Authors: Matthias Jürgens Stephan Brand Rüdiger P Laubender Julia Seiderer Jürgen Glas Martin Wetzke Johanna Wagner Simone Pfennig Cornelia Tillack Florian Beigel Maria Weidinger Fabian Schnitzler Martin E Kreis Burkhard Göke Peter Lohse Karin Herrmann Thomas Ochsenkühn
Publish Date: 2010/04/29
Volume: 45, Issue: 7, Pages: 721-731
Abstract
Medical records of n = 1110 patients with inflammatory bowel diseases were screened for patients with fistulizing and stricturing CD Study inclusion required diagnosis of stenosis made within 6 months of diagnosing fistulas CDassociated NOD2 and IL23R variants were genotyped Similarly we prospectively investigated 42 patients presenting with fistulizing CDIn the retrospective study n = 333 CD patients fistulas were found in 145 435 patients and stenoses in 223 670 patients Concomitant stenosis was diagnosed in 125 patients with fistulas resulting in a positive predictive value PPV of 862 for fistulas predicting intestinal stenosis p = 553 × 10−11 OR 574 95 CI 322–1050 In logistic regression analysis presence of fistulas OR 451 95 CI 254–801 p = 268 × 10−7 and disease duration OR 109 95 CI 105–113 p = 319 × 10−6 were strongly associated with intestinal stenosis NOD2 genotype information but not IL23R status increased the PPV for the correct diagnosis of stenosis PPV = 899 All homozygous carriers 100 of NOD2 variants with fistulizing CD were diagnosed with stenosis 1007fs homozygotes were found more often among patients with fistulas and stenoses than in patients without stenoses and fistulas p = 000037 Similar results were found in the prospective analysis in which 833 of the patients with fistulizing CD had concomitant stenosis
Keywords: