Y-ECCO Literature Review: Raquel Oliveira

Methods and key findings

This prospective observational study included 296 outpatients with active IBD [203 Crohn’s Disease (CD) and 93 Ulcerative Colitis (UC)] who were initiating biological therapy. Considering 32 patients who switched biological therapies during the study timeframe, 328 biological interventions were assessed: 140 with an anti-TNF, 123 with vedolizumab (VDZ) and 65 with ustekinumab (UST). Stool features were assessed at baseline [faecal calprotectin (FC), moisture and microbial load and composition]. The main outcome was endoscopic remission, evaluated at week 14 for patients with UC and at week 24 for patients with CD. Before the endoscopic assessment, 177 patients (120 CD, 57 UC) provided another stool sample for reassessment.

Microbiome composition in active IBD is linked to disease location

At baseline, disease location, stratified between ileal IBD, colonic IBD (Montreal L2 CD and UC) and ileocolonic IBD, was associated with genus-level microbial community variation. Other covariates had an impact on microbiome diversification, such as faecal moisture content and patients’ inflammatory status [serum C-reactive protein (CRP) levels]. FC levels and previous biological exposure were not associated with microbiome composition variation. Accordingly, disease location, faecal moisture and serum CRP levels were the drivers of microbiota variation. Regarding microbiome composition, 65.9% of the patients harboured a Bacteriodes 2 (Bact2) microbiota at baseline. Bact2 is a microbiota enterotype that groups the key characteristics of a dysbiotic microbiota, including an overall decrease in microbial load, depletion of butyrate-producing bacteria and association with systemic (CRP) and gastrointestinal (FC) inflammation [3, 7]. An increased risk of hosting a Bact2 community was found for loose stools (higher faecal moisture content), systemic inflammation and ileal disease. Patients with a colonic disease presentation harboured the dysbiotic Bact2 enterotype less frequently.

Anti-TNF modulates patient microbiome composition and enterotype prevalence

Inter-individual differences explain most of the microbiome variation in any population. Beyond this, here, only the biological therapy administered was found to be associated with genus-level microbial community variation. Neither disease location nor treatment response had a significant contribution. A significant change in gut microbiota composition was seen only with anti-TNF therapy for all interventions, independent of endoscopic remission achievement. This translated into a minor enterotype switch, with 12.8% of carriers shifting away from the potentially dysbiotic Bact2 enterotype. Furthermore, 20 taxa with significant changes in quantitative abundances after anti-TNF treatment were identified: 19 gut-associated butyrogens (previously identified as essential for intestinal homeostasis [8]) increased and Veillonella (postulated to be a potential opportunist in inflammatory conditions [3]) decreased. Anti-TNF treatment was also associated with a significant increase in faecal microbial load, suggesting the post-anti-TNF gut environment to be more favourable for gut microbial community re-establishment. Importantly, microbial load remained significantly lower than in healthy controls. In summary, anti-TNF was the strongest microbiota-modulating agent. VDZ and UTK were mostly administered as second- or third-line treatments after anti-TNF failure.

Anti-TNF treatment results in higher therapeutic response rates among patients hosting a dysbiotic microbiota

In patients with a dysbiotic (Bact2) enterotype at baseline, a switch to a eubiotic enterotype was associated with a positive endoscopic outcome with anti-TNF and VDZ. However, remission rates for Bact2 carriers were higher with anti-TNF therapy than with VDZ, even when only first-line interventions were considered. In fact, a Bact2 carrier status, as well as higher FC levels at baseline, lowered the chances of achieving remission with VDZ.

Assessment of Bact2 and associated stool features enables prediction of response.

Finally, the authors developed a model to predict treatment response to biological therapies, using a combination of clinical data (age, gender, body mass index, smoking, haemoglobin, albumin, CRP, IBD diagnosis, disease location and duration, biological history), faecal characteristics (microbial load, moisture and FC) and Bact2 carrier status. This model achieved a balanced sensitivity (67.5%) and specificity (67.6%) and an accuracy of 73.9%. Interestingly, 53.29% of the non-responders were predicted by this model to remain non-responders with any alternative therapy, suggesting that a subset of patients would be refractory to any kind of biological intervention.

Discussion

This study from Caenepeel et al. reveals a potential for using stool characteristics to predict therapeutic outcomes and personalise IBD therapy. Importantly, the analyses conducted in this study do not allow establishment of the chain of events between restoration of eubiosis and endoscopic remission. A longer follow-up period would be necessary to assess the persistence of eubiosis and its association with sustained remission. Additionally, given the prospective, observational nature of this study, VDZ and UST were mostly administered as a second- or third-line treatment, after anti-TNF failure, which could have influenced the evaluation of the microbiota-modulating effect of these therapies.

Conclusion

In conclusion, the authors show a link between IBD location and microbiota dysbiosis and demonstrate that anti-TNF therapies have favourable microbiota-modulating effects. The novel and reasonable model that they developed for prediction of treatment response to biological therapies, which included faecal characteristics, could be helpful in informing personalised therapy choice.

References

1. Ni J, Wu GD, Albenberg L, Tomov VT. Gut microbiota and IBD: Causation or correlation? Nat Rev Gastroenterol Hepatol 2017;14:573–84.
2. Pascal V, Pozuelo M, Borruel N, et al. A microbial signature for Crohn's disease. Gut 2017;66:813–22.
3. Vieira-Silva S, Sabino J, Valles-Colomer M, et al. Quantitative microbiome profiling disentangles inflammation- and bile duct obstruction-associated microbiota alterations across PSC/IBD diagnoses. Nat Microbiol 2019;4:1826–31.
4. Joossens M, Huys G, Cnockaert M, et al. Dysbiosis of the faecal microbiota in patients with Crohn's disease and their unaffected relatives. Gut 2011;60:631–7.
5. Kolho KL, Korpela K, Jaakkola T, et al. Fecal microbiota in pediatric inflammatory bowel disease and its relation to inflammation. Am J Gastroenterol 2015;110:921–30.
6. Tedjo DI, Smolinska A, Savelkoul PH, et al. The fecal microbiota as a biomarker for disease activity in Crohn's disease. Sci Rep 2016;6:35216.
7. Vandeputte D, Kathagen G, D'Hoe K, et al. Quantitative microbiome profiling links gut community variation to microbial load. Nature 2017;551:507–11.
8. Sokol H, Pigneur B, Watterlot L, et al. Faecalibacterium prausnitzii is an anti-inflammatory commensal bacterium identified by gut microbiota analysis of Crohn disease patients. Proc Natl Acad Sci U S A 2008;105:16731–6.

Raquel Oliveira - Short Biography

Raquel Oliveira is currently undertaking a Residency Program in Gastroenterology, Hepatology and Endoscopy at the University Hospital Centre of the Algarve, in Portugal. She has worked closely with the IBD clinic and undertaken clinical research since early in the Program and will soon start her formal training in bowel ultrasonography with IBUS. Raquel has also been collaborating with the Portuguese IBD Study Group and ECCO in both national and international projects.