Prevalence and diagnostic role of antineutrophil cytoplasmic antibodies in inflammatory bowel disease.

BACKGROUND: Antineutrophil cytoplasmic antibodies (ANCA) are of proven diagnostic value in a variety of vasculitides, where they are also thought to play a pathogenic role. ANCA has also been detected in the serum of patients with idiopathic inflammatory bowel disease (IBD), both ulcerative colitis (UC) and Crohn's disease (CD), and primary sclerosing cholangitis (PSC) with or without concomitant IBD. Although the prevalence in PSC and UC is reported to be up to 85%, a much lower prevalence of around 10-20% has been reported in CD. AIM: To determine ANCA prevalence in a group of British patients with IBD and evaluate their use as a serological marker to distinguish between UC and CD. METHODS: A total of 99 UC-only patients (44 males, median age 50) and 41 CD patients (11 males, median age 47) were tested for ANCA using an alkaline phosphatase technique at a 1:5 serum dilution. Controls were other diarrhoeal diseases including 17 coeliac disease (4 males, median age 41), 23 irritable bowel syndrome (5 males, median age 42), 5 infectious colitis (2 male, median age 64) and 36 healthy volunteers (13 males, median age 43). RESULTS: ANCA was detected in 42/99 (42.4%) UC patients but in only 2/41 (5%) CD (P < 0.0001). All ANCA were perinuclear in distribution. No ANCA was detected in the control sera. The sensitivity of the test for the diagnosis of UC was 42% with a specificity of 98%. In patients with UC, no association was found between presence of ANCA and age, sex, disease extent, treatment or activity. However, ANCA-positive UC patients had longer median duration of disease (50 months vs. 29 months, P = 0.037). Both CD ANCA-positive patients had colonic involvement, but one also had ileal disease. Both had inactive disease and one was on mesalazine. CONCLUSIONS: ANCA is highly specific for UC and may be a helpful diagnostic test in distinguishing UC from CD and other diarrhoeal illnesses. Although ANCA positivity may reflect disease heterogeneity within UC, no association with clinical features or treatment of UC was demonstrated and it is therefore unlikely to play a pathogenic role. The correlation with disease duration needs further investigation.