Y-ECCO Literature Review: Susanna Meade
Prominence of ileal mucosa-associated microbiota to predict postoperative endoscopic recurrence in Crohn’s disease
Sokol H, Brot L, Stefanescu C, et al.
© Susanna Meade
The aetiopathogenesis of CD is multifactorial but includes the interaction between the microbiome and the host’s immune response. Up to 80% of patients with Crohn’s Disease (CD) require surgery during their lifetime and many factors are associated with postoperative recurrence (POR). Differential abundance of bacterial species is seen in patients with IBD compared with healthy individuals and several studies have suggested an association between microbiota composition and CD recurrence [1–3]. Altered mucosal gene expression and abundance of specific microbiota are associated with, and specific to, ileal CD .
Given the above, Sokol et al aimed to investigate the role of ileal mucosa-associated microbiota in postoperative endoscopic recurrence (ER). They conducted a prospective multicentre cohort study alongside the REMIND group study investigating predictive factors for POR . Patients >18 years old requiring intestinal resection for active ileal or Ileocolonic CD between 2010 and 2017 were included. Patients received postoperative therapy according to a pre-set algorithm accounting for established risk factors for POR. Colonoscopy with Rutgeerts scoring (Ri) was performed at six months (Ri ≥2 indicating POR).
Two ileal biopsies were taken at resection (M0) and at colonoscopy (M6). Microbial DNA was extracted from the samples and then analysed by 16S sequencing. The 16S rRNA gene encodes a ribosomal subunit universal to most bacteria, within which are variable regions specific to each species, enabling bacterial classification . A logistic regression model was used to detect factors associated with POR. The LEfSE pipeline, a validated metagenomics-based statistical method, was used to compare microbiota composition at the two different time points.
A total of 287 patients had biopsies taken at either time point and after quality control 201 patients were included. Eighty-seven patients had data from M0 and M6.
Among these 87 patients there was a statistically significant difference in beta diversity, but not alpha diversity, between samples taken at M0 and M6. However, a difference in alpha diversity was noted when patients were stratified according to their Rutgeerts score at M6 (Ri <2 vs ≥2). Differential representation within each phylum was noted. Gammaproteobacteria decreased whilst Alphaprotebacteria increased, and similarly, Bacilli decreased but Clostridiales increased (both Firmicutes). Clostridiales are associated with healthy mucosa and this increase was less pronounced in patients with Ri ≥2. Increased Enterococcae and reduced Doreae were associated with ER. Faecalibacterium prausnitzii and Proteus did not predict POR.
A total of 135 samples were taken at M6. There was a statistically significant difference in beta diversity when comparing patients with and without ER (but not with clinical recurrence – HBI ≥/<4). The alpha diversity in patients with ER was numerically reduced. Higher Rutgeerts scores were associated with increased Proteobacteria and a reduction in Firmicutes.
M0 samples were then reviewed to assess for predictors of POR at M6. Patients who had received antibiotics pre-operatively were excluded (57/153) given the changes noted in alpha and beta diversity in this cohort. Alpha diversity at M0 did not predict ER at M6. A multivariable analysis was performed accounting for the predetermined predictors of POR (male gender, smoking, prior resection, anti-TNF therapy postoperatively) . In patients who did not receive anti-TNF prophylaxis postoperatively, the presence of Gammaproteobacteria, Ruminiclostridium 6 and Ruminococcus gnavus at M0 were significantly associated with ER (odds ratios 16.7, 0.02 and 21.8). On inclusion of those receiving anti-TNF therapy, only Ruminiclostridium 6 remained significantly protective (OR 0.17), whilst Corynebacterium was significantly associated with POR (OR 9.8).
A predictive score of POR was created using these two taxa and the pre-set predictors of POR. This correlated with Rutgeerts scores in patients not receiving anti-TNF therapy, whereby a greater number of risk factors were associated with a higher proportion of patients with Ri ≥2: 14%, 53% and 86% with ≤1, 2 or >2 risk factors respectively. When anti-TNF therapy was included in the score as a protective factor, the correlation persisted. The random forest method was also performed, combining several other taxa most correlated with disease and the predetermined risk factors for POR. Correlations were strongest with Streptococcus, R. gnavus and Gammaproteobacteria. Correlations were unaffected by the added clinical parameters but were lost in patients receiving anti-TNF therapy.
This is the largest study to investigate postoperative changes within the microbiome in patients with CD and is novel in its specific assessment of the ileal microbiota.
By comparing microbiota composition at M0 and M6 the authors demonstrated that postoperatively there is a reduction in the classical dysbiosis seen in ileal CD  that is less pronounced in the presence of POR. This adds to the theory that active CD is associated with dysbiosis. It is perhaps unsurprising that there was no correlation between alpha/beta diversity and clinical recurrence; clinical scores have lower specificity for active disease when compared to endoscopy and, postoperatively, new diagnoses may mimic IBD (e.g. bile acid malabsorption or small intestinal bacterial overgrowth).
It is interesting that Faecalibacterium prausnitzii and Proteus, bacteria identified in previous studies as predictors of POR [3, 8], did not show a correlation in this study. The abundance of F. prausnitzii was too low to draw meaningful conclusions. Geographical location and antimicrobial prophylaxis are proposed as explanations with regards to Proteus. Changes in microbiota composition were noted in patients prescribed preoperative antibiotics, suggesting that postoperative antibiotics may have similar effects. Imidazole use postoperatively is best practice  and patients in this cohort did not appear to receive antimicrobial prophylaxis. Excluding patients on antibiotics from the analysis could have arguably removed patients with the most aggressive disease phenotype, thus affecting the results. Effect sizes were also largely lost once anti-TNF therapy had been accounted for. Additionally, bacterial function is affected to a greater degree than bacterial composition in patients with IBD in comparison with healthy individuals, so analysis of species proportions focusses on only part of the picture . These factors pose questions with regard to the generalisability of these data in a real-world setting. They highlight the effects of routine CD care on the microbiome and therefore the challenges we face in using the microbiota as a biomarker for disease activity, demonstrating that we are way off completing our understanding of how the microbiome may predict disease in a given individual.
That being said, using two different models the authors demonstrated that ileal microbiota could predict POR. The most compelling data were obtained in patients not receiving anti-TNF therapy, perhaps highlighting a possibility to select patients for postoperative prophylaxis in a group presumably deemed at lower risk of POR.
- Dey N, Soergel DAW, Repo S, Brenner SE. Association of gut microbiota with post-operative clinical course in Crohn's disease. BMC Gastroenterol. 2013;13:131.
- Wright EK, Kamm MA, Wagner J, et al. Microbial factors associated with postoperative Crohn's disease recurrence. J Crohns Colitis. 2017;11:191–203.
- Mondot S, Lepage P, Seksik P, et al. Structural robustness of the gut mucosal microbiota is associated with Crohn's disease remission after surgery. Gut. 2016;65:954–62.
- Haberman Y, Tickle TL, Dexheimer PJ, et al. Pediatric Crohn disease patients exhibit specific ileal transcriptome and microbiome signature. J Clin Investig. 2014;124:3617–33.
- Auzolle C, Nancey S, Tran-Minh ML, et al. Male gender, active smoking and previous intestinal resection are risk factors for post-operative endoscopic recurrence in Crohn's disease: results from a prospective cohort study. Aliment Pharmacol Ther. 2018;48:924–32.
- Ames NJ, Ranucci A, Moriyama B, Wallen GR. The human microbiome and understanding the 16S rRNA gene in translational nursing science. Nurs Res 2017;66:184–97.
- Morgan XC, Tickle TL, Sokol H, et al. Dysfunction of the intestinal microbiome in inflammatory bowel disease and treatment. Genome Biol 2012;13:R79.
- De Cruz P, Kamm MA, Hamilton AL, et al. Crohn's disease management after intestinal resection: a randomised trial. Lancet. 2015;385:1406–17.
- NICE. Crohn's disease: management. NICE Guideline [NG129] 2019.
Susanna Meade – Short biography
Susanna Meade is a UK gastroenterology registrar currently working at Guy’s and St Thomas’ Hospital in London. She is currently taking time out of training to undertake research at King’s College London and is particularly interested in postoperative recurrence and prophylaxis of Crohn’s Disease.