Impact of advanced endoscopic imaging on Barrett's esophagus in daily clinical practice

Several advanced imaging techniques have been proposed to improve the visualization of dysplastic regions within Barrett's epithelium, with some evidence for the use of narrow-band imaging (NBI) and acetic acid chromoendoscopy (AAC).

We retrospectively analyzed consecutive cases of Barrett's esophagus, diagnosed using white-light endoscopy and confirmed histologically by the presence of intestinal metaplasia, between April 2007 and April 2010 in a large community hospital. A change in practice was then instituted, whereby a Barrett's team consisting of specialist endoscopists was formed in an attempt to standardize and improve the quality of surveillance. Barrett's epithelium was inspected with both white-light imaging and NBI in all patients. Where the length of Barrett's epithelium was 3 cm or more, AAC was also used. One and a half percent acetic acid was sprayed onto the Barrett's segment and loss of aceto-whitening observed after a 2-minute period. Any abnormal areas noted during advanced imaging underwent target biopsy sampling. We subsequently compared the dysplasia detection rate in Barrett's epithelium identified between April 2011 and April 2014 after these changes. Observed differences between the cohorts were analyzed with the Fisher exact test and the Student t test.

From 2007 to 2010 Barrett's esophagus was identified during 560 gastroscopies in 392 individual patients. The mean maximal Barrett's esophagus recorded length was 4.4 cm (range, 1-10), with an average of 4.7 esophageal biopsy specimens taken per endoscopy. In comparison, from 2011 to 2014 Barrett's esophagus was identified during 856 endoscopies in 630 patients. From 2011 to 2014 the Barrett's team performed 85% of all procedures using the aforementioned techniques. The mean maximal Barrett's esophagus length was 3.8 cm (range, 1-16), with an increased average of 5.8 biopsy specimens per endoscopy taken (P < .01). Both cohorts were comparable in age and gender distribution. Our data demonstrated no significant difference in the relative frequencies of occurrence of dysplasia detected between both cohorts of patients. From 2007 to 2010 dysplasia was detected in 11.0% (n = 43) of patients. This consisted of low-grade dysplasia in 7.7% of patients and high-grade dysplasia or cancer 3.3%. From 2011 to 2014 this compared with dysplasia in 11.3% (n = 71) of patients, with low-grade dysplasia in 9.4% and high-grade dysplasia or cancer in 1.9%.

Our data show that the use of NBI and AAC in the imaging of Barrett's esophagus did not result in an increased detection rate of dysplasia in routine clinical practice. These findings concur with the recommendations of existing Barrett's esophagus surveillance guidelines, which advocate the continued use of quadratic biopsy sampling within general surveillance programs.