Background: Imbalance in the faecal microbiota with a reduction in biodiversity; dysbiosis, has been identified in inflammatory bowel disease (IBD). Our aim was to study and compare the faecal microbiota in paediatric patients with newly diagnosed untreated IBD with the microbiota of healthy children and paediatric patients with gastrointestinal symptoms but no IBD. We also aimed at studying the microbiota related to IBD subgroups and treatment.
Methods: Faecal samples were collected from 235 children and adolescents. Eighty had Crohn's disease (CD), 27 ulcerative colitis (UC) and 3 IBD unclassified, 50 were non-IBD patients and 75 were healthy children between two and 18 years. The microbiota was analysed using a 16s rRNA DNA based test with the GA-map technology, measuring probe signal intensity (PSI) of 54 DNA probes targeting 300 bacteria on different taxonomic levels. Using non-parametric methods, we selected six probes where the PSI was lower in IBD compared to non-IBD patients. For each of these six probes, IBD patients were given 1 point if their PSI was lower than the median PSI value of non-IBD patients. The points were summarized as a Score ranging from 0–6 points. Logistic regression was used to model possible associations between this Score and risk of having IBD.
Results: Most bacterial PSI were reduced in IBD and non-IBD patients (p<0.001) compared to healthy controls. IBD patients had reduced abundance of Firmicutes (Eubacterium, p=0.006; Holdemanella, p=0.038), Tenericutes and Bacteroidetes (Parabacteroidetes p=0.02), p=0.002, and Bifidobacterium, p=0.02, compared to the non-IBD patients. CD patients had lower abundance of Mycoplasma (p=0.045) than UC patients. IBD patients with extensive disease (L3/E3) had more Clostridiales (Ruminococcus gnavus), p=0.02, and CD patients with L3 had more Proteobacteria, p=0.04, than patients with limited disease. IBD patients who later received TNF blockers, 64/110, had lower diversity at baseline for Firmicutes, Tenericutes (Mycoplasma, p=0.009), and Bacteroidetes, p=0.015, compared to IBD patients who were treated with conventional medications, 46/110. Patients who reached 3 or more points using the Score were 2.2 times more likely to have IBD compared to non-IBD (OR=2.1, 95% CI 1.1–4.5, p=0.027).
Conclusions: Microbiota profiles may be of value for stratification of paediatric IBD into diagnostic and prognostic subgroups. A severe dysbiotic microbiota profile in newly diagnosed IBD is associated with a severe phenotype with more extensive disease and subsequent need of TNF blocker treatment.