P701 The contribution of genetics in differentiating inflammatory bowel disease type unclassified
T. Billiet*1, I. Cleynen1, V. Ballet2, K. Claes1, F. Princen3, S. Singh3, M. Ferrante2, G. Van Assche2, S. Vermeire2
1Department of Clinical and Experimental Medicine, KU Leuven, Translational Research in GastroIntestinal Disorders, Leuven, Belgium, 2University Hospitals Leuven, Gastroenterology - Translational Research Center for Gastrointestinal Disorders (TARGID), Leuven, Belgium, 3Prometheus Laboratories, Department of Research and Development, San Diego, United States
A definitive diagnosis of Crohn's disease (CD) or ulcerative colitis (UC) in patients who initially present with inflammatory bowel disease type unclassified (IBDU) remains challenging. Most often, a combination of clinical (presence of rectal sparing without local therapy; ileal disease/backwash ileitis; perianal abscess; segmental colitis; stenotic disease), histopathological and sometimes serology is used in clinic. We investigated if the combination of clinic, pathology, serology and genetics would improve differential diagnosis in these patients.
We retrospectively identified 60 patients diagnosed with IBDU. On the basis of histopathology, 21 of these were later reclassified as CD, 22 as UC and 17 remained IBDU at the end of follow-up (with a median follow-up time of 12.6 yrs). For each patient a clinical score ranging from 0-5 (sum of the clinical factors mentioned above) was calculated and a serum samples were analyzed for pANCA and several antimicrobial antibodies (ASCA IgA, ASCA IgG, CBir1, OmpC, Fla2 and FlaX (Prometheus Laboratories Inc.)). All antimicrobial antibodies were divided into quartiles and quartile sum scores (QSS) were calculated for each patient. We also genotyped patients for the 163 IBD loci through immunochip and calculated a genetic risk score (GRS) for specific CD vs UC loci (higher values more indicative for CD and lower for UC). All markers were compared between the different groups.
The median time (IQR) to definitive diagnosis was longer in the CD patients (9.6 (4.9-12.9) yrs) than in the UC patients (2.1 (0.8-8.8) yrs, P=0.003). Both the clinical score and QSS could clearly distinguish the CD group from the UC and IBDU group (P=0.03 and P=0.04, Kruskall-Wallis test) but not the UC from the IBDU group. The GRS and pANCA status did not differ between groups (P>0.45). Logistic regression identified the clinical score and QSS to be independent predictors for diagnosing CD (P<0.01) with OR (95% CI) of 2.7 (1.3-5.7) and 1.3 (1.1-1.5) respectively and the accuracy of this prediction increased (AUC of 0.7 to 0.78 in ROC) when both were combined. A similar approach for the UC patients could only identify the clinical score as a predictor (P=0.01) with an OR of 0.4 (0.2-0.8).
In patients with IBD-unclassified, a combination of clinical factors and antimicrobial antibodies is superior for determining evolution to CD. The current validated genetic risk panel of 163 susceptibility loci does not have an added value in making this distinction. This is probably due to the fact that there is a significant overlap between CD and UC-risk alleles and the fact that only very few genes are specific for CD or UC.