We have read the letter of Drut and Rúa and agree that “patchy villous atrophy” is another way of saying “patchy areas of atrophy alternating with normal mucosa,” as shown in our article (1). In fact, we described children with celiac disease (13.7% at diagnosis and 42.9% after gluten challenge) in whom multiple duodenal biopsy specimens showed varying degrees of villous atrophy with some completely normal biopsy specimens (Tables 2 and 3 and Fig. 1). Only a few of our patients had isolated involvement of the duodenal bulb with normal distal biopsy specimens. As reported in our article, we have always taken four or more duodenal samples. The novelty of our observation is that in children, not only after gluten challenge, as reported by Vogelsang et al. in adults (2), but also at diagnosis, the bulb is sometimes the only area showing villous atrophy and crypt hyperplasia. We now routinely biopsy this area during endoscopy for possible CD. We cannot exclude the possibility of more distal jejunal mucosal lesions in patients with apparent isolated bulb involvement because we did not biopsy distal to the descending duodenum in our study. Drut and Rúa have a great experience with biopsy specimens of the distal duodenum and jejunum (more than 8,000 small bowel biopsy specimens; how many celiacs?) but have not biopsied the duodenal bulb mucosa. We also have had much experience with one- or two-hole peroral capsules for the diagnosis of CD (3,4) but no longer use this method. Endoscopy is faster, does not require x-ray exposure, permits direct visualization of the duodenum, and allows for collection of many biopsy samples, including the bulb. The use of endoscopy in the diagnosis of CD is now widely accepted and is considered better than the use of the Crosby capsule. Drut and Rúa affirm that “proper orientation and interpretation of endoscopically obtained material has been stressed in the literature.” This is certainly true, but by examining more than four samples, we feel that we obtain a correct diagnosis in 98% of patients (5) and also detect cases of patchy atrophy that might be missed by peroral capsules. We think that the presence of typical symptoms in patients with patchy lesions limited to the bulb mucosa is probably due to the existence of more distal areas of villous atrophy. The presence of antiendomysium and antitransglutaminase autoantibodies in patients with patchy villous atrophy supports the diagnosis of celiac disease. Finally, we find that the Marsh classification, modified by Oberhuber et al. (6), is useful in classifying the morphologic aspects of mucosa in patients with CD. Margherita Bonamico Paolo Mariani Enina Thanasi Mirella Fern Raffaella Nenna Claudio Tiberti Barbara Mora Maria Cristina Mazzilli Fabio Massimo Magliocca University of Rome “La Sapienza”, Rome, Italy