Anti‐Saccharomyces cerevisiae mannan antibodies in inflammatory bowel disease: comparison of different assays and correlation with clinical features

Abstract

Anti-Saccharomyces cerevisiae mannan antibodies have been proposed as a new serological marker associated with Crohn's disease. However, their clinical value is still unclear; furthermore, a standardization of anti-S. cerevisiae mannan measurements is lacking. In this study, we aimed to assess the correlation between anti-S. cerevisiae mannan detection and specific clinical features in Crohn's disease and ulcerative colitis. Moreover, we tested the concordance of four different anti-S. cerevisiae mannan assays. Serum samples from 196 patients with Crohn's disease, 197 patients with ulcerative colitis and 100 unrelated healthy controls were tested for anti-S. cerevisiae mannan with a standard enzyme-linked immunosorbent assay method (Lille) by one of the authors (VP). Subsequently, 60 randomly selected serum samples (27 Crohn's disease, 28 ulcerative colitis and five healthy controls) were tested for anti-S. cerevisiae mannan with three different commercial kits. With the Lille assay, anti-S. cerevisiae mannan were detected in 100 of 196 patients with Crohn's disease (51%; P < 0.0001 vs. controls), 32 of 197 patients with ulcerative colitis (16%; P < 0.02 vs. controls), and six of 100 controls (6%). No correlation between presence of anti-S. cerevisiae mannan and specific clinical features was found in both ulcerative colitis and Crohn's disease patients. The percentages of anti-S. cerevisiae mannan detected with four different assays ranged from 28 (Bouty) up to 43% (Inova), but these differences did not reach statistical significance. The concordance rate of anti-S. cerevisiae mannan detection in the four assays was very low (11 concordant results of 60 samples, 18.3%) (k = 0.15). No improvement of the concordance rate was obtained by modifying the suggested cut-off values (k = 0.20). In this study, we confirm that anti-S. cerevisiae mannan are significantly more frequent in Crohn's disease patients compared with ulcerative colitis patients (P < 0.0001) and controls. However, no correlation with clinical features was found in both ulcerative colitis and Crohn's disease. The low prevalence of anti-S. cerevisiae mannan, at least in our population, and the low concordance rate between different assays, makes the clinical role of this marker questionable.