RESUMO
BACKGROUND: Barrett's esophagus is generally accepted to be a premalignant condition. Previous studies have suggested the use of methylene blue (MB) chromoendoscopy to aid the identification of dysplasia in Barrett's esophagus surveillance programs, but a recent study has raised the concern that MB might induce oxidative damage of DNA. OBJECTIVE: The aim of this study was to compare MB directed biopsies (MBDB) with our current standard, which is random 4 quadrant biopsies (RB). DESIGN: A randomized prospective crossover study. SETTING: Single center. PATIENTS: Patients with a diagnosis of dysplasia identified in Barrett's esophagus within a 2-year period before entering the study. INTERVENTIONS: Either 4 random quadrant biopsies taken every 2 cm through the length of the Barrett's esophagus or MBDB from unstained or heterogenously stained mucosa. MAIN OUTCOME MEASUREMENTS: The number of patients with a diagnosis of dysplasia by each intervention. LIMITATIONS: Thirty-six percent of eligible patients declined the invitation to participate. RESULTS: Thirty patients completed the crossover study. The median length of Barrett's esophagus was 5 cm (interquartile range [IQR] 3-9 cm). At baseline histology, grades were as follows: 17 low-grade dysplasia (LGD), 3 high-grade dysplasia (HGD), and 10 no dysplasia. At completion, there were 10 LGD, 8 HGD, and 12 no dysplasia. Overall, dysplasia was identified in 17 of 18 patients by RB and in 9 of 18 by MBDB (McNemar test, p = 0.02). CONCLUSIONS: Our study showed MBDB to be significantly less sensitive in detecting dysplasia than RB in Barrett's esophagus. Hence, we discourage its use during routine surveillance of Barrett's esophagus.