity of high confidence M-NBI diagnoses were 98.1, 85.7 and 99.4 %, respectively. The false negative case was a pale mucosal lesion with tissue diagnosis of signet-ring cell carcinoma. Exclusion of pale mucosal lesions increased the accuracy, sensitivity and specificity of high confidence M-NBI diagnoses to 99.4, 100 and 99.4 %, respectively. We therefore propose a practical strategy targeting non-pale mucosal lesions.
Conclusions: With a refined strategy considering its limitations, M-NBI can act as an "optical biopsy" in screening endoscopies.