The gut microbiome seems an integral part in maintaining the tight junction integrity [11]. Evidence has suggested that perturbations in the gut microbiome can lead to an increase in gut permeability, a decrease in the thickness of the protective mucus layer which culminates in pathogen invasion [12]. The goblet cells, found in the intestinal lining, produce a thick mucus layer which acts as a mechanical barrier, as well as chemical, in that it supports the presence of antibacterial proteins, such as secretory IgA and lactoferrin [13,14]. Therefore, a loss of goblet cells, as found in chronic bowel inflammation, leads to a loss in mucosal integrity and impaired barrier function. This culminates to an increase in bacterial translocation altering the T-cell profile and pro-inflammatory cytokines leading to tissue damage, a destabilised microbiome and further tissue damage. Short-chain fatty acids (SCFA), mainly butyrate, are produced via selective bacterial fermentation of resistant carbohydrates. SCFA play an important role in the maintenance of the epithelial barrier and regulation of the immune system, through increased intestinal IgA production, induction of tolerogenic dendritic cells and higher regulatory T-cell percentages [15]. Indeed, bowel inflammation has been associated with increased rates of intestinal regulatory T lymphocytes, however it remains unclear which subsets of regulatory T cell exert proinflammatory effects in UC [16].

From an immune perspective, the gut microbiome is in constant communication with the immune system to help with immune tolerance and disease pathogenesis [17]. Breakdown of the epithelial innate immune function may also lead to the invasion of the intestinal lining by pathogenic bacteria. Indeed, defective autophagy impairs lysosomal digestion and clearance of pathogenic bacterial strains. Activation of different signalling cascades, mediated by the Toll-like receptors (TLR) located on the surface of intestinal epithelial cells and the downstream nuclear factor-κB (NF-κB) pathway, leads to the production of cytokines and recruitment of the adaptive immune system, thus enhancing bacterial clearance. Thus, genetic polymorphisms of TLR, inhibition of the NF-kB pathway and decreased release α-defensin seen in variants of the caspase recruitment domain/nucleotide-binding oligomerisation domain (CARD/NOD family) all contribute to increased intestinal permeability and subsequent bowel inflammation, as seen in UC [18].

Through advancing next generation sequencing technologies, we have been able to understand both the composition of the gut microbiota and also its functionality [19]. Figure 1 summarises the critical roles of the microbiome in drug metabolism, absorptive and secretory capacities of the intestinal lining. Although data remain heterogenous, consistent findings of perturbations in the gut microbiota in patients with UC compared with healthy controls persist. In an inception cohort, Schirmer et al. evaluated the microbial taxonomic changes of 405 paediatric new-onset, treatment-naïve UC patients, who received either corticosteroids or 5-aminosalicylic acid drugs. Interestingly, after 52 weeks of follow-up, poor response to therapy correlated with an increased resemblance to the bacterial taxa of the oral cavity, such as higher levels of H. parainfluenzae, likely explained by intestinal inflammation and strain-specific adaptation [20]. However, it remains unclear whether such microbial changes may be the cause or the consequence of active bowel inflammation.

Interestingly, microbial signatures may differ between UC and CD despite their similar clinical and epidemiologic profiles. In a Spanish cohort of 34 patients with CD and 33 with UC, those afflicted with CD showed, in faecal samples, higher rates of unstable microbial communities compared to patients with UC, as well as increasingly altered microbiota composition, such as a lower relative abundance of Faecalibacterium and a higher relative abundance of Fusobacterium and Escherichia. The composite microbial biomarkers were validated as a diagnostic tool in Spanish and Belgian cohorts, demonstrating a sensitivity of 80% and a specificity of 90.9% in distinguishing CD from UC [21].

The microbiota of a patient with pouchitis differs from a patient without pouchitis. Specific patterns have found persistence of Fusobacter and Enterobacters associated with the disease state and the absence of specific bacteria such as Streptococcus species in the inflamed pouch [22]. Clinically, pouchitis can be treated with antibiotics and hence provides plausibility that the microbiome may influence the course of pouchitis. Of interest, mucosal inflammation in the pouch is concentrated in areas where bacterial concentration is highest [23]. From an immune perspective, it has been noted that pouchitis-derived bacterial sonicates from metronidazole-sensitive bacterial species stimulate healthy patients’ mononuclear cells significantly more than corresponding sonicates from non-pouchitis patients [24].

Potentially, the closest direct link to the microbiome having an influence on pouchitis came from a study which highlighted that baseline microbiome prior to colectomy could predict those that developed pouchitis and those that did not and hence provides suggestions that altering the gut microbiota may influence the pouch functionality [25].

Dietary prebiotics are defined as ‘a substrate that is selectively utilised by host micro-organisms conferring a health benefit’ [26]. It is hypothesised that prebiotics may improve gut inflammation by selective stimulation of protective members of gut microbiota, improvement of the intestinal permeability and increased production of SCFA [27].

Hafer et al. studied the effect of lactulose at a daily dose of 10 g added to standard therapy in 7 patients with active UC, compared to 7 UC patients receiving standard therapy without the use of a placebo. They noted that the Inflammatory Bowel Disease Questionnaire (IBDQ) score improved from 123 ± 20 to 171 ± 18 (p = 0.026) in the lactulose group [28].

Two clinical trials evaluated the effects of an oligofructose-enriched inulin compound (Beneo™ Synergy 1) in UC patients with mild to moderate disease, receiving concomitant mesalazine therapy. Casellas et al. noted that, at day 14, levels of faecal calprotectin improved in five of seven patients (70%) who received the prebiotic, compared to two of eight patients (25%) in the placebo arm [29].

Valcheva et al. assessed the alterations of the gut microbiota composition and activity in 25 patients with mild to moderate UC treated with different doses of oligofructose-enriched inulin over a nine-week period. The primary outcome was clinical response and/or remission. The primary outcome was achieved in 77% of patients receiving the high-dose prebiotic product (15 g per day) compared to 33% in the low-dose group (7.5 g per day). High-fructan dose was associated with Bifidobacteriaceae and Lachnospiraceae abundance, however such microbiota modifications did not correlate with improved disease scores. Interestingly, the trial showed that a prebiotic course resulted in higher butyrate levels, with strong negative correlations between butyrate levels and clinical symptoms [30]. Existing trials assessing the use of prebiotics in the treatment of UC lack sufficient power to change clinical practice, however data regarding its potential efficacy and safety profile are encouraging.

Probiotics are ‘live microorganisms which confer a health benefit on the host when administered in adequate amounts’ [26]. Probiotics are traditionally composed of one or more bacterial strains.

Derwa et al. performed a meta-analysis of eight trials targeting induction of remission in active UC as a primary outcome (n = 651), as well as six trials assessing prevention of relapse in quiescent UC (n = 677) [31]. Types of probiotics varied between E. coli Nissle 1917 (5 studies), Bifidobacterium longum 356 (1 study), Lactobacillus rhamnosus GG (1 study), a multistrain probiotic containing a combination of lactic acid bacteria, streptococci and bifidobacterial (3 studies) and other combined formulations (4 studies). In the single trial comparing probiotics with 5-ASAs for induction of remission in 116 patients, no difference was seen in the primary outcome of failure to achieve remission (RR = 1.24; 95% CI = 0.70–2.22), similar to the pooled 7 randomised placebo-controlled trials (RR = 0.86; 95% CI = 0.68–1.08). Rates of adverse events were comparable in both analyses. Interestingly, in a subgroup analysis of the multistrain probiotic containing a combination of lactic acid bacteria, streptococci and bifidobacteria studies, 56.2% of 162 patients randomised to the probiotic failed to achieve remission, compared with 75.2% of 157 patients who received placebo (RR = 0.74; 95% CI = 0.63–0.87). The authors measured a number needed to treat of 5 to prevent one patient with active UC failing to achieve remission, without significant heterogeneity between studies (I² = 0%, p = 0.52). In a pooled group of 140 patients, the E. coli Nissle 1917 compound did not demonstrate a statistically different benefit compared to placebo (RR = 1.56; 95% CI = 0.44–5.53). In a randomised, double-blind trial, seven weeks of E. coli Nissle 1917 or placebo was added to conventional therapy in patients with active UC and initially treated with one week of ciprofloxacin or placebo (25 patients in each of the four groups). Surprisingly, rates of clinical remission and treatment persistence with the study drug were significantly lower in patients treated with E. coli Nissle 1917 and without a previous antibiotic cure, despite similar adverse events in all groups [32]. In regard to maintenance of remission, use of probiotics in 342 patients was not shown to decrease rates of UC relapse compared with the use of 5-ASAs (RR = 1.02; 95% CI = 0.85–1.23) in 278 patients and placebo (RR = 0.62; 95% CI = 0.33–1.16) in 57 patients.

Astó et al. conducted a meta-analysis of randomised controlled trials (RCTs) examining the effects of probiotics, prebiotics and synbiotics on human UC [33]. Rates of remission in 602 patients with active UC were unchanged between the probiotics and placebo groups. In trials defining UC remission, the beneficial effects of probiotics were estimated to be statistically significant compared to placebo (RR = 1.55, 95% CI = 1.13–2.15) with decreased heterogeneity between trials (I² = 29%). On further subgroup analysis (n = 424), patients with active UC who received Bifidobacterium-containing probiotics were more likely to be in remission compared to those on placebo (RR = 1.73; 95% CI = 1.23–2.43, p = 0.002). In comparison, no difference in UC remission was seen between probiotics without Bifidobacterium strains and control groups (n = 168). In trials assessing the multistrain probiotic containing a combination of lactic acid bacteria, streptococci and bifidobacteria in combination with standard therapy (n = 348), significantly higher rates of UC remission were seen in the probiotic group compared to the control group (RR = 1.99; 95% CI = 1.25–3.15, p = 0.003).

The faecal concentrations of SCFA were measured in two trials and were significantly increased in one pilot study with active UC patients after supplementation of Bifidobacterium-fermented milk (n = 20) [34]. In a trial assessing 46 inactive UC patients, SCFA concentrations did not differ significantly between probiotic (Streptococcus faecalis T-110, Clostridium butyricum TO-A and Bacillus mesentericus TO-A) and placebo groups at any time over the six months, however a higher butyrate/acetate ratio was observed throughout the follow-up period in patients who relapsed compared to those who remained in remission [35]. Decreased Bifidobacterium species was observed in 195 inactive UC patients prior to relapse in one study assessing the effects of Bifidobacterium breve fermented milk [36]. Furrie et al. explored the effects of a synbiotic formulation containing B. longum and oligofructose-enriched inulin; improved endoscopic scores and significantly higher levels of bifidobacterial rRNA on mucosal biopsies were observed in 9 patients receiving this synbiotic compared to 9 patients on placebo [37].

Antibiotics are seldom used in clinical practice for the management of UC. A few initial studies suggested a potential clinical benefit of adding antibiotics, such as ciprofloxacin or tobramycin, to conventional therapy in patients with active UC of all severities [38,39]. Khan et al. performed a meta-analysis of nine RCTs assessing the efficacy of antibiotics in adult patients with active UC [40]. Efficacy outcomes were mostly clinical with limited reporting of biochemical and endoscopic outcomes. Overall, the authors noted a statistically significant benefit favouring antibiotics over placebo (RR 0.64; 95% CI = 0.43–0.96, p = 0.03), however antibiotic regimens were significantly heterogeneous between trials. Two RCTs assessed the use of antibiotics in acute severe UC, a condition perceived at high risk of bacterial translocation, and showed no short-term clinical benefit of adding metronidazole (trial of 39 patients) or metronidazole and tobramycin (trial of 39 patients) compared to placebo [41,42]. Internationally recognised guidelines either do not recommend their use or do not mention them as a potential therapeutic option in the management of adult patients with UC [43,44].

Since 2015, four placebo-controlled RCTs [45,46,47,48] and multiple cohort studies have been published [49,50,51], with meta-analyses suggesting a positive impact of faecal microbiota transplantation (FMT) in the induction of remission in UC patients with mild-moderate disease [52].

Costello et al. conducted a meta-analysis of the four placebo-controlled RCTs, totalling 277 patients. Clinical remission was achieved in 28% of pooled donor FMT groups compared with 9% of patients in placebo groups (OR = 3.67; 95% CI = 1.82–7.39) [53]. A Danish open-label pilot study has examined the efficacy of oral FMT capsules in patients with active UC; over a 50-day course of oral FMT capsules, all of the seven patients achieved clinical response at weeks 4 and 8, as well as significant improvements in quality of life and faecal calprotectin levels [49].

Paramsothy et al. performed gastrointestinal microbial community profiling in 81 UC patients treated with colonoscopy delivered FMT versus placebo. Bacterial diversity in samples before and after FMT administration was higher in recipients who achieved remission compared with those who did not. Remission after FMT was associated with a relative microbial abundance of Eubacterium hallii and Roseburia inulivorans, compared with higher levels of Fusobacterium gonidiaformans, Sutterella wadsworthensis and Escherichia species in patients not determined to be in remission post-FMT. The former patient group exhibited increased production of SCFA and secondary bile acids, while the latter group showed higher levels of heme and lipopolysaccharide biosynthesis. A relative abundance of Bacteroides species in donor FMT stools was associated with higher rates of remission in recipients, likely explained by antagonist interactions between Bacteroides and Prevotella species [54].

The beneficial effects of FMT in UC may also be mediated by other members of the intestinal microbiota, such as viruses and fungi. In a small cohort of nine UC patients, a numerical trend towards reduction in eukaryotic viral richness was observed in FMT responders compared with non-responders (p = 0.056) [55]. The pro-inflammatory role of Candida species was highlighted in a prospective trial of 39 patients. A relative abundance of Candida species pre-FMT was associated with increased bacterial diversity, which likely implies a microbiota more amenable to FMT engraftment. A reduction of Candida species post-FMT administration correlated with improved clinical and endoscopic outcomes; such an impact was not reproduced in patients receiving placebo [56]. A better understanding of the FMT-induced modifications of bacterial taxonomy, as well as trans-kingdom interactions, will hopefully improve the selection process of FMT donors and recipients, thus improving the overall efficacy of FMT in the management of active UC.

An increasing number of studies have highlighted the importance of macronutrients in the aetiology of IBD [57,58,59], however data is limited regarding their impact on the gut microbiota. In one study that explored the role of animal-based diets, it was found that the increase in the abundance and activity of Bilophila wadsworthia on the animal-based diet group supports a link between dietary fat and the overgrowth of microbes able to prompt the host to IBD [60]. Table 1 highlights the published studies assessing the impact of diet in the incidence and treatment of UC.

Welters et al. evaluated the effects of enteral inulin on ileo-anal pouch functioning by studying epithelial gene expression, cell turnover and mucosal morphology. Twenty patients were given 24 g of inulin daily for three weeks, then a four-week wash-out period, and a placebo for three weeks. Inulin supplementation did not significantly alter pouch mucosal functioning because neither epithelial homeostasis nor epithelial gene expression was significantly altered; however, the author concluded that enteric supplementation with 24 g/day of inulin led to a decrease of inflammation-associated factors, with an increase in butyrate production, decrease of secondary bile acids and significant decrease in the endoscopic and histologic pouch disease activity index score [79]. This is yet to be linked to the role in the microbiome but could provide some interesting mechanistics as to the importance of prebiotics in pouch integrity.

In a meta-analysis of the four RCTs, it was shown that probiotics have no effect on the maintenance of remission in pouchitis. However, when exploring individual studies, it has been demonstrated that 6 g/day of a multistrain probiotic containing a combination of lactic acid bacteria, streptococci and bifidobacteria can help maintain remission [80] and prevent acute pouchitis [81,82]. There is a single small trial suggestive that probiotics may be effective for acute pouchitis [83] but this was not replicated in a RCT [84] and hence its position in acute pouchitis remains uncertain.

When exploring some of the microbial mechanisms that changed during probiotic treatment of the pouch, Gionchetti et al. noted that faecal concentration of lactobacilli, bifidobacteria and S. thermophilus increased significantly from baseline levels only in the group receiving the multistrain probiotic containing a combination of lactic acid bacteria, streptococci and bifidobacteria (p < 0.01) [81]. The same group in a follow-up study found that the patients that benefited from the multistrain probiotic were associated with faecal colonization with probiotics [83]. In the other probiotic study reporting on changes in the gut microbiome, Kuisima et al. highlighted that a trial of Lactobacillus GG supplementation (10 LGG, 10 placebo) for 3 months changed the pouch microbiota, but was ineffective as primary therapy for a clinical or endoscopic response [85].

Antibiotics remain the mainstay treatment for both acute and chronic pouchitis. In a meta-analysis it was highlighted that antibiotics could achieve remission in nearly three-quarters of cases [86]. However, these are based on a number of small randomised controlled trials and observational studies. When exploring the microbial changes that occur following antibiotics, the evidence is very heterogenous, meaning consistent signals are not yet found. Furthermore, the mechanisms that underpin the microbial changes that are responsible for the beneficial effect remain poorly understood. In one of the few mechanistic studies exploring the role of antibiotics and pouchitis, it was highlighted that antibiotics lead to an antibiotic-resistant microbiome to include abundance of facultative anaerobic bacteria genera to include Escherichia, Enterococcus and Streptococcus. This was associated with low virulence which helps to maintain remission [87]. This therefore suggests that antibiotics may be decreasing the virulence of the bacteria contributing to pouchitis.

There have been a number of small studies that have explored the potential of faecal microbiota transplantation for chronic pouchitis. Overall, a meta-analysis highlighted that as yet there was a lack of evidence for its effectiveness in the treatment of chronic pouchitis [88]. The problem with many of these studies is the heterogeneity in study design, the methodology of delivering the faecal transplant, the relatively small number of patients and the variability of diagnosis of pouchitis. Interpretation of these studies remains challenging, especially understanding mechanisms that may underpin therapeutic success.

Due to the small number of studies related to pouchitis, larger studies are needed to guide our understanding of future therapies to treat pouchitis. It is likely that a multicentre approach with mechanistic work that aims to understand the pathophysiology of pouchitis will help guide our treatment strategies and begin to personalise them.