The Prevalence of Erosive Esophagitis in Children: A Population-based Study

Abstract

The Prevalence of Erosive Esophagitis in Children: A Populationbased Study Mark A. Gilger, Hashem B. El-Serag, Craig L. Dietrich Symptoms of gastroesophageal reflux disease (GERD) occur in 2-7% of children. The manifestations of GERD can be limited to symptoms (e.g. heartburn, regurgitation) or can be more complicated, such as erosive esophagitis (EE), esophageal strictures or Barrett’s esophagus. The prevalence of such GERD complications in children is unknown. PURPOSE: To determine the prevalence of endoscopic findings of EE in children.METHODS: All children aged 0 to 18 years who underwent EGD that was recorded in the Pediatric Endoscopic Database System Clinical Outcomes Research Initiative (PEDS-CORI) between 1999 and 2002 were included. Endoscopic reports that were incomplete, missing demographic features, indications for endoscopy or endoscopic findings were excluded. EE was defined either descriptively (erythema, erosion, ulcer) or by Los Angeles classification. Esophageal biopsy was not evaluated. RESULTS: 8038 children who underwent EGD fulfilled the inclusion and exclusion criteria. Of those, 1070 (13.3%) had EE. The mean age of children with children with EE was 12.6 (+/ 5.6) yrs vs. 10.0 (+/ 6.0) yrs in those without EE (p=<0.0001). There were 575 (53.7%) males with EE as compared with 3315 (46.7%) in those without EE (p=.0002). EE was found in 106 (9.9%) children aged 0-2 years, 73 (6.8%) children aged 3-5 years, 277 (25.9%) children aged 6-12 years, and 432 (40.4%) children aged 13-17 years. Hiatal hernia was found in 108 (10.1%) of children with EE, compared to 228 (3.2%) in those without EE (p=<0.0001). Mucosal abnormalities were identified in 390 (36.4%) of children with EE vs. 1327 (19.0%) of those without EE (p=<0.0001). The prevalence of Barrett’s esophagus, esophageal stricture, ulcer, prior surgery, nodules, foreign body/retained food and anatomical abnormalities were not significantly different between children with EE and those without. CONCLUSIONS: The prevalence of erosive esophagitis is slightly higher inmales and increases with age. In contrast to erosive esophagitis in adults, there were no significant variations according to race or ethnicity. Hiatal hernia is the only clinical observation that predicts the presence of EE. **470 The Presence of Helicobacter pylori in the Intestinal Mucosa is Closely Associated with Ulcerative Colitis Like Crohn’s Disease (UCLCD) Phenotype Dulciene M.M. Queiroz, Adriana G. Oliveira, Maria das Gracas P. Sanna, Andreia M.C. Rocha, Gifone A. Rocha Sr., Silvia B. Moura, Renato Dani, Maria de Lourdes A. Ferrari, Frederico P. Marinho, Liano S. Moreira, Lucia P.F. Castro Crohn’’s disease (CD) is a heterogeneous disorder with diverse clinical manifestations. Recently, fibrostenosing CD was associated with a serologic response to gut microbiota such as Pseudomonas fluorescens, ASCA and mutated allele of NOD2. We investigated whether the presence of Helicobacter in the intestinal mucosa was associated with specific disease phenotypes. We studied 117 patients: 74 controls and 43 with CD (22 with fibrostenosing, 16 with perforating and 12 UCLD). Fragments from the ileum and 6 different colon regions were obtained at colonoscopy. Helicobacter was investigated by culture, Helicobacter 16S rRNA, ureAnested PCR and characterized by 16S rRNA sequencing. Gastric HP infection was diagnosed by C-UBT and serology. Seven HP strains were isolated from the intestinal mucosa of 1 (2.2%) control and 6 (14.0%) CD patients (1with perforating and 5withUCLD) (p=0.02). SpecificHPDNAwas detected in the intestinal biopsies of 7 (9.5%) controls and 17 (39.5%) CD patients (p=0.0003).No otherHelicobacter species was identified.HPwasmore frequently isolated (47.1%v3.2%, p=0.004) or detected by PCR (58.3%v 22.6%, p=0.03) in the intestinal mucosa of patients with UCLD. If we considered only the group of gastric HP-positive patients, PCR was positive in 18.4% of the controls and in 100% of UCLD CD patients (p=0.00007). The positive results of culture (0.0% v 28.6%, p=0.009) and PCR (18.2% v 47.6%, p=0.04) were negatively associated with fibrostenosing disease. Positive culture was also inversely associated with ileal localization of CD (p=0.006). No association was seen between fistulating disease and positive culture (6.3% v 18.5%, p=0.4) and PCR (25.0%v 37.0%, p=0.4). Positive culture and PCRdid not associate with diarrhea in any subgroup. Gastric HP infection was similar in control (38-51.4%) and CD (21-48.7%) patients (p=1.0). However, the mean IgG titer against HP was lower (p=0.006) in CD patients (51.0U/ml) than in controls (239.7 U/ml). Furthermore, a significant difference was seen (p=0.04) between UCLD (103.8U/ml) and stenosing disease (24.5U/ml). We demonstrated associations between CD clinical phenotypes and the presence of viable HP or HP DNA in the intestinal mucosa that may indicate an abnormal higher number of HP in the intestinal mucosa or HP intestinal colonization in UCLD patients. Humoral immune response to HP infection was also influenced by CD phenotype, with a low reactivity to HP infection in stenosing disease phenotype.