Therapeutic Manipulation of the Microbiome in IBD: Current Results and Future Approaches

Opinion Statement

Despite recent major strides in our understanding of the genetic and microbial influences that contribute to the development of the inflammatory bowel diseases (IBDs), their etiology continues to be enigmatic. Results from experiments in animal models of IBDs overwhelmingly support a causal role of the microbiota in these diseases, though whether such a cause-effect relationship exists in human IBDs is still uncertain. Therefore, virtually all currently-approved, and most often prescribed, treatments for IBDs are directed toward the over-active immune response in these diseases rather than the intestinal bacteria. Nevertheless, there is an important need for non-immunosuppressive therapies that may present a more favorable risk-benefit profile such as those that selectively target the disruptions in gut microbiota that accompany IBDs. This need has led to clinical trials of various microbial-directed therapies including fecal microbial transplant, antibiotics, probiotics, and prebiotics. Unfortunately, these published studies, many of which are small, have generally failed to demonstrate a consistent benefit of these agents in IBDs, thus leading to slow acceptance of microbe-focused treatments for these conditions. In this article, we review and summarize the microbial basis for IBDs and the results of the most recent trials of fecal microbial transplant, antibiotics, probiotics, and prebiotics in IBDs. We also comment on possible safety concerns with these agents, speculate on why they have failed to show efficacy in certain clinical settings, and propose strategies to improve their usefulness.

Introduction

IBDs, including Crohn’s disease (CD) and ulcerative colitis (UC), are chronic, immune-mediated disorders of the gastrointestinal tract that cause significant morbidity and are particularly prevalent in developed countries, where approximately 4 in 1,000 individuals are afflicted [1]. CD is characterized by patchy, transmural inflammation of any portion of the gastrointestinal tract and is often complicated by strictures, abscesses and fistulae. UC, on the other hand, is restricted to the superficial layers of the colon. There are currently no medical cures for IBDs and curative surgical resection is only an option for UC patients. A prevailing hypothesis is that IBDs are caused by overaggressive immune responses to intestinal microbiota in genetically predisposed individuals. This hypothesis is supported by recent, large genome wide association studies that have identified susceptibility loci in or near genes involved in innate and adaptive immune responses to microbes [2].

Intestinal microbiota in the pathogenesis of inflammatory bowel diseases

While the roles of resident intestinal microbes in the pathogenesis of IBDs has been suspected for decades, molecular analysis of the human intestinal microbiome over the past five years using culture-independent, high-throughput DNA sequencing methods has significantly accelerated our understanding of how the composition and functions of the gut microbiota correlate with IBDs and has brought us a step closer to determining the etiology of these diseases. With this more complete knowledge of the altered gut microbes in IBDs, often referred to as dysbiosis, we are now poised to determine whether these differences are the cause or result of intestinal inflammation and ultimately design therapies that target the dysbiosis.

Inflammatory bowel diseases are associated with altered gut microbial composition

Numerous cross-sectional analyses of gut bacterial composition in patients with IBDs have been previously published and results are summarized elsewhere [3]. While study designs and findings are somewhat variable between individual studies, the most consistent results are that bacteria from the Proteobacteria phylum are increased and those from the Firmicutes phylum are decreased in feces from patients with IBDs compared with non-IBD controls. Moreover, overall bacterial diversity is consistently decreased during intestinal inflammation. However, a similar dysbiosis also occurs in patients with Clostridium difficile colitis [4] suggesting that such changes in microbiota may not be unique to IBDs and that microbiota-directed therapies that prove successful in treating C. difficile colitis may also be effective for some IBDs.

A recently-published study of intestinal bacterial composition in IBDs included 447 pediatric patients with newly-diagnosed Crohn’s disease and 221 non-IBD controls, the largest study to date [5]. The authors detected increased Enterobacteriaceae, Pasteurellaceae, Veillonellaceae, Neisseriaceae, and Fusobacteriaceae in ileal and rectal biopsies from CD patients vs. controls. On the other hand, members of the Bifidobacteriaceae, Erysipelotrichales, Bacteroidales, and Clostridiales were decreased in CD patients. These results in ileal and rectal biopsies are generally consistent with previously published data in feces from adult IBD patients. Interestingly however, the authors of this pediatric study found that CD was associated with relatively few changes in fecal bacterial composition compared with the more robust changes observed in tissue biopsies. These data suggest that fecal bacterial communities are different from those near the mucosa and may not be as relevant to the pathogenesis of CD. However, similar to all previously-published studies, this one was cross-sectional by design and therefore cannot distinguish whether the observed changes in bacterial composition are the cause or the result of CD.

A subsequent study of the afore-mentioned pediatric cohort, but also including ulcerative colitis patients, indicates that the bacterial compositional changes occur independently of gut inflammation and therefore provides indirect evidence that bacterial dysbiosis may cause disease in humans [6]. The authors detected altered bacterial composition in ileal biopsies from IBD patients with colon inflammation, but without ileal inflammation. Moreover, they demonstrated that low levels of members of the Veillonella genus of the Firmicutes phylum in ileal tissue predicted six month steroid free remission rates with an adjusted odds ratio of 0.23. Therefore, it is conceivable that altered bacterial community composition may contribute to the pathogenesis of IBDs. Longitudinal studies of large numbers of individuals pre- and post-diagnosis would be needed to answer this question more definitively.

The vast majority of published microbiome studies in IBD have focused exclusively on bacteria, likely because the currently-available tools for high-throughput DNA sequence data analysis are much more robust for bacteria compared with those for other intestinal microbes such as viruses, archaea, and fungi. Nonetheless, we are beginning to understand how non-bacterial microorganisms play a role in IBDs. Though fungi account for only 0.02–0.03% of the fecal microbiota [7], fungal species richness and diversity are higher in mucosal biopsies from CD patients vs. controls, a finding that is opposite of what has been observed with bacteria [7,8]. Moreover, several fungal species, including Candida spp., Giberella moniliformis, Alternaria brassicicola, and Cryptococcus neoformans, are increased in tissues from CD patients vs. controls [8]. Similar to fungi, only a few studies have been published examining the populations of viruses in patients with IBD. CD patients may harbor increased numbers of bacteriophages in inflamed tissue and feces, though no specific viruses have been associated with human IBDs to date [9,10].

Inflammatory bowel diseases are associated with altered gut microbial function

While the composition of the intestinal microbiota likely plays an important role in IBDs, it is becoming increasingly clear that the effects of microbes on host pathophysiology depend more upon their aggregate functional capacity as determined by the presence and expression of specific genes. For example, in one study, only 2% of bacterial genera were perturbed in IBD whereas 12% of bacterial pathways were affected [11]. Several studies have shown that bacterial genes encoding proteins in specific functional pathways are enriched in fecal and mucosal specimens from patients with IBDs vs. controls. For instance, fecal bacterial communities from IBD patients contain fewer genes in biosynthetic pathways and more genes in oxidative stress and nutrient transport/uptake pathways compared with controls [5,11]. Thus, bacterial communities in IBD patients may shift their function towards pathways that promote survival rather than growth in the inflamed environment. Whether functional changes also occur in viruses, fungi, or archaea remains to be determined.

Fecal microbial transplant as therapy for inflammatory bowel diseases

Although the causal nature of microbial dysbiosis in IBDs has not been definitively established, investigators have pressed forward with attempts to treat IBDs by correcting the dysbiosis. For instance, fecal microbial transplant (FMT), in which unfractionated fecal microbiota from a healthy donor is instilled into the intestinal tract of a sick individual with the intent of curing the disease, has recently been proven as a highly effective therapeutic option for recurrent C. difficile colitis [12]. FMT is thought to work by correcting the relatively depleted, homogeneous intestinal microbiota that facilitated the outgrowth of toxin-expressing C. difficile organisms. Since the dysbiosis that occurs during C. difficile colitis and IBDs is generally similar, and assuming that this dysbiosis is a cause of IBDs, a concept that still remains to be definitively proven, it is reasonable to suspect that FMT may also effectively treat IBDs.

The first published use of FMT for IBD was in 1989 in a Kansas City surgeon with UC who entered a medication-free remission within 1 week of the transplant [13]. Since then, numerous retrospective case series on this topic have been published, with most of the reports of successful treatments from a single author. The first published prospective trial of FMT in IBD included nine pediatric patients with UC that were treated with four FMT enemas per day for five consecutive days [14]. The majority of patients had a clinical response four weeks later. No control group was included. In a separate trial, one-time administration of FMT via nasogastric tube in four pediatric patients with UC showed no benefit, but the inability of gut microbes to survive the gastric environment may have been a limiting factor [15]. An uncontrolled prospective trial of FMT delivered during colonoscopy to 6 adults with “medically refractory” UC resulted in improved clinical scores in 50%, but none entered remission [16]. Analysis of the post-transplant fecal bacterial composition in this study showed increased species richness, decreased Proteobacteria, and increased Bacteroidetes. However, microbiota changes did not correlate with clinical improvement. The first prospective trial of FMT for CD included 30 Chinese patients with “refractory” CD who were treated with one FMT delivered to the jejunum during endoscopy. Half of the patients were in clinical remission one year later though no control group was included [17]. Based on results from these uncontrolled studies, it is apparent that certain subsets of IBD patients might benefit from FMT though the optimal source of donor feces, route of delivery, frequency of administration, and patient selection criteria are unclear.

Given the possible benefit of FMT in IBD noted in these pilot studies, randomized, placebo controlled trials of FMT in patients with active UC are currently being conducted. The first results of these has been published in abstract form [18]. In this study, 61 adults with active UC were randomized to FMT or placebo enemas once weekly for six weeks. The primary endpoint of remission at week seven with a Mayo endoscopic subscore of 0 occurred in 23% of patients in the active treatment arm vs. 7% in the placebo arm (p=0.15). Based on this interim analysis, the investigators decided to stop the study early due to futility. Sixteen of the 31 patients receiving active treatment subjectively felt better at seven weeks and were therefore offered weekly FMT enemas for six additional weeks. Overall, 33% of patients in the active treatment arm were in remission if treatment was extended. Interestingly, bacterial diversity was not affected by FMT, but butyrate producing bacteria including Ruminococcus, Blautia, and Lachnospiraceae were increased in feces from patients that responded to FMT [19]. This change is important since butyrate is a short chain fatty acid that enhances intestinal epithelial cell integrity and anti-inflammatory immune responses.

In summary, the use of FMT to treat IBDs has met with variable success thus far. Based on small pilot trials in humans, it is rational to suspect that FMT might at least transiently correct the dysbiosis associated with IBDs. However, it is still unclear whether the dysbiosis is a primary or secondary event and therefore whether correcting the dysbiosis will truly impact the disease. The placebo controlled trial (considered the gold standard in clinical therapeutics) of FMT in UC failed to meet its primary endpoint. However, on average, more patients who were treated with FMT achieved clinical and endoscopic remission compared with placebo suggesting that certain subgroups of IBD patients may benefit from this treatment. In addition, optimizing the route, frequency, and length of treatment, pretreatment preparation (antibiotics, bowel prep), processing of the donor specimen, as well as identifying the most beneficial microbes in the donor feces could lead to the design of more effective microbial-based therapies for IBDs that are easier to mass produce, store, and administer for prolonged periods.

Antibiotics as primary and adjunctive treatment

Whereas FMT attempts to correct the microbial dysbiosis by introducing bacteria into the intestinal ecosystem, antibiotics typically perturb the gut microbes by removing large communities of bacteria and have long been used for primary therapy of CD and pouchitis, as well as for superinfections, bacterial overgrowth, septic complications and preoperative preparations of IBD [20,21]. Unfortunately, this historic use of antibiotics has relatively weak support by large randomized, placebo-controlled trials [22,23]. Several meta-analyses uphold a role for antibiotics in treating both Crohn’s disease and ulcerative colitis [24,25] and antibiotics are universally accepted as the primary treatment of pouchitis.

Crohn’s disease

Metronidazole, ciprofloxacin, the combination of these two agents, and rifaximin have reasonably good support for treating CD, particularly Crohn’s colitis and ileocolitis. In one randomized, placebo-controlled study, patients with active colonic and ileocolonic CD treated with metronidazole for four months significantly improved, but those with isolated ileal involvement did not [26]. Trials of ciprofloxacin for six weeks [27] or for six months [28] likewise demonstrate benefit. The largest trial combining metronidazole and ciprofloxacin did not show improvement over budesonide alone in patients with ileal or ileocecal CD [29], but these antibiotics likely work better in the colon than the ileum. In a dose ranging trial involving 402 patients with moderately active CD, 800 mg of rifaximin b.i.d. was superior to placebo (62% vs. 43% remission), with a surprising lack of benefit from 1200 mg b.i.d. [30]. Early stage disease, colonic involvement and elevated CRP values were predictors of best response to rifaximin [30]. A two-year, 213 patient placebo-controlled trial of triple antibiotics (clarithromycin, rifabutin and clofazimine) designed to treat Mycobacterium avium paratuberculosis showed benefit at four months (66% of those taking antibiotics were in remission vs. 50% taking placebo), but no lasting benefit one year after stopping therapy [31]. Unfortunately, patients were not tested for M. paratuberculosis colonization, so there is no evidence of a direct M. paratuberculosis effect rather than broad antibiotic effect on resident microbiota.

Antibiotics, particularly metronidazole and ciprofloxacin, are also beneficial with and without anti-TNF agents for perianal fistulae complicating CD. Forty to 50% improvement rates in fistula symptoms have been reported for high-dose metronidazole [32] or ciprofloxacin [33] alone, although success is variable [33]. The addition of ciprofloxacin to either infliximab [34] or adalimumab [35] significantly improved healing of perianal fistulae, although fistula closure can be transient after cessation of antibiotic treatment. Similar improvement was seen with antibiotics added to azathioprine in an uncontrolled study [36].

Antibiotics have a potential but still unclear role in the evolving strategies to prevent postoperative recurrence of CD. Initial studies by Rutgeerts et al showed a transient decrease in endoscopic recurrence using high-dose (20 mg/kg) metronidazole [37] or ornidazole [38] for three and six months, respectively, beginning immediately after resection. The addition of metronidazole to azathioprine was reported to improve outcome compared with azathioprine alone [39], although a smaller, probably underpowered, similarly designed study showed only a nonsignificant trend toward improvement in endoscopic scores at six and 12 months after surgery [40]. A small pilot study of ciprofloxacin for postoperative prevention showed a trend toward protection for ciprofloxacin based on six month colonoscopy scores, but more patients taking antibiotics dropped out vs. placebo [41].

Ulcerative colitis

Use of narrow spectrum antibiotics, such as ciprofloxacin or metronidazole, in ulcerative colitis is not supported by the few available trials. Many of these studies were performed in hospitalized patients with severely active disease [42–44]. One six-month study reported benefit of ciprofloxacin over placebo as adjunctive therapy in patients treated with corticosteroids and mesalamine [45], but overall there is limited justification for single antibiotics in routine management of ulcerative colitis patients.

In contrast, use of broad spectrum antibiotics in ulcerative colitis appears more promising, although this approach is still not well defined. Combined amoxicillin, tetracycline and metronidazole for two weeks led to one year outcomes superior to placebo in a Japanese study [46]. Follow up mucosal biopsy microbiota profiles showed improvement at three months in the majority of the antibiotic-treated patients, with decreased Fusobacterium varium concentrations, compared to placebo-treated subjects [47]. Likewise, approximately one-half of children with moderate to severe refractory ulcerative colitis responded to a two to three week combination of metronidazole, amoxicillin and doxycycline +/− vancomycin in a small open-labeled study in hospitalized patients [48]. In a similarly refractory 28-patient adult cohort, a 10-day course of rifaximin 400 mg twice daily did not improve overall clinical outcomes [49]. These broad spectrum antibiotic trials need to be expanded to explore outcomes in refractory patients as an alternative to chronic biologic therapies.

Pouchitis

Antibiotics are the primary approach to treating acute and chronic pouchitis following colectomy for UC with creation of an ileal pouch/anal anastomosis (IPAA) [50,51]. A number of antibiotic regimens are effective in inducing clinical remission, although supportive high quality controlled clinical trials are limited. Acute pouchitis occurs in up to 50% of UC patients following IPAA and typically responds to antibiotics. Only metronidazole and ciprofloxacin have been studied in adequately powered prospective controlled trials. In an acute pouchitis study, both ciprofloxacin 500 mg twice daily and metronidazole 20 mg/kg/day for 14 days were effective, but ciprofloxacin was superior and had fewer adverse effects (0 vs. 33%) than metronidazole [52]. In a crossover trial of chronic, unremitting pouchitis, metronidazole 400 mg t.i.d. for two weeks was superior to placebo in reducing stool frequency but did not improve endoscopic or histologic parameters [53]. Despite anecdotally positive clinical experience with rifaximin in pouchitis, it induced only a non-significant 25% remission vs. 0% placebo in a pilot study though it may have been underdosed and underpowered [54]. In an open-label study, symptomatic pre-pouch ileitis associated with pouchitis responded to the combination of metronidazole 800 mg/day and ciprofloxacin 1 gram daily [55].

Chronic antibiotic-dependent or refractory pouchitis is clinically more challenging to treat. A number of combination antibiotic strategies have been proposed to induce and/or sustain remission, at times in parallel with probiotic maintenance, but these are mostly small uncontrolled trials. Combined metronidazole 800 mg/day and ciprofloxacin 1 gram/day for 28 days led to an 82% remission rate and improved quality of life scores in an open label study [56]. Similarly, combined ciprofloxacin 1 gram/day and rifaximin 2 grams/day for 14 days [57] had 88% response. The combination of ciprofloxacin 1 gram/day and tinidizole 15 mg/kg/day for four weeks led to 87.5% remission [58]. Although these studies are small and uncontrolled, combination antibiotic therapy appears to be effective for single-antibiotic refractory pouchitis.

Probiotics

Probiotics, defined as microorganisms with beneficial effects on the host, have widespread appeal for patients wishing to use “natural” approaches. Unfortunately, very few properly designed clinical trials for commercial probiotics have been performed [23,51,59,60]. Nevertheless, the potential use of well-chosen commensal enteric bacterial species with protective effects on colonization resistance, mucosal protection and induction of protective immune responses is exciting and could open new therapeutic approaches for IBD management.

Ulcerative colitis

Although a Cochran systematic review concluded that insufficient evidence is available to support probiotic induction and maintenance of remission in UC [61], clinical trials support a potential role for certain probiotic formulations in UC. In two independent relatively large trials E coli 1917 Nissle was similar to low dose mesalamine (1.2 or 1.5 gm/d) in maintaining remission in UC [62,63]. The combination probiotic, VSL#3, maintained remission in a one-year trial [64]. However, other probiotic preparations are not consistently beneficial in maintenance studies [65,66].

VSL#3 induced remission of active UC in several randomized placebo-controlled studies as primary or adjunctive therapy. In 77 patients with mild/moderate UC, VSL#3 was superior to placebo in inducing remission (43% vs. 16%) by 12 weeks [67]. Of 144 patients taking mesalamine, azathioprine and/or methotrexate, the group receiving adjunctive VSL#3 had a 48% remission rate vs. 32% receiving placebo [68].

Crohn’s disease

Despite abnormal bacterial profiles in CD patients, currently available probiotics are not consistently effective in inducing and maintaining remission or preventing relapse after surgical resection [23,60,69,70]. Several probiotic formulations, including Lactobacillus rhamnosus GG, Lactobacillus johnsonii LA1 or VSL#3 are ineffective at reducing post-operative recurrence of CD [71–74]. However, decreased mucosal levels of a different beneficial bacterial species, Faecalibacterium prausnitzii, that is not included in previously-tested probiotics predict postoperative relapse rates [75]. These studies suggest that the effectiveness of probiotics in treating CD may be critically dependent on the choice of component bacterial strains and that current probiotics may be ineffective because they lack the proper immunoprotective enteric commensal bacterial species such as F. prausnitzii. Optimism for the use of probiotics to treat CD was bolstered by a report that showed efficacy of combining a probiotic (Bifidobacterium longum) with a prebiotic in treating CD [76].

Pouchitis

The very positive results of VSL#3 in pouchitis establish a proof of principle for using probiotics in managing at least this particular manifestation of IBD [50,77–79]. High-dose (1.8 × 10¹² CFU) VSL#3 daily for 9 months dramatically decreased flares of chronic relapsing pouchitis following antibiotic-induced remission compared with placebo (100% vs. 15%) [80]. Interestingly, all patients in the VSL#3-treated group relapsed after stopping treatment. Similar results were noted in a 12 month study [81]. However, an open label US study showed only minimal effects (19% remission) in antibiotic-dependent pouchitis patients [82]. Gionchetti et al also showed that VSL#3 for one year beginning within one week of ileostomy closure decreased onset of pouchitis from 40% in the placebo group to 10% in the treated group [83]. Unfortunately, other agents have not been studied extensively, so it is unclear if these results are unique to VSL#3, although a different probiotic mixture, Ecologic 825, appeared to improve mucosal barrier function as measured by ex vivo bacterial translocation [84]. Lactobacillus GG had no efficacy in acute pouchitis [85], although VSL#3 showed some benefit in an open label study [86].

Prebiotics

Physiologic dietary management of IBDs appeals to patients who are eager to avoid the toxicity and expense of the potent systemic immunosuppressive therapies currently in use. Prebiotics are non-absorbed carbohydrate polymers, such as inulin, fructo-oligosaccharides, and galacto-oligosaccharides that promote growth and metabolic activity of presumed beneficial gut bacteria, most notably Bifidobacterium and Lactobacillus species. Prebiotics can induce bacterial production of short chain fatty acids including butyrate, which has well-documented immunoregulatory effects. These include histone deacetylase and PPARγ activities that inhibit proinflammatory cytokines, and activation of regulatory T cells through G protein receptor 43 and GPR109a [87–89]. In addition, butyrate is the primary metabolic fuel for distal colonic epithelial cells. Multiple studies demonstrate efficacy of oral prebiotics, either alone or in combination with probiotics (synbiotics) in experimental colitis [90], but proof of efficacy in human CD in UC patients is still limited [23,90,91]. A controlled four-week trial of fructo-oligosaccharides demonstrated no effect on remission rates in patients with active CD [92]. In a small controlled trial of patients with active UC, oligofructose-enriched inulin decreased fecal calprotectin levels but did not enhance remission [93]. Combined prebiotic and probiotic Bifidobacterium longum in patients with active CD improved clinical and histologic scores and transiently decreased mucosal TNFα concentrations [76]. An open label, uncontrolled study of combination probiotics and psyllium fiber showed a 70% response and 60% remission rates in steroid refractory CD patients [94], although the combination of four probiotics and four prebiotics had no benefit in preventing postoperative recurrence of CD in a small, placebo-controlled study [95]. A small randomized, controlled trial demonstrated that inulin decreased histologic and endoscopic severity of pouchitis, but did not affect clinical parameters [96]. These studies demonstrate a potential of prebiotics for clinical efficacy, but further investigation is needed.

Potential adverse effects of therapeutic manipulation of the microbiota in IBDs

Given recent data that implicate commensal intestinal microbiota in a wide range of pathological processes, potential short and long term adverse outcomes of various microbiome manipulations must be considered. In addition to transferring microbial pathogens with FMT, mouse studies demonstrate the possibility of transferring undesired disease phenotypes, such as obesity, metabolic syndrome, and fatty liver [97–100]. Cases of febrile reactions and worsening disease activity in studies of FMT for IBDs have been reported [14,15]. Similarly, sepsis with probiotic bacterial strains has been reported, which is particularly concerning for immunocompromised patients with increased mucosal permeability [101]. However, sepsis from probiotics is exceptionally unusual in IBD patients. Increased intestinal bloating is relatively common following prebiotic and probiotic administration, can limit dose escalation, and lead to increased dropout rates in clinical trials compared with placebo groups [92,102].

Complications from acute and chronic exposure to antibiotics are also of concern. For instance, antibiotics, particularly those with broad spectrum activities can cause diarrhea, Candida overgrowth, and C. difficile infection. Emergence of resistant organisms, particularly vancomycin resistant Enterococcus, methicillin-resistant Staphylococcus aureus and multi-drug resistant Enterobacteriaceae is a particularly worrisome consequence of long term antibiotic exposure. In addition, considerable epidemiologic data implicate early life exposure to antibiotics as a risk factor for developing CD but not UC [23,103,104].

Strategies to improve microbial-targeted therapies in IBD

Despite a strong rationale, therapeutic results of using antibiotics, probiotics, prebiotics and FMT for human IBDs have been generally disappointing (Table 1). While it is possible that the observed dysbiosis, enhanced uptake of bacterial products, and increased antimicrobial immune responses in IBDs are secondary phenomena, it is more likely that current therapeutic approaches to restoring normal microbiota are suboptimal. Several strategies can be used to improve current treatments. First, personalized microbial-based therapies based on an individual’s particular intestinal microbiota profile can be designed to reduce/eliminate a person’s pro-inflammatory bacterial, viral or fungal species while simultaneously increasing the number or function of decreased anti-inflammatory species. In addition, a more rational approach to selecting “probiotic” microbial species in IBDs is needed. For instance, microbes with pre-clinical evidence of anti-inflammatory potential in IBDs such as F. prausnitzii and certain Clostridium and Bacteroides species subsets, should be components of the next generation of probiotic therapies [89,105,106]. Second, alternative novel approaches to microbial-based therapeutics should be developed, including administering synthetic bacterial products rather than viable organisms, using recombinant bacteria to secrete protective proteins directly to the distal ileum and/or colon [107], and using bacteriophages in targeted viral therapies.

Table 1.

Summary of Selected Clinical Trials of Microbe-Targeted Therapies in IBDs

	UC		CD		Pouchitis
	Benefit	Trial Design	Benefit	Trial Design	Benefit	Trial Design
FMT	− +	RCT [18] Case Series [14,15]	+	Case Series [17]
Antibiotics
• Ciprofloxacin/Metronidazole	+/−	RCT [44,45]	+ (colonic/perianal)	RCT [26–28]	+	RCT [52,53]
• Rifaximin	−	RCT [57]	+ (colonic)	RCT [30]	−	RCT [54]
• Broad Spectrum	+	RCT [46,48]
Probiotics
• VSL#3	+	RCT [67,68]	−	RCT [74]	+	RCT [80,83]
• Other	−	RCT [66]	−	RCT [71–73]	−	RCT [85]
Prebiotics	−	RCT [93]	−	RCT [92,95]	−	RCT [96]

Conclusions

Great strides have recently been made in understanding the composition and function of the microbes that exist in the gastrointestinal tract. Importantly, we have also begun to recognize how the intestinal microbiome impacts human health and disease. Armed with this information, we are now better positioned than ever before to develop rational strategies that manipulate the microbiota for therapeutic purposes. Huge potential exists to optimize microbial-based therapies of IBDs if we continue efforts to identify specific resident microbes and microbial functions that play pathogenic and protective roles in IBDs and then design effective methods to modulate concentrations and functions of these microbes to suppress inflammation in individuals based on their unique microbial profiles. Such microbial-targeted treatments offer more physiologic approaches to treating IBD patients compared with lifelong potent immunosuppression, though may also be used as adjuncts to immunosuppressive treatments. Furthermore, potential exists for nontoxic approaches to prevent onset of disease in individuals at high risk of developing CD and UC, such as offspring and siblings of IBD patients.