Predicting Complicated Crohn's Disease and Surgery

Phenotypes, Genetics, Serology and Psychological Characteristics of a Population-Based Cohort

J. D. Ryan; M. S. Silverberg; W. Xu; L. A. Graff; L. E. Targownik; J. R. Walker; R. Carr; I. Clara; N. Miller; L. Rogala; C. N. Bernstein


Aliment Pharmacol Ther. 2013;38(3):274-283. 

In This Article


Participant demographic and psychological characteristics and Montreal Classification are reported in Table 1 and Table 2 for all 182 members of the Manitoba IBD Cohort with CD. The baseline assessment for disease phenotype includes the first 3 years from disease diagnosis (T1). The baseline demographic and psychological assessment took place at Cohort Study enrolment, at a median of 4.3 years after diagnosis, and the 5-year follow-up measurement was done at a median of 9.3 years following diagnosis (T2). Sixty-two per cent of participants were females. Seventy-eight per cent had ileal involvement at study enrolment (either isolated or combined with colon involvement). Fifty-five per cent of study enrolees had stricturing and/or penetrating disease at enrolment, and 65% had developed complications by the final assessment. By a median 9 years post diagnosis, 42% had undergone surgery. Persons diagnosed under the age of 40 were more likely to have perianal disease (P = 0.004), but age at diagnosis was not associated otherwise with phenotypic expression, or development of complicated CD. A history of smoking was associated with a lower incidence of complicated CD and perianal disease (P = 0.05), but was not associated with the need for surgery.

Relationship of Psychological Factors With Disease Characteristics

Most subjects reported some level of adverse childhood experience (73%) and high childhood adversity impact tended to be more common in those who underwent surgery (P = 0.06). Subjects who underwent surgery were also less likely to report being married or in common-law relationships (P = 0.02). No other psychological variables were associated with disease location, complicated CD or surgery. High adherence to medications was modest (57%) and was not associated with the outcomes of interest. Subjects with a family history of IBD (46%) did not have any significant differences in phenotype or rate of surgery.

Relationship of Serological and Genetic Markers and Disease Location

Persons with disease confined to the ileum (L1) vs. those with colonic (L2) and ileocolonic (L3) combined were more likely to be CBir1 positive at T1 (P = 0.02) and ASCA IgG positive at T2 (P = 0.04) (Table S1). No genetic markers were associated with disease location, although the presence of NOD2 rs2076756 genotype tended to associate with isolated ileal location (odds ratio (OR) 1.54; 95% CI: 0.99–2.42, P = 0.06) (Table S2). Multivariate analysis indicated that only ASCA IgG at T2 remained a significant predictor of CD location and specifically disease of the terminal ileum (L1) (OR 2.2; 95% CI: 1.07–4.54, P = 0.03) (Table 3). No further correlations between psychological, phenotypic, serological and genetic factors were present in our population. Upper GI (L4) disease and extra-intestinal manifestations were uncommon in our cohort.

Predictors of Complicated Disease

Disease location did not predict the development of complications. NOD2 rs2066847 was the only SNP associated with complicated behaviour (OR 3.25; 95% CI: 1.32–8.01, P = 0.01). Subjects at T1 and T2 who were ASCA IgA (P = 0.01; P = 0.002 respectively) and ASCA IgG positive (P = 0.0001, P = 0.0006 respectively) were more likely to have stricturing or penetrating disease. Multivariate analysis of complicated behaviour (stricturing/penetrating disease) indicated that ASCA IgG remained the only significant predictor (OR = 3.01; 95% CI:1.28–7.09; P = 0.01), while NOD2 rs2066847 trended towards an association (OR 3.08; 95% CI: 0.87–10.89, P = 0.08). Sex, age at diagnosis, perianal disease behaviour, family history, adherence to medications and all other psychological traits were not correlated with complicated disease behaviour. Conversely, those subjects with a positive pANCA at T2 (P = 0.04) and with a history of smoking (P = 0.05) appeared to be protected from stricturing/penetrating complications; however, these associations were not significant when included in the multivariate analysis. Antibody testing using a quartile sum score analysis[42] was not predictive of any outcomes (data not shown). The positive predictive value ASCA IgG for complicated disease is 61.4%, and the negative predictive value of a negative ASCA IgG is 66.2%.

Predictors of Surgery

Stricturing/penetrating behaviour was significantly associated with the need for surgery (P < 0.0001). Perianal disease activity was marginally associated with surgery (P = 0.06). NOD2 rs2066847 was the only SNP predictive of surgery (OR 2.39; 95% CI: 1.06–5.38) and was weakly associated with early surgery (OR 1.96, 95% CI: 0.95–4.07, P = 0.07). Surgery was associated with ASCA IgA at T1 (P = 0.02) and T2 (P = 0.03) and ASCA IgG at T1 (P = 0.002) and at T2 (P = 0.006). Although stricturing/penetrating disease was predictive of surgery, we excluded complicated disease from our multivariate analysis of surgery as it was one of our independent outcomes. On multivariate analysis of predictors of surgery, there was an association with ASCA IgG (OR 2.66, 95% CI, 1.40–5.06, P = 0.003). Sex, age at diagnosis, disease location, family history, smoking, adherence to medications and all other psychological variables were not significant in either model for early surgery or surgery overall.