Methylene blue staining in Barrett's esophagus

We read with interest the article by Dave et al. on the usefulness of methylene blue staining in Barrett's esophagus. This article has followed two well-conducted studies by Canto et al. and Kiesslich et al. involving over 100 patients, on the value of methylene blue staining in Barrett's esophagus published within the last 12 months in this journal.

These two studies have shown encouraging results with high specificity and sensitivity for the detection of intestinal metaplasia, although they have differed in the dysplasia detection rate. It was unfortunate that Dave et al. enrolled only 9 patients and the study was terminated on safety grounds, which was not a problem with the above-mentioned studies. There are several limitations to the study, which the authors themselves have pointed out: limited patient number, nonrandomization, absence of a validated questionnaire to assess patient discomfort, and the use of midazolam, which may be associated with retrograde amnesia. More importantly, photographs of the staining pattern were not obtained. Interpretation of the staining pattern plays a major role in this kind of a study. This could vary with individual endoscopists. Canto et al. have described the staining pattern as diffuse and nondiffuse staining, depending on the percentage area stained. From our personal experience (data due to be published) we have also noted focal staining and heterogeneous staining patterns. Ideally biopsies should be targeted toward the stained and unstained mucosa for comparison, after spraying methylene blue. These results should be compared with random biopsies without using methylene blue staining. It is not surprising that Dave et al. could not reproduce the high specificity and sensitivity achieved by the other two groups of investigators with the above-mentioned limitations. The learning curve is also an important confounding factor in this study with a limited number of patients. Perhaps a more valid assessment could have been made if the patient number recruited were appropriately larger.

The aim of any identification procedure for Barrett's metaplasia is first to reduce the number of biopsy specimens taken and target them to suspicious areas, and second to help in the identification of groups at high risk for esophageal adenocarcinoma.