terologist in the United States. Furthermore, after reading the above letter, I find very little factual discrepancy be- tween our viewpoints. The major points of my editorial were: 1) Antireflux surgery is more successful in the hands of experienced operators; 2) Complications of reflux disease are rare and unpreventable by any intervention; 3) Compli- cations and adverse events occur with antireflux surgery; 4) Laparoscopic antireflux surgery is no more efficacious than the open procedure; 5) The strongest data for surgery pertain to patients with ulcerative esophagitis; 6) The economic arguments favoring antireflux surgery are severely flawed; and 7) Antireflux surgery can be tremendously beneficial to carefully selected patients. I hope that these are the messages that the readership extracts from the editorial. Regarding controlled data di- rectly comparing laparoscopic antireflux surgery to modern medical therapy, I maintain that no relevant studies have been reported, and on that basis, I advise caution. Regarding the multitude of uncontrolled trials and observations dem- onstrating the efficacy of antireflux surgery, I only quote Sackett from his work on evidence-based medicine, “Ther- apeutic reports with controls tend to have no enthusiasm, and reports with enthusiasm tend to have no controls” (1). Peter J. Kahrilas, M.D. Department of Medicine Northwestern University Medical School Chicago, Illinois REFERENCE 1. Sackett DL. Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest 1989;95:2s– 4s. Reprint requests and correspondence: Peter J. Kahrilas, M.D., Northwestern University Medical School, Division of Gastroen- terology and Hepatology, Department of Medicine, Searle Build- ing 10th Floor, Suite 541, 303 East Chicago Avenue, Chicago, IL 60611. Received Oct. 22, 1999; accepted Jan. 24, 2000. Gastric Mucosal Lesions in Celiac Patients TO THE EDITOR: We read with great interest the article by Diamanti et al., who studied gastric mucosal lesions in patients with untreated and treated celiac disease (CD) by means of endoscopic biopsies and counts of gastric intra- epithelial lymphocytes (IEL). The authors concluded that celiac patients presented a similar prevalence of gastric mucosal changes compared with a control population of nonceliac patients, and that CD is rarely associated with lymphocytic gastritis and follicular gastritis. They also con- sidered that the previous studies about these gastric mucosal lesions in celiac patients were based on small numbers of cases or exhibited methodological defects (1). The data shown in this article are very surprising in many respects. Among the 104 celiac patients enrolled in this study, 24 subjects were on gluten free-diets (GFD) for many years. We have recently shown the disappearance of gastric mu- cosa-associated lymphoid-tissue (MALT) in celiac patients after 8 months of strict GFD (2); therefore, we believe that patients on GFD are not suitable to be enrolled in this kind of study. Diagnosis of CD was based on a combination of clinical symptoms: a flat intestinal mucosa, and a positive clinical response to GFD. Moreover, the diagnosis was confirmed by positivity of antigliadin (AGA; only IgA? both IgA and IgG?) or antiendomysium antibodies (AEA) but not in all patients. Recent studies indicate that the first diagnosis of CD is the result of a clinical picture that includes intestinal flattened mucosa, positivity of IgA-class AEA (which have a specificity near although 100%, not by AGA determina- tion alone) in the absence of serum IgA deficiency; never- theless, even if the ESPGAN criteria considered mandatory a control biopsy after GFD only in patients asymptomatic at first presentation such as first-degree relatives of celiac patients (3), a second small-bowel biopsy to check the recovery of intestinal lesions after GFD is considered nec- essary by most adult gastroenterologists. Performing all diagnostic tests is very important because an unexplained diarrhea with intestinal villous atrophy can be due to a parasitic infestation such as Giardia Lamblia, that is not evaluated in differential diagnosis, also considering that the high prevalence of Helicobacter pylori (H. pylori) infection, observed in patients of lower socioeconomic status suggests that the studied population is probably exposed to many infectious agents. The study protocol for gastric lesions seems inadequate, because few biopsy samples were ob- tained from the body and antrum, and not in all patients. Thus, we believe that gastric lesions have probably been underestimated. Diagnosis of follicular gastritis and detec- tion of lymphoid aggregates need a complete gastric histo- logical assessment by means of at least three biopsies in every gastric area. In our previous studies we performed, in all patients, three biopsies in the gastric body and three in gastric antrum, to detect gastric MALT (2, 4). The high prevalence of H. pylori infection in this population is very surprising. IgG anti-H.p. antibodies are not a specific marker of gastric infection of the bacteria in comparison with his- tological assessment. Patients without the histological pres- ence of the bacteria or negative 13 C-urea breath test, even in presence of IgG-class-specific antibodies, cannot be consid- ered H. pylori-positive. Moreover, it is very strange that, in presence of so high a prevalence of infection and in so wide a range of age in the studied population, lesions such as 1364 Letters to the Editor AJG – Vol. 95, No. 5, 2000