Introduction Historically, Brunner's glands in the bulb were thought to cause histological interpretation difficulties, however recent studies have demonstrated that this area maybe the only site to demonstrate villous atrophy (VA) and thus detect Coeliac Disease (CD). This study evaluates the diagnostic yield of taking duodenal bulb biopsies in coeliac patients compared with controls.

Methods Patients undergoing clinically indicated oesophogastrodudoenoscopy (OGD) were prospectively recruited from a single tertiary referral centre between November 2008 and December 2011. Indications for OGD included positive coeliac serology, family history of coeliac disease, diarrhoea, iron deficiency anaemia, abdominal pain and weight loss. All biopsies were graded using the Marsh criteria, with patients being assigned to one of three groups: Group 1 (CD: New Diagnosis), Group 2 (CD: Remission) and Group 3 (Controls).

Results 550 patients (360 female) with median age 51 (range 15–89 years) were prospectively recruited. 153 had newly diagnosed celiac disease, 91 established celiac disease, and 306 controls. New diagnosis celiac disease (9%, p<0.0001) and established celiac disease (14%, p<0.0001) were more likely than controls to have VA in the bulb alone (Abstract OC-049 table 1). Overall, when comparing the histological lesion of the bulb against the distal duodenum, 36/91 (40%) with established celiac disease (p<0.0001) and 35/153 (23%) newly diagnosed (p<0.0001) had a discrepancy in the severity of the lesion between the two sites compared with 22/306 (7%) controls. In all, 28/36 with established celiac disease and 24/35 newly diagnosed had the more severe lesion in the bulb. One patient in the control group had VA. This patient was HIV positive with positive tTG and negative EMA, however the HLA status was incompatible with coeliac disease.

Conclusion This is the largest prospective study evaluating the value of a duodenal bulb biopsy strategy. VA may only be present in the duodenal bulb. In this study 14/153 (9%) of newly diagnosed coeliac disease patients and 13/91 (14%) of CD remission patients demonstrated VA in the bulb alone. We suggest that endoscopists should consider taking a duodenal bulb biopsy in patients suspected of having coeliac disease and in reassessment cases.

Competing interests None declared.