Introduction Data suggests that an additional duodenal bulb (D1) biopsy may increase the diagnostic yield for coeliac disease (CD) by up to 10%. However no consensus exists on necessity of D1 biopsy. One reason may be that it is not clear if these patients with Ultra-Short Coeliac Disease (USCD) have the same phenotype or are at risk of the same consequences as conventional CD. We aimed to assess the clinical phenotypes of patients with USCD compared to those with conventional disease.

Method All patients attending a specialist CD endoscopy list were invited to take part. All patients had duodenal biopsy taken as routine. Patients had standard quadrantic biopsies taken from the second part of the duodenum (D2) and at least one biopsy taken from D1. Biopsies were analysed separately according to the Marsh classification system. Marsh 3 disease was required to diagnose CD. Patients had concurrent tissue tranglutaminase (tTG) and endomysial antibodies (EMA) and total IgA. All patients with VA were followed up in the CD specialist clinic where routine haematology, biochemistry, HLA typing and DXA scans were requested. Presenting symptoms and immunology were compared for all presentations. Haematology and biochemistry results were compared to a control group of patients that had CD excluded with normal serology and histology.

Results 1378 new presentations (62% female; mean age 50.3) underwent duodenal biopsy. 268 (19.4%) new diagnoses of CD were made (66% female; mean age 41.6). CD patients were significantly younger than controls (p < 0.0001). 25/268 (9.3%) of new CD patients had USCD. Univariate analysis showed fewer USCD patients had diarrhoea than conventional CD (3.8 vs 24.0%, P < 0.0001). Decision tree analysis to identify USCD showed the absence of diarrhoea was the single discriminating factor (Adj P = 0.018). Rates of osteoporosis (p = 0.78) and anaemia (p = 0.14) were equal. Ferritin deficiency (P = 0.007) and folate deficiency (P = 0.003) rates were higher in conventional CD than USCD and controls. On multivariate analysis, patients with USCD were younger than those with conventional CD 36.6 vs. 42.1 (AOR 0.97 (0.94–0.998) P = 0.03), had lower tTG titres (AOR 0.89 (0.81–0.98) P = 0.02) and had higher folate levels (AOR 1.17 (1.01–1.36) P = 0.03) compared to conventional disease.

Conclusion USCD appears to represent early disease with younger age and lower tTG titres and may represent a milder form of CD with lower rates of diarrhoea and folate deficiency. Long term follow up of patients with USCD is required to fully assess the clinical impact of diagnosis.

Disclosure of interest None Declared.