Prevalence of Barrett Esophagus in Adolescents and Young Adults With Esophageal Atresia
We thank Vergouwe et al1 for their comments about our recent prospective multicenter study published in a recent issue of this journal. They expressed concerns about the accuracy of the diagnosis Barrett esophagus (BE), suggesting a possible overestimation of the prevalence of BE in our cohort.
They first suggested that, after esophageal atresia (EA) repair, the anatomy is modified with a gastroesophageal junction (GEJ) often located proximal to the diaphragm. This statement is, however, not supported on published literature although it is usually recognized, based on clinical practice that the surgeon may tract on the lower-pouch to ensure an esophageal anastomosis without excessive tension. We agree that endoscopic landmarks, GEJ and BE may be more difficult to recognize and describe in this population, but all our patients underwent upper endoscopy under general anesthesia, which allows comfortable landmarks description, and were performed by experimented gastroenterologists from EA expert centers. In the protocol, different solutions for an accurate diagnosis was offered including standardized staining techniques (acetic acid or electronic chromoendoscopy) and sending landmarks’ picture for a centralized confirmation.
To answer a second criticism, the study protocol indeed defined a BE regarding to the true position of the GEJ as defined by Sharma et al2; mentioning the Z-line in the published article was a mistake and we apologize for that.
Another suspected flaw of our study was that we found histological evidence of BE in 51 patients, whereas only 44 patients had suspected BE at endoscopy. This is in line with Sistonen et al3 who described in their adult EA population 21 BE at histology, with only 11 macroscopic BE. We are convinced that our seven patients without endoscopic suspicion of BE had indeed BE because four of them had cardial and fundic-type mucosa whereas two had exclusively fundic-type mucosa. Some can argue that fundic mucosa may indicate the presence of a hiatal hernia; multiple biopsies are therefore necessary, as in our study protocol, to minimize sampling error.4 Nevertheless, regarding only endoscopic metaplastic columnar epithelium confirmed histologically, we still have a high rate of BE in our population (37%). Our low incidence of intestinal metaplasia in repaired EA (2%) is in accordance with the literature (see Table 2)1: 0% to 3% in pediatric population and 0% to 15% in adult population.
A last point concerned hiatal hernia in EA who was suspected to be underdiagnosed in our study. The low prevalence of hiatal hernia we found in this study could be explained by the fact that more than 40% of our population had antireflux surgery previously.
Following Vergouwe et al conclusion, our study highlights the importance of accurate recognition of endoscopic anatomical landmarks in the diagnosis of BE, which was done here as correctly as possible following strict definition of GEJ. This allows us to be confident about the high prevalence of BE we found in adolescents with EA.