Prospective validation of magnified endoscopic examination with image-enhanced endoscopy as an optical biopsy for differentiating superficial duodenal epithelial tumor and non-neoplastic lesion in duodenum

INTRODUCTION: Superficial duodenal epithelial tumors (SDETs) and non-neoplastic lesions (NNLs) often appear similar on conventional endoscopy, complicating real-time differentiation and leading to frequent biopsies. However, duodenal biopsy is invasive and may hinder subsequent endoscopic resection. We previously developed a diagnostic algorithm using magnified endoscopic examination with image-enhanced endoscopy (IEE-ME) to distinguish SDETs from NNLs. This study aimed to prospectively validate the diagnostic performance of this algorithm in clinical practice. METHODS: In this single-center prospective observational study, we evaluated an IEE-ME-based diagnostic algorithm for esophagogastroduodenoscopy. Lesions were classified according to predefined criteria, including superficial structure, white opaque substance, demarcation line, and enlarged marginal epithelium. Endoscopic diagnoses were made immediately by 11 endoscopists with varied experience, who were blinded to the final histopathological diagnoses. The diagnostic performance was assessed using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy. RESULTS: Among 241 lesions analyzed, 84 and 157 were histologically diagnosed as SDETs and NNLs, respectively. The diagnostic algorithm demonstrated excellent performance, with a sensitivity of 97.6% (95% confidence interval [CI]: 91.7-99.7%), specificity of 95.5% (95% CI: 91.0-98.2%), PPV of 92.1% (95% CI: 85.0-96.0%), NPV of 98.7% (95% CI: 95.0-99.7%), and overall accuracy of 96.3% (95% CI: 93.0-98.3%). CONCLUSION: This prospective validation showed that IEE-ME, applied through a structured algorithm, enables highly accurate and non-invasive differentiation between SDETs and NNLs. Given its excellent NPV and accuracy, this approach may serve as an optical biopsy method that reduces the need for conventional biopsies in clinical practice.