Objective: To evaluate the diagnostic utility of routine gastric biopsies taken at random versus targeted biopsies with methylene blue staining for the diagnosis of intestinal metaplasia.
Material and methods: We performed a validation study in 376 patients. We performed 2 antral biopsies, 2 gastric body biopsies, 1 biopsy for urease test and additional biopsies if demanded. One hundred and one patients underwent 2 biopsies after methylene blue staining. A total of 1,486 biopsies were analyzed. Frequencies of histological and endoscopic diagnosis and the validation of endoscopic diagnosis of intestinal metaplasia with and without staining against histology, which followed updated Sydney and Vienna criteria, were determined
Results: Metaplasia was found in 37.23% ofcases and in 6.38% was incomplete in antrum and body, moderate or severe. The endoscopic diagnosis had: sensitivity 46%, specificity 91%, positive predictive value (PPV) 75% and negative predictive value (NPV) 74%. The low sensitivity indicates the need for gastric biopsies. Staining had: sensitivity 95%, specificity 67%, PPV 84% and NPV 89%, with significant difference for staining (P < 0.05). The typical endoscopic appearance of intestinal metaplasia always coincided with staining and histology. Dysplasia was found in 0.53% ofcases, gastric cancer in 1.8% and Helicobacter pylori in 56%.
Conclusion: Directed biopsy staining is superior to those taken at random for diagnosing metaplasia. Metaplasia was not found when endoscopy was normal and typical endoscopic appearance of metaplasia agreed with staining and histology.