Abstract
Purpose
To critically examine the evidence for and against reversibility of gastric intestinal metaplasia (GIM) after Helicobacter pylori eradication, reframe the two-decade controversy, and propose a hypothesis-generating framework for risk-stratified management.Methods
Narrative synthesis of meta-analyses, cohort studies, randomized controlled trials, and mechanistic studies (2007-2026) identified through PubMed and Web of Science.Results
The evidence has swung from early meta-analyses (2007-2011) concluding against IM reversibility to recent syntheses (2023-2025) demonstrating regression with prolonged follow-up. In early 2026, two reviews reached opposing conclusions: Song and Yu proposed an oncogenic memory model of persistent mucosal injury, epigenetic and immune alterations, and microbiome dysbiosis, while Chen et al. maintained that part of GIM exhibits meaningful reversibility. Extending this framework, we propose a fourth, largely irreversible dimension-histological memory-encompassing spasmolytic polypeptide-expressing metaplasia (SPEM) progression and stem cell niche alteration, which deepens the mucosal injury component of oncogenic memory. Known predictive factors (OLGA/OLGIM stage, subtype, age, eradication timing, and oncogenic memory burden) are organized into a unified reversibility gradient that reframes the binary question in risk-stratified terms, from which a hypothesis-generating three-tier surveillance framework is derived.Conclusion
GIM reversibility is a continuous function of lesion characteristics, host factors, time since eradication, intervention intensity, and oncogenic memory burden rather than a binary property. The clinically relevant question is whose GIM, under what conditions, over what time frame, and with what residual risk reduction. The frameworks are hypothesis-generating, not validated clinical recommendations; prospective validation of risk-stratified surveillance represents the most urgent research priority.