Separate Duodenal Bulb Biopsies Increase Detection of Abnormal Histology but Do Not Increase Diagnostic Yield of Celiac Disease in All-comers to Endoscopy

Abstract

Purpose: Celiac disease (CD) detection is increased when separate duodenal bulb and distal duodenal samples are obtained in a high pre-test probability population (i.e., patients with positive serology or established CD). This study examines whether CD detection is increased by sampling the duodenal bulb in a low pre-test probability population. Methods: Patients who underwent separate sampling of the bulb during endoscopy at 3 Mayo Clinic sites (Rochester, MN; Scottsdale, AZ; Jacksonville, FL) between January 1, 2011 and December 31, 2011 were included. Records were reviewed for age, gender, pathology, serology, biopsy indication, and adherence to a gluten free diet (GFD) for 4 weeks prior to biopsy. Histology consistent with CD included Marsh stages 1, 2, and 3. Statistical analysis was performed using Χ2 test. P values were two-sided and considered significant if < 0.05. Results: Final analysis included 737 patients - 464 (63%) female, mean age 50 years (range 2-88). The most common biopsy indications were abdominal pain (47%), diarrhea (36%), nausea (23%), weight loss (17%) and anemia (16%). CD or Dermatitis Herpetiformis was known in 33 (4%) and 72 (10%) were on a GFD. Celiac serology was checked in 331 patients; 32 (9%) had at least one positive celiac serology -- 14/330 (4.2%) for TTG IgA, 13/128 (10%) for TTG IgG, 3/98 (3%) for Endomysial Antibody IgA, 10/91 (11%) for Deamidated gliadin peptide (DGP) IgA, and 6/90 (6.7%) for DGP IgG. A total of 15 patients (2%) in this cohort were newly diagnosed with CD. Villous atrophy (VA) consistent with CD was detected in 22 patients (3%) -- 14 (1.8%) new CD diagnosis and 8 (1%) ongoing CD. In 3 of these patients (0.4%), separate bulb sampling detected a new CD diagnosis (n=1) and ongoing CD activity (n=2) which would have been missed if only the distal duodenum was sampled. Most patients (n = 447, 61%) had normal pathology in the duodenum (Tables 1 and 2). Abnormal histology was detected in 290 patients (39%) and was more commonly detected in the bulb - 262 (36% [95%CI 32-39]) vs 119 (16% (95% CI 14-19%)), p < 0.0001. In all patients with abnormal bulb histology (n = 262), the most frequent abnormality was chronic peptic duodenitis (119, 60%). In patients with a normal distal duodenum, 171 (23%) had abnormal histology isolated to the bulb.Table 1Table 2Conclusion: Abnormal duodenal histology is not uncommon and is more frequently detected in the bulb; the most common histology being chronic peptic duodenitis. In a low pre-test probability patient population (i.e., <10% with positive serology or known CD), individual analysis of the bulb does not sufficiently increase detection of CD (0.4% diagnostic yield) and would not warrant the cost associated with this additional anatomic site biopsy.