Can we accurately diagnose minute gastric cancers (≤5 mm)? Chromoendoscopy (CE) vs magnifying endoscopy with narrow band imaging (M-NBI)

Chromoendoscopy (CE) is relatively ineffective at identifying the cancer-specific morphological characteristics of minute gastric cancers less than or equal to 5 mm in diameter, and on its own is insufficient to make an accurate diagnosis. The aim of this study is to assess the diagnostic performance of magnifying endoscopy with narrow band imaging (M-NBI) for minute gastric cancers. The minute cancer group comprised consecutive endoscopic submucosal dissection-resected minute gastric cancers histologically measured as no larger than 5 mm in diameter. The non-cancer group comprised consecutive non-cancer lesions no larger than 5 mm in diameter. The two groups were subject to retrospective analysis to evaluate the diagnostic ability (sensitivity, specificity, and diagnostic accuracy) and reproducibility of CE and M-NBI. The results for CE versus M-NBI were as follows: sensitivity 43.7 % (95 % CI, 26.5-61.0 %) versus 78.0 % (95 % CI, 64.0-92.0 %); specificity 81.6 % (95 % CI, 72.6-90.6 %) versus 92.9 % (95 % CI, 87.0-98.9 %); and diagnostic accuracy 69.9 % (95 % CI, 61.0-78.6 %) versus 88.3 % (95 % CI, 82.0-94.5 %). The sensitivity and diagnostic accuracy were, therefore, significantly higher for M-NBI than for CE. The inter-observer variability was kappa = 0.08 for CE and kappa = 0.56 for M-NBI, while the intra-observer variability was kappa = 0.38 and kappa = 0.65, respectively. M-NBI has greater sensitivity and reproducibility than CE for the diagnosis of minute gastric cancers.