Endoscopic Signs in Celiac Disease: Reply

I am grateful to Dr. Brocchi and colleagues for their interest in our article and for their comments. They offer several constructive points for discussion. With regard to the issue of the classification of endoscopic signs, I think that it is necessary to clarify a few points. First of all, I agree with Brocchi and colleagues that histological damage usually occurs in advance of the macroscopic appearance of damage. In fact, we showed that a normal appearance of the second part of the duodenum may also conceal slight histological damage (Table 1 in [1]). Secondly, there are no existing classifications of endoscopic lesions in celiac disease. Before carrying out our study, we noted in our clinical practice that in patients with celiac disease, there was a certain correlation between the endoscopic appearance of the duodenum and histological damage. As the duodenal folds decreased and the appearance of the mucosa deteriorated, we observed a progressive worsening of the histological damage. We therefore conducted the study with arbitrary assessment of the grading of the increasing duodenal endoscopic damage, from a normal appearance of the duodenum to an absence of duodenal folds, with or without a “mosaic” or “micronodular” appearance of the mucosa. With regard to the sign defined as “granular mucosa” in the second part of the duodenum, this is an endoscopic sign that has not been reported in the published literature [2] [3] [4], in which duodenal mucosa shows a “granular” appearance, with normal presence of duodenal folds. This sign is very different from the sign defined as “micronodular mucosa”, in which it is clearly observed that the duodenal mucosa mimics submucosal nodules (see Figure 1 in [1]), as clearly described recently by Brocchi et al. [2]. In our experience, “granular mucosa” is an endoscopic feature which, although strictly related to several small-bowel diseases (such as celiac disease, Crohn's disease, and Giardia lamblia infestation), is more prevalent in celiac disease, just as “scalloping” of the duodenal folds is more prevalent in celiac disease, although it is not a specific marker of the disease [5].

I also agree with the comments made by Brocchi et al. regarding the sensitivity of endoscopic signs in diagnosing celiac disease. Certainly, an overall consideration of the endoscopic signs may be of greater clinical relevance than any single endoscopic sign in suggesting a diagnosis of celiac disease. However, our study offers another message about the single sign: in patients who have mild clinical complaints or are asymptomatic, it is not surprising to find slight endoscopic damage or normal features in the duodenum. I think that this is a very important point in clinical practice, since some endoscopists may be tempted not to obtain biopsy samples in the absence of severe endoscopic damage, with a definite risk of not making the correct diagnosis. Recognizing any of the features shown in our study should therefore always prompt the endoscopist to obtain duodenal biopsy specimens. In addition, an absence of any of these features does not exclude celiac disease, and biopsy samples should always be obtained when there is a suspicion that the disease may be present.

References

• 1 Tursi A, Brandimarte G, Giorgetti G M, Giglioblanco A. Endoscopic features of celiac disease in adults and their correlation with age, histological damage and clinical form of the disease.  Endoscopy. 2002;  34 787-792
• 2 Brocchi E, Tomassetti P, Misitano B. et al . Endoscopic markers in adult celiac disease.  Dig Liver Dis. 2002;  34 177-182
• 3 Dickey W, Hughes D. Prevalence of celiac disease and its endoscopic markers among patients having routine upper gastrointestinal endoscopy.  Am J Gastroenterol. 1999;  94 2182-2186
• 4 Bardella M T, Minoli G, Redaelli F. et al . Reevaluation of duodenal endoscopic markers in the diagnosis of celiac disease.  Gastrointest Endosc. 2000;  51 714-716
• 5 Shah V H, Rotterdam H, Kotler D P. et al . All that scallops is not celiac disease.  Gastrointest Endosc. 2000;  51 717-720