P147. Serological and genetic markers as predictors of complicated disease in patients with Crohn's disease
E. Hoefkens1, M. Ferrante2, F. Princen3, V. Ballet2, I. Cleynen1, P. Rutgeerts2, S. Vermeire2, 1KU Leuven, Department of Clinical and Experimental Medicine, TARGID, Leuven, Belgium, 2University Hospitals Leuven, Department of Gastroenterology, Leuven, Belgium, 3Prometheus Laboratories Inc., Research and Development, San Diego, United States
Prediction of disease progression in Crohn's disease (CD) would help to individualize management and treatment. Some clinical parameters have been reported but the use of serology and genetics have been proposed to improve accuracy. We studied baseline clinical parameters, together with a serology panel and NOD2 genotypes in a prospective cohort of new CD patients and hypothesized that molecular markers could improve accuracy.
Out of 78 newly-diagnosed CD patients, 69 patients reached a follow-up of at least 6 months (median FU 26 months; IQR 15–46) and were included in this study. Serum and DNA was taken at diagnosis and analyzed for microbial peptides: anti-Saccharomyces cerevisiae (ASCA), anti-bacterial flagellins (CBir1, Fla2 and FLAX) and anti-outer membrane porin of E. Coli (OmpC) (Prometheus Labs) and were genotyped for NOD2-variants (p.Arg702Trp, p.Gly908Arg and p.Leu1007fsX1008). Associations between clinical parameters (age at diagnosis <40 years, ileal location, need for steroids, deep ulcers on endoscopy, smoking), immune responses to microbial peptides (both the presence and the quartile sum score), NOD2, and disease progression were evaluated.
During follow-up, 10 patients progressed from B1 (inflammatory disease) to B2 (stenosing) or B3 (penetrating) behaviour after a median of 8 months (5–20) and 13 needed abdominal resection after a median of 11 months (5–37). Kaplan–Meier indicated the presence of ASCA IgG as the only risk factor for progression of disease behaviour (LogRank p = 0.054).
A high serology load (quartile 3+4) was a risk factor (LogRank p = 0.023) for the need for abdominal resection. Individual risk factors for resection were a positive value for ASCA IgA (LogRank p = 0.015), CBir1 (p = 0.012), B2/B3 behaviour at diagnosis (p < 0.001) and a mutation in one of the 3 snps (p = 0.024). In Cox regression multivariate analysis, ASCA IgA [Odds ratio (OR) 6.6 (95% CI 1.1–39.4), p = 0.039] and B2/B3 behaviour at diagnosis [OR 7.8 (95% CI 1.7–36.8), p = 0.009] remained significant.
The proportion of patients needing abdominal surgery significantly increased with increasing quartile sum score for antibodies (linear-by-linear association p = 0.01) and none of the 19 patients in the lowest serology quartile sum score needed abdominal surgery.
In these newly-diagnosed CD patients, we demonstrate that the presence and magnitude of immune responses to different microbial antigens and the presence of mutations in NOD2 are associated with a higher risk for surgery.