A randomized comparison of methylene blue-directed biopsy versus conventional four-quadrant biopsy for the detection of intestinal metaplasia and dysplasia in patients with long-segment Barrett's esophagus.

Objectives: Methylene blue (MB) selectively stains specialized intestinal metaplasia (SIM) and may assist in surveying a columnar-lined esophagus for Barrett's esophagus associated dysplasia.

Methods: This is a prospective, randomized crossover study comparing 4-quadrant random biopsies (4QB) versus MB-directed biopsies for the detection of SIM and dysplasia in 48 patients with long segment Barrett's esophagus (LSBE). Patients randomly underwent two endoscopies over a 4-wk time period with either 4QB or MB-directed biopsies as their first or second exam. Our aim was to correlate stain intensity with histology.

Results: The sensitivity of MB for SIM and dysplasia was 75.2% and 83.1%, respectively. The yield of 4QB for identifying nondysplasia SIM was 57.6% (523/917) and for dysplasia was 12% (111/917). Dark staining was significantly associated with histologic grade (P < 0.007). The final diagnosis was correct in 43 (90%) patients using MB and in 45 (94%) using 4QB. The 4QB technique missed dysplasia in 3 of 21 patients while MB biopsies missed dysplasia in 5 of 21 patients. The discordance between the two techniques was not significant (P= 0.727, McNemar's test). The mean number of biopsies taken during 4QB was 18.92 +/- 6.36 and with MB was 9.23 +/- 2.89 (P < 0.001).

Conclusion: MB requires significantly fewer biopsies than 4QB to evaluate for SIM and dysplasia. Dark staining correlates more with HGD than LGD in our experience. While MB is not more accurate than 4QB, MB may help to define areas to target for biopsy during surveillance endoscopy in patients with LSBE.