Prospective evaluation of narrow-band imaging endoscopy for screening of esophageal squamous mucosal high-grade neoplasia in experienced and less experienced endoscopists.

Narrow-band imaging (NBI) is a novel, noninvasive optical technique that uses reflected light to visualize the organ surface. However, few prospective studies that examine the efficacy of NBI screening for esophageal cancer have been reported. To compare the diagnostic yield of NBI endoscopy for screening of squamous mucosal high-grade neoplasia of the esophagus between experienced and less experienced endoscopists. Patients with a history of esophageal neoplasia or head and neck cancer received NBI endoscopic screening for esophageal neoplasia followed by chromoendoscopy using iodine staining. Biopsy specimens were taken from iodine-unstained lesions and the histological results of mucosal high-grade neoplasias served as the reference standard. The primary outcome was the sensitivity of NBI for detecting new lesions. The secondary outcome was the positive predictive value of NBI and the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of NBI in a per lesion basis. A total of 350 patients (170 by experienced endoscopists and 180 by less experienced endoscopists) underwent endoscopic examination. A total of 42 new mucosal high-grade neoplastic lesions (25 in the experienced endoscopist group and 17 in the less experienced endoscopist group) were detected. In the per-lesion-based analysis, the sensitivity was significantly higher in the experienced endoscopist group (100%; 25/25) compared with the less experienced endoscopist group (53%; 9/17) (P < 0.001). The positive predictive value of NBI was higher in the experienced endoscopist group than in the less experienced endoscopist group (45%, 25/55 vs. 35%, 9/26), although the difference was not significant (P = 0.50). The sensitivity of NBI in the less experienced endoscopist group was 43% in the former half of patients, and increased to 60% in the latter half of patients. In the per-patient-based analysis, the sensitivity of NBI was significantly higher in the experienced endoscopist group (100%) than in the less experienced endoscopist group (100 vs. 69%, respectively; P = 0.04). The positive predictive values of the experienced endoscopist group and the less experienced endoscopist group were similar, and were 48 and 47%, respectively. In conclusion, compared with the gold standard of chromoendoscopy with iodine staining, the sensitivity of NBI for screening of mucosal high-grade neoplasia was 100% with the experienced endoscopists but was low with the less experienced endoscopists. Electronic chromoendoscopy with NBI is a promising screening tool in these high-risk patients with esophageal mucosal high-grade neoplasia, particularly when performed by endoscopists with experience of using NBI.