Abstract
Background and purpose
Gastric reactive hyperplasia (RH) is a common benign lesion of the gastric mucosa that can be resolved by conservative treatment without endoscopic intervention. Some RH lesions are indistinguishable from low-grade intraepithelial neoplasia (LGIN) lesions of gastric mucosa under endoscopy. The aim of this study was to investigate the morphological features of RH lesions under magnifying endoscopy combined with narrow-band imaging (ME-NBI).
Methods
A retrospective study of 653 patients with superficial suspicious lesions of gastric mucosa was performed. According to the pathological results of biopsies, the final included lesions were divided into the RH group (n = 88) and LGIN group (n = 138). We analysed the microvascular and microsurface patterns of these lesions under ME-NBI, extracted the most significant combination of endoscopic features of RH lesions, and evaluated their diagnostic performance.
Results
ME-NBI characteristics that could distinguish RH lesions from LGIN lesions after univariate analysis were included in multivariate logistic regression. The results showed that ten characteristics, including intervening part (IP) length homogeneity, type III gastric pit pattern and homogeneity of marginal crypt epithelium (MCE), were statistically significant. Receiver operating characteristic (ROC) analysis showed that the triad of these features was the best combination for diagnosing RH lesions with an AUC of 0.886 (95% confidence interval; 0.842–0.929), the sensitivity of 85.5% and specificity of 79.5%.
Conclusions
The triad of IP length homogeneity, type III pit pattern and MCE homogeneity under ME-NBI helps endoscopists to identify gastric RH lesions, thereby avoiding unnecessary biopsy and repeat endoscopy due to misjudgment of neoplastic lesions.
Acknowledgments
We thank the Digestive Endoscopy Center of Shanghai General Hospital and the Digestive Endoscopy Center of Shanghai Sixth People’s Hospital for their assistance.
Author contributions
Study concept and design: Jing Wang and Hui Zhou. Data acquisition: Wenlu Niu, Leheng Liu, Zhixia Dong, Tiancheng Mao, Xinjian Wan, Hui Zhou and Jing Wang. Manuscript drafting: Wenlu Niu, Hui Zhou, Jing Wang and Leheng Liu. Critical review of the manuscript: Jing Wang, Hui Zhou and Xinjian Wan. Statistical analysis: Wenlu Niu, Leheng Liu and Xiaowan Wu. All the authors participated in the interpretation of study results and approved the final version of the manuscript.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Correction Statement
This article has been republished with minor changes. These changes do not impact the academic content of the article.