Gastroduodenal Mucosal Injury in Liver Cirrhosis: A Composite Score Analysis of Site-Specific Distribution and Risk...

Background: Liver cirrhosis is associated with portal hypertension and its gastrointestinal complications, but non-variceal gastroduodenal mucosal injury and its relationship to Helicobacter pylori (H. pylori) infection and hepatic disease severity remain incompletely characterized. Methods: This retrospective, cross-sectional study included 162 patients undergoing first upper gastrointestinal endoscopy (61 with cirrhosis, 101 controls), assessed using a composite gastroduodenal mucosal injury score (range 0-16) across the antrum, gastric corpus, duodenal bulb, and second duodenal portion. Results: Cirrhosis was independently associated with a higher injury burden after adjusting for H. pylori status, nonsteroidal anti-inflammatory drug (NSAID), antiplatelet, anticoagulant, and proton pump inhibitor (PPI) use (adjusted count ratio [aCR] = 1.33, 95% CI: 1.09-1.63, p = 0.006); and H. pylori infection was independently and separately associated with higher injury burden (aCR = 1.25, 95% CI: 1.01-1.54, p = 0.037), while none of the four medication classes showed an independent association. This excess injury was concentrated almost exclusively in the gastric corpus (adjusted aCR = 3.04, p < 0.0001), with no significant differences at other sites. Injury severity did not correlate with Child-Pugh class or tests of the liver's biosynthetic capacity (all p > 0.05). Only patients with concurrent cirrhosis and H. pylori infection showed significantly elevated injury relative to unexposed controls (adjusted aCR = 1.87, 95% CI: 1.38-2.53, p < 0.0001), with a significant, exploratory cirrhosis × H. pylori interaction (aCR = 1.56, p = 0.037). Conclusions: Cirrhosis is therefore associated with a corpus-predominant pattern of gastroduodenal injury, compatible with, though not proof of, a portal hypertension-related mechanism, and shows a significant positive multiplicative interaction with concurrent H. pylori infection; given the exploratory nature of this analysis, these findings are hypothesis-generating and warrant prospective confirmation before informing endoscopic screening or infection-screening practice.