DOP69 Longitudinal gut microbiome dynamics in relation to disease flares in Inflammatory Bowel Disease, pilot data from the IBD-Tracker study

Abstract

Abstract Background The pathogenetic mechanisms that lead to IBD flares are unknown and the absence of reliable predictive biomarkers results in a therapeutic approach focused on treating active flares rather than prevention. While IBD is known to be associated to changes in gut microbiota, the longitudinal dynamics of the microbiome in relation to disease flares and its biomarker potential remain poorly understood. We initiated the IBD -racker study which aims to include 100 patients with weekly fecal sampling along other biomaterials and clinical data. The aim is to elucidate patient-specific dynamics of the gut microbiome, including changes during IBD flares, strain replacements, and the potential for personalized prediction of flares based on microbiome. Here we present pilot microbiome data from the first 5 patients that completed the collection. Methods A comprehensive analysis of microbiome dynamics was conducted by collecting 52 weekly fecal samples, calprotectin measurements and clinical disease scores from five IBD patients. Metagenomic sequencing was employed on 248 samples, followed by biobakery4 tools for profiling of microbiome composition and function. Stability and dynamics of gut microbiota were evaluated through analysis of autocorrelation and stable states, while bacterial strains were profiled using markers and metagenomic assembly and binning based approaches. Finally, microbiome signatures associated with IBD flares were identified using multivariate generalized regression models correcting for patient-driven heterogeneity. Results The core gut microbiome of IBD patients (43 to 82 patient-specific highly prevalent species) is highly personalized and largely stable over time, regardless of flares (patient ID R2 > 0.4 in ADONIS of microbiome composition and function). The bacterial strains of these core species cluster by patients, indicating flares do not perturb them. Despite this stability of the core microbome, ~60% of identified bacterial taxa were present in <70% of weekly samples, indicating that a large part of the microbiome is dynamic. Differential abundance analysis revealed 32 bacterial species genomic bins (e.g. Clostridia pathobionts) and one pathway significantly associated with IBD flares (FDR < 0.1). Intriguingly, several species (e.g. Faecalibacterium sp. SGB15346, Clostridium inoocum) showed switches between high and low abundance stable states correlated with flares, and certain bacteria (e.g., Blautia wexlerae) underwent strain replacement during flares, which was not reversed post flare. Conclusion Our results indicate that IBD flares perturb a large part of the non-core gut microbiome, implying that personalized longitudinal monitoring of microbiota has potential for prediction and monitoring of IBD flares.