White light and/or magnifying endoscopy with narrow band imaging for superficial nonampullary duodenal epithelial tumors

Abstract

Background and aim A reliable optical diagnosis of superficial nonampullary duodenal epithelial tumors (SNADETs) to guide optimal treatment strategy is lacking. The aim of this study was to simulate the treatment outcomes based on optical diagnosis using white light imaging (WLI) or magnifying endoscopy with narrow band imaging (MNBI) and to evaluate the method to extract optimal lesions suitable for cold snare polypectomy (CSP) or not. Methods We created a decision tree model using WLI and MNBI diagnosis for ideal treatment for SNADETs. Optical diagnoses of Vienna category 3 lesions (C3), category 4/5 (C4/5) were defined based on the WLI scoring system or the MNBI pattern diagnosis. Ideal treatments were CSP for C3 < 10mm, and endoscopic mucosal resection, endoscopic submucosal dissection or surgery for lesions ≥10mm or C4/5. Ideal treatment results based on optical diagnosis were analyzed according to actual pathological results. Results A total of 218 lesions with pre-operative diagnosis of SNADETs were included for analysis. Percentage of ideal treatment for C3 < 10mm based on WLI or MNBI or both was 88%, 83% and 97%. When WLI and MNBI diagnoses predicted the same histology, the specificity, positive predictive value and accuracy were significantly higher than those of WLI or MNBI alone (p < .05). Conclusions A decision tree model using lesion diameter, WLI scoring and MNBI pattern diagnosis is hypothetically useful to select ideal lesions for CSP among SNADETs. Optical diagnosis using both WLI and MNBI is more useful than WLI or MNBI alone.