Quantitative Diagnosis of Atrophic Gastritis by Probe-Based Confocal Laser Endomicroscopy.

Aims. The aims of this study were to characterize nonatrophic and atrophic gastric mucosa under conventional endoscopy and probe-based confocal laser endomicroscopy (pCLE) modes and to define quantitative diagnostic parameters for these lesions under pCLE. Method. In phase I, 64 patients with gastric mucosal lesions diagnosed by gastrointestinal endoscopy were enrolled in the study. Normal mucosa and suspicious lesions were evaluated under normal white light imaging (WLI) and pCLE mode. Descriptive characteristic of gastric mucosal inflammation and atrophy under pCLE were defined according to the histology. In phase II, the criteria for nonatrophic gastritis (NAG) and chronic atrophic gastritis (CAG) under pCLE were used to diagnose the mucosal lesions in 431 patients. Diagnostic accuracy of each endoscopy modes was evaluated by measuring the concordance with histology. Result. A total of 64 patients with 187 positions were enrolled in the first part of this study. According to the histological diagnosis, the vessel diameter was increased in the NAG (11.18 +/- 0.1 mu m) and CAG (13.21 +/- 0.29 mu m) groups compared to the normal group (10.58 +/- 0.13 mu m); meanwhile, the distance between glands was 17.75 +/- 0.51 mu m in the normal group, 22.38 +/- 0.45 mu m in the NAG group, and 34.66 +/- 0.82 mu m in the CAG group, which increased significantly compared to nonatrophic mucosa. In order to differentiate atrophic mucosa from nonatrophic mucosa in real time, the cutoff value between these two kinds of lesions was >30 mu m in distance between glands. In phase II, 431 patients with 431 positions were evaluated under pCLE by using the criteria above. The sensitivity, specificity, PPV, and NPV for the diagnostic parameter were 90.3%, 78.8%, 85.1%, and 85.8%. The consistency of pCLE (Kappa value = 0.698) with histology was much better than WLI (Kappa value = 0.393). Conclusion. pCLE shows high potential for the diagnosis of gastric inflammation and atrophy based on quantitative criteria and has the ability to be a substitute for histology in the diagnosis of diffuse lesions in the stomach.