Local lesion characteristics rather than H. pylori infection status predict endoscopic submucosal dissection bleeding: a large, confounder-adjusted study and a revised perspective.

backgroundWhether Helicobacter pylori (H. pylori) infection increases the risk of intraoperative bleeding during endoscopic submucosal dissection (ESD) for early gastric cancer (EGC) remains controversial. Previous studies often used a binary H. pylori classification and often inadequately controlled for local lesion characteristics, location, size, and invasion depth. This study aimed to re-evaluate the predictive value of H. pylori infection status through refined stratification and systematic adjustment of local factors.

methodsThis retrospective cohort study consecutively enrolled 799 EGC patients. Based on histological, serological, urea breath test, and endoscopic findings, patients were categorized into three groups: H. pylori-negative (n = 179), prior infection (evidence of prior infection without current activity, n = 454), and active infection (positive histology or urea breath test, n = 166). The primary outcome was intraoperative bleeding requiring active intervention. Multivariate logistic regression models were constructed, incorporating local lesion characteristics and key clinical factors. Interaction, subgroup, and sensitivity analyses were performed to verify the robustness of the findings.

resultsNo significant difference in intraoperative bleeding rates was observed among the three groups (28.5 vs. 26.7 vs. 24.7%, P = 0.728). Multivariate analysis identified larger lesion size and submucosal invasion as independent risk factors. A potential association with dual antiplatelet interruption was noted but based on only four events. Higher platelet count and lesion location in the lower third of the stomach were protective factors. H. pylori infection status was not independently associated with bleeding risk. Interaction, subgroup, and sensitivity analyses consistently supported this conclusion.

conclusionIn this large-sample study with rigorous adjustment for confounding factors, H. pylori infection status was not an independent risk factor for intraoperative ESD bleeding. Our findings suggest that risk assessment should prioritize local lesion characteristics and key clinical factors over H. pylori infection status. This hypothesis-generating observation awaits validation in prospective multicenter studies.