S1695 Can an IBD Polygenic Risk Score Function as an IBD Polygenic Prognosis Score? Modeling Genetic Biomarkers for Prognostic Insight in Established IBD

Introduction: With over 240 polymorphisms identified, the genetic architecture underlying risk of developing IBD is well characterized, but the prognostic potential of polygenic risk polymorphisms to predict future IBD disease trajectory remains unclear. We hypothesized that multiyear patterns of healthcare utilization, organized as mean annual healthcare charges ≥3 years, would serve as a proxy for multiyear IBD severity, facilitating identification of genetic polymorphisms associated with more severe disease trajectories. Methods: Models for genomic severity biomarker identification were generated from consented IBD patients, followed over multiple years, with multiyear mean healthcare charges as a severity phenotype. The genomic model for a Polygenic Prognosis Score repurposed a Polygenic Risk Score (PRS) based on leveraging 2 different datasets as base data. PLINK v1.90b7 and R version 4.4.0-4.4.1 packages were employed to plot single nucleotide polymorphisms (SNPs) per 4 high risk genes (ATG16L1, NOD2, IRGM, IL23) against mean healthcare charges, and downstream statistical analyses. Results: Among the 868 IBD patients (male 53%; Crohn’s disease 55%), IBD risk polymorphisms occurred in most patients, with IL23R polymorphisms being most common. We observe a scattered distribution of SNPs across the mean multiyear healthcare charge spectrum, supporting no association with charge severity and explaining lack of correlation between PRS and healthcare charges. We only find a weak, yet significant positive correlation (Spearman’s rank: ρ = 0.08, P < 0.05) between the burden of NOD2 SNPs and multiyear health charges as supported by the Kruskal-Wallis test P <0.05, Dunn’s post-hoc testing showing P < 0.05 between (1) 1-2 NOD2 SNPs vs 3-4 NOD2 SNPs and (2) 3-4 NOD2 SNPs vs No NOD2 SNPs. This finding is similarly seen in all 4 genes when combined, yet the group with at least 13 SNPs is not in any significant Dunn’s post-hoc pairwise comparison. Conclusion: An increased burden of high-risk alleles is not associated with increasing multiyear healthcare charges in established IBD patients. However, an increased burden of up to 3-4 NOD2 SNPs may be correlated with higher multiyear healthcare charges. This study confirms that genetics may offer only limited prognostic value for prediction of disease severity in established IBD.