Diagnostic value of non-magnifying gastroscopy followed by targeted biopsy for detection of early gastric cancer: Multicenter prospective study

Abstract Non-magnifying endoscopy remains the most essential tool for detecting early gastric cancer (EGC), but there is still a lack of widely accepted diagnostic methods. We established an optimized EGC detection protocol using non-magnifying gastroscopy-guided targeted biopsy of morphological suspicious lesions and evaluated its effectiveness. This study included 5738 participants across four medical centers in China. Targeted biopsies were performed on the following suspicious lesions for high-grade neoplasia (HGN) under screening non-magnifying gastroscopy: 1) ulcerative lesions; 2) esophagogastric junction reddish lesions outside the atrophic area; 3) pale or well-demarcated lesions outside the atrophic area, except polypoid lesions smaller than 5 mm; 4) elevated lesions with clear borders or uneven top within the atrophic area; and 5) flat/depressed lesions with irregular borders or uneven surface or ocher color under narrow-band imaging (NBI) within the atrophic area. Sensitivity for detecting gastric HGN and the positive predictive value (PPV) of targeted biopsy were calculated. The targeted biopsy method demonstrated a sensitivity of 90.9% (50/55, 95% confidence interval [CI] 83.1%-98.8%) and a PPV of 5.4% (50/931, 95% CI 3.9%-6.8%) for diagnosing HGN on per-lesion assessment. Lesions that met the suspicious morphological criteria carried a significant risk to be HGNs, even after adjusting for age, sex, and other risk factors associated with gastric cancer (adjusted odds ratio = 42.03, 95% CI 11.14–158.63, P < 0.001). Targeted biopsy of suspicious lesions for HGN with non-magnifying gastroscopy can be used as a primary clue for detecting EGC.