Authors' Response

Abstract

To the Editor: We thank Dr Roggenbuck et al (1) for the interest in our recent article on pancreatic autoantibodies and autoantibodies against goblet cells in pediatric patients with inflammatory bowel disease (IBD). In our study, conducted only in pediatric patients with IBD, the combined prevalence of exocrine pancreatic antibody (PAB) and recombinant pancreatic antibody (rPAB) was significantly higher in pediatric patients with Crohn disease (CD) (34% and 35.9%) and in pediatric patients with ulcerative colitis (UC) (20.4% and 24.5%) compared with controls. In our study, the presence of PAB and rPAB was not associated with disease phenotype in either CD or UC (1). It should be noted that Bogdanos et al (2) enrolled a mixture of patients with IBD up to 87 years; only 18.3% of the CD patients were in A1 group (younger than 17 years) and indeed, there were no patients with UC younger than 17 years. Therefore, it is difficult to compare the data of the 2 studies; hence, age may affect location and behavior also. In our patients with CD, the presence of PAB and rPAB (34% and 35.9%) was comparable with the results of Bogdanos et al (30.2%). In UC, we found higher prevalence of rPAB and ASCA (24.5% and 26.5%) than Bogdanos et al (8.8% and 6.9%). These differences may be because of the younger age of our patients with IBD (median age 13.9 vs 36 years [CD] and 12.5 vs 47 years [UC]) and the higher percentage of sclerosing cholangitis in our patients with UC (20.9%) (3), although in other studies similar prevalence of PAB was noted in UC (22.7% (4) and 23.3% (5)). It should be noted that Bogdanos et al found relative low antibodies against Saccharomyces cerevisiae prevalence in their CD group (35.5%). Based on previous studies, there are conflicting results of association between PAB and CD phenotype (2,4,5). In agreement with the results published by Bogdanos et al (2), we observed numerically higher PAB and rPAB positivity in patients with ileocolonic CD (60% and 59.4%) than in patients with colonic (28.6% and 27%) and ileal disease (11.4% and 13.5%); however, the difference did not reach significance. Regarding CD behavior we could not demonstrate any relation, perhaps because of relatively higher proportion of patients with B1 phenotype and use only indirect immunofluorescence (IIF) method. We agree with Bogdanos et al that the combined use of IIF and enzyme-linked immunosorbent assay may improve the sensitivity of PAB. Further simultaneous multicentric studies conducted in pediatric and adult patients with IBD are necessary to investigate anti-glycoprotein 2 and anti-CUB and zona pellucida–like domain-containing protein 1 and their phenotype association separately using either IIF or enzyme-linked immunosorbent assay.