Article Text
PDF
Abstract
Introduction Gut microbiome ‘dysbiosis’ is an established feature of early IBD. Paediatric inception studies have linked increased abundance of oral pathobionts within the gut with more severe disease in UC patients. The paucity of adult treatment-naïve longitudinal microbiome data makes it challenging to link baseline microbiome profile to treatment outcomes. We present 16S rRNA and metagenomics data from baseline stool samples mapped to 2-year outcomes from a treatment naïve UC cohort diagnosed in our rapid-access IBD inception clinic.
Methods Stool was collected pre-diagnosis for sequencing (DNA Genotek OM-200 kits) and faecal calprotectin (FCP). Microbial DNA was extracted. 16S rRNA (targeting the V4 domain) and shotgun metagenomic sequencing were undertaken using the QIIME workflow (16sRNA), and the HUMAnN 3.0 workflow (metagenomics). Differential abundance analyses were undertaken in MaAsLin2 with a False discovery rate (FDR) <0.2 relevant and <0.05 highly significant. Outcome data was collected prospectively with escalation to immunomodulator or advanced therapy representing failure of conventional treatment.
Results 40 treatment naïve UC patients underwent 16S sequencing (mean 45000 reads), with metagenomic data in 33 (mean 3.6 million reads). A minimum two years’ follow up data was available.
10 patients had progressed beyond 5ASA treatment. Shannon alpha diversity was reduced in these patients compared to UC patients who did not require escalation, though this was only significant in the shotgun dataset (16S median 4.8 vs 4.4 p=0.11 Shotgun 2.7 vs 1.71 p=0.004).
In the 16S dataset, failure of conventional therapy was associated with significant enrichment of the genera Fusobacterium, Haemophilus, Veillonella, Escherichia.shigella and Aggregatibacter. Depletion of multiple genera, including Alistipes, Bifidobacterium, Oscillibacter and Fusicatenibacter was observed (figure 1).
Differential microbial abundance (Log2 fold change) in UC patients failing conventional therapy.
The metagenomic dataset was interrogated to assess for species and strain level changes. These did not reach FDR significance thresholds, but in accord with 16S rRNA data, reductions were seen in Oscillibacter ER4, Fusicatenibacter saccharivorans, Alistipes putredinis, Alistipes onderdonkii and Bifidobacterium longum.
Conclusion Reduced alpha diversity at UC onset was associated with the need to escalate from 5ASA treatment within 2 years. Enrichment of Haemophilus and Fusobacterium was seen in the therapy failure cohort, alongside depletion of key short chain fatty acid producers such as Bifidobacterium and Alistipes.