HOW COMMON ARE VILLOUS ATROPHY AND CELIAC DISEASE? SIX YEAR ANALYSIS OF DUODENAL BIOPIES IN CHILDREN
Molly A O'Gorman, Brian Stanga, Shawn Reese, Charles Hoff, Linda S Book, Pediatric Gastroenterology, Primary Children's Medical Center/University of Utah Medical Center, Salt Lake City, UT
Children with Celiac Disease (CD) may have no, atypical, or minimal gastrointestinal symptoms. The Marsh score is the established histological grading system for CD. We hypothesize that unsuspected CD occurs in children who have symptoms requiring evaluation with esophagogastroduodenscopy (EGD) and that Marsh III (total/subtotal villous atrophy) abnormalities are highly predictive of CD.
To deteremine the prevalence of duodenal histologic abnormalities consistent with CD in children undergoing routine EGD, to identify symptoms associated with Marsh II and III lesions, and to determine how often a diagnosis of CD was made.
Duodenal biospy reports representing 1690 procedures were reviewed between 1995–2000. We determined the number of abnormal duodenal biopsies for this six year period and determined the Marsh score on all abnormal biopies. We then did 1–5 year follow-up evaluation for clinical symptoms, growth and EMA testing.
Histologic abnormalities occurred in 593/1690 (35.0%) of all duodenal biopsies. Marsh II changes occurred in 2/1690 (0.1%) biopies and Marsh III, 136/1690 (8.0%) biopies. The remaining 455/1690 (27.0%) biopies showed other abnormalities such as eosinophilia, inflammatory changes, lymphangiectasia, and apoptosis. Review of 108 available charts showed that 25/108 (24.0%) patients had a confirmed diagnosis of CD, and fulfilled ESPGHAN criteria. In 17 patients without a diagnosis of CD, who were available for follow-up, none were EMA positive but all complained of continued gastrointestinal symptoms 1 to 5 years later. The most common symptoms were diarrhea, abdominal pain, nausea and vomiting; none of the patients had failure to thrive.
Histologic abnormalities of the duodenum are common in children undergoing esophagogastroduodenscopy for gastrointestinal symptoms. We found that 24% of abnormal biopies and 8% of all biopies were consistent with a diagnosis of CD. However, serologic (EMA) testing failed to confirm a CD diagnosis, suggesting either that these patients have EMA negative CD or that Marsh III histological abnormalities are more nonspecific than generally considered. Goals for the future would be continued EMA testing, HLA typing and repeat esophagogastroduodenscopy in these patients.
ALL CHILDREN WITH TYPE I DIABETES SHOULD BE SCREENED FOR CELIAC DISEASE
Michelle Pietzak, Adam Wolf, Christine Rongey, Francine Kaufman, Lynda Fisher, Pisit Pitukcheewanont, Debra Devoe, Jody Krantz, Irene Berti, Tanya Gerarduzzi, Mary Thorpe, Debbie Kryszak, Karoly Horvath, Alessio Fasano, Pediatrics, Childrens Hospital Los Angeles, University of Southern California Keck School of Medicine, Los Angeles, CA; Pediatrics, University of Maryland, Baltimore, MD
Celiac disease (CD) is an immune-mediated enteropathy which occurs in genetically susceptible individuals who ingest gluten. Children with other autoimmune diseases, such as type I diabetes (IDDM), are at higher risk to develop CD. The perception in the U.S. is that CD is rare, especially among non-caucasians, despite the presence of the genes and dietary grains.
to screen children with IDDM for CD in an ethnically diverse population.
Subjects completed a medical questionnaire. Serum antigliadin IgG and IgA (AGA), and human anti-tissue transglutaminase IgA (htTG) were measured by ELISA. Anti-endomysium IgA (AEA) was detected by indirect immunofluorescence on human umbilical cord. Total IgA was measured on those only AGA IgG+. Small bowel biopsy was recommended to patients htTG+, AEA+, AGA IgG+ and IgA deficient, and AGA IgG+/IgA+ under the age of 3 years.
23 of 189 subjects (12%) had postive serology. The majority of the seropositives were Caucasian (78%), followed by Latino (13%), Persian and mixed ethnicities (9%). The most common reported symptoms in the seropositives were abdominal pain and joint pain (22% each), bloating, diarrhea and fatigue (13% each) and constipation (9%). 65% had no GI symptoms but had evidence of subclinical malabsorption (iron and fat-soluble vitamin deficiencies and nocturnal hypoglycemia). Family histories in the seropositives were significant for thyroid disease (35%), arthritis and IDDM (22% each) and osteoporosis, stomach and duodenal ulcers and irritable bowel syndrome (13% each). None had a family history of CD. Of the 12 biopsied, 10 were consistent with CD and two were indeterminate. CD3 staining for intraepithelial lymphocytes in one indeterminate biopsy reclassified it as Marsh Grade I.
CD is very common in children with IDDM. These patients should be screened for CD regardless of their ethnicity, absence of gastrointestinal symptoms or lack of a family history of CD.
IS TISSUE TRANSGLUTAMINASE ANTIBODY PREDICTIVE OF CELIAC DISEASE IN CHILDREN WITH NORMAL SMALL BOWEL BIOPSIES?
Mahmoud Sabri, Seema Khan, Division of Gastroenterology, Children's Hospital of Pittsburgh, University of Pittsburgh, School of Medicine, Pittsburgh, PA
The high sensitivity (92%) and specificity (98%) of tissue transglutaminase antibody (t-TG Ab) for diagnosis of celiac disease (CD) prompted us to evaluate its diagnostic value in children with normal small bowel morphology compared to children with CD.
We reviewed retrospectively clinical data and histology pertaining to children referred for evaluation of positive t-TG Ab. We divided the patients into group (gp) A: normal small intestinal mucosal biopsies, and gp B: biopsies compatible with CD. Tissue transglutaminase IgA antibodies were determined by using the ELISA and a commercially available kit using guinea pig derived liver t-TG (QUANTA Lite™ t-TG).
Data related to 15 children were reviewed. Eight (3 females) of 15 (53%) children had no abnormalities on small bowel biopsies and comprised gp A and their median age at presentation was 11 y (4.5–14). Seven of the eight (88%) children presented with constipation, vomiting, diarrhea, dyspepsia, and/or weight loss. The median duration of symptoms prior to evaluation was 1.37 y in gp A versus 0.06 y in gp B. The seven children (4 females) in gp B had a median age at presentation of 7 y (5–17), 3 presented with constipation, vomiting, diarrhea, and weight loss. The five asymptomatic children, in this series, were screened because of associated diagnosis of Down syndrome (1) in gp A, IDDM (1), and a family history of CD (3) in gp B. The mean t-TG value was 50 IU/L (33–117) in gp A compared to 117 IU/L (31–200) in gp B (p=NS). All patients in gp B had either partial villous atrophy or blunted villi and intraepithelial lymphocytes on small bowel biopsies consistent with CD. In follow up, the 7 symptomatic patients in group A were diagnosed with functional constipation (3), Crohn's disease (1), cyclic vomiting and food allergy (1), and (2) had spontaneous resolution of symptoms.
We speculate that positive t-TG Ab in children with normal small bowel biopsies may be either false positive or represent latent CD. An endoscopic re-evaluation with biopsies should be considered for those with persistently elevated t- TG Ab.