Chronic atrophic gastritis in the fundus was assessed endoscopically by autofluorescence imaging [34].

Metachronous GC occurred in 12 of the 82 patients (14.6%) and was significantly associated with “open” type fundic atrophic gastritis (HR 4.88; 95% CI 1.32–18.2, p = .018). There is ongoing debate about the adequate assessment of gastritis for risk estimation and the necessity of surveillance for patients with risk gastritis. Rugge et al. [35] suggested application of the Operative Link for Gastritis Assessment (OLGA) system for prediction of neoplastic progression. Ninety-three patients followed for more than 12 years by upper gastrointestinal endoscopy and serum sampling for pepsinogens and H. pylori status developed invasive or intra-epithelial neoplasia only if they have been defined as at “high risk” according to the OLGA staging system, at inclusion DNA Damage inhibitor (OLGA III or IV). The mean pepsinogen I/II ratio was correlated with respective result of the OLGA stage. The assessment of the OLGA stage at inclusion allows LY294002 us to predict the OLGA stage at the end point as well as the incidence of neoplasia [36]. Another group from the Netherlands performed endoscopic surveillance in patients with either IM or dysplasia in the gastric mucosa [37]. Biopsies were taken both nontargeted from suspicious areas and

targeted from antrum, angulus, corpus, and cardia. At surveillance endoscopy, the highest prevalence of premalignant mucosal conditions was in the angulus (40%), then in the antrum (35%), and in the corpus (33%). High-grade dysplasia was present in targeted biopsies only. It was concluded that for adequate surveillance both targeted and nontargeted biopsies are necessary. Several authors assessed the cost-effectiveness of endoscopic surveillance of higher risk patients (e.g. patients with IM or gastric ulcer) [38,39]. In a decision model, endoscopy was scheduled yearly for a period of 10 years after new diagnosis of IM in the stomach, for a cohort of 10,000 American patients (compared with no

surveillance) [38]. With an estimated GC incidence of 1.8% per year, 556 and 3738 endoscopies were needed to detect one case of GC and to see more prevent one cancer-related death, respectively. Incremental cost-effectiveness ratio of endoscopic surveillance compared with nonsurveillance was US $ 72519 per life year gained. In a similar analysis on patients with gastric ulcer and an estimated cancer incidence of 2.6%, costs were US $ 146700 per quality-adjusted life year. The probability of cost-effectiveness was only 25.2% and was not evident unless the prevalence of undetected malignancy exceeds 6% [39]. For noninvasive stratification of patients at high risk for GC development, the analysis of serum pepsinogens combined with anti-H. pylori antibodies remains the best option.