CROHN’S disease (CD) causes complications requiring surgery in nearly 50% of affected patients. As there is currently no cure, early detection and interventions are essential. A new study introduces the Crohn’s and Colitis Canada-Genetic, Environmental, Microbial (GEM) Integrative Risk Score (IRS), which aims to predict the risk of CD in healthy first-degree relatives (FDR) of individuals with the disease.
The study used data from the GEM study, which included 3,270 healthy FDRs from North America and Israel. After quality control, 2,619 participants were included, with 61 (2.3%) developing CD over a median follow-up of 6.8 years.
The GEM-IRS model, developed using random survival forest techniques, combined demographic data, physiological biomarkers, and faecal microbiome data. It achieved a Harrell’s concordance index of 0.951 (95% CI: 0.925–0.976) in the North American training cohort. Upon validation in the pooled testing cohort, the model maintained strong performance, with a concordance index of 0.789 (95% CI: 0.713–0.865). The GEM-IRS outperformed other models, including gradient boosting survival (0.777) and penalised Cox models (0.717).
The cumulative incidence of CD was significantly higher in those with scores in the highest quartile of the GEM-IRS, with a hazard ratio of 6.42 (95% CI: 3.10–13.30) compared to the lower quartiles. The unadjusted hazard ratio per standard deviation increase in the score was 2.69 (95% CI: 2.06–3.53), consistent even after adjusting for prespecified covariates.
The model performed well even in subgroups with low baseline faecal calprotectin (FCP<50 μg/g or FCP<100 μg/g) and low lactulose-mannitol ratio (LMR≤0.025), indicating its predictive ability in asymptomatic individuals. FCP and LMR were the most significant contributors, followed by microbial composition and functional pathways. Specific microbial genera linked to increased CD risk included Holdemania, Gemella, and Ruminococcus torques.
The GEM-IRS is the first model to integrate biomarkers of inflammation, gut barrier function, and microbiome data to predict CD risk. It demonstrated strong predictive performance across diverse cohorts, suggesting its potential for early identification of high-risk individuals. Further validation is needed, but this model could guide interventions to prevent CD onset.
Ada Enesco, EMJ
Reference
Lee SH et al; Crohn’s and Colitis Canada-Genetic, Environmental, Microbial (CCC-GEM) Project Research Consortium. Development and validation of an integrative risk score for future risk of Crohn’s disease in healthy first-degree relatives: a multicenter prospective cohort study. Gastroenterology. 2025;168(1):150-153.
