Keywords

The role of bacteria and antibiotics in the pathophysiology and treatment of ulcerative colitis has been postulated for over 60 years [1–3]. The first case reports of the use of antibiotics to treat inflammatory bowel disease (IBD) were published in the 1940s. Multiple studies since then have demonstrated a role for antibiotics in the treatment of perianal Crohn’s disease (CD) and complications of IBD such as peritonitis, abscesses, and bacterial overgrowth [4–6]. The role for antibiotics in the treatment of ulcerative colitis (UC) has not been as clearly delineated.

Dysbiosis and microbiome alterations in the intestinal tract have been appreciated in several diseases, both within the gastrointestinal tract and beyond. The development of early bacterial populations in newborns can be modified by the delivery method at birth and antibiotic exposure, and these changes in bacterial composition may be long lasting [7–9]. Such modifications appear to have a significant impact on the early stages of postnatal immunologic development, potentially predisposing individuals to the development of autoimmune diseases such as type I diabetes, food allergies, and asthma [10–12].

Microbial composition has been demonstrated to play a significant role in IBD and other chronic diseases. Murine studies have demonstrated that mice raised in a germfree environment, i.e., lacking enteric flora, do not develop colitis despite a genetic predisposition to do so. These mice only develop colitis once specific commensal bacteria have been introduced [13, 14]. In humans with colitis, alterations in the location and concentration of intraluminal bacteria in relation to areas of inflammation have been appreciated [15, 16]. Modifications in microbial composition of the enteric microbiome have been demonstrated in several studies of humans with IBD, with increases in Bacteroides, Escherichia coli, and Clostridium species, as well as downregulation of Bifidobacterium and Lactobacillus species [4, 15, 17, 18]. There is a growing evidence demonstrating that these changes in flora may not only differ between different patients and phenotypes of IBD but also may differ significantly in local regions of the bowel in a single patient due to the presence or lack of inflammation in that specific region of the intestine [19, 20]. The composition of the gut microbiome may also have far-reaching implications, with studies suggesting a link between intestinal bacteria and the risk of type I diabetes mellitus, cardiovascular disease, and eczema [21, 22].

Given the growing body of evidence suggesting an effect of dysbiosis of the gut microbiome in IBD and other chronic medical conditions, medical therapy aimed at modulating these bacterial populations may have a significant impact on the resultant disease. Downregulating potentially deleterious organisms while upregulating beneficial strains could affect the degree of inflammation and alter the course of disease. This chapter will explore the current evidence that exists for therapies designed to directly modulate the gut flora in ulcerative colitis (UC), focusing on three classes of such treatment: (1) probiotics, which consist of live bacterial agents used to modify the composition of enteric flora; (2) prebiotics, or oligosaccharides and other compounds designed to affect the growth of particular bacteria within the gut; and (3) antibiotics, or targeted pharmaceutical agents given to either limit the growth or kill specific bacteria.

Probiotics

The hypothesis that specific strains of bacteria could be used to modulate the behavior of other bacteria and provide benefit to the host organism is often attributed to Elie Metchnikoff. In 1907, he proposed that the prolonged life spans of Bulgarians were due to lactic acid and other compounds produced by Bacillus species and other bacteria they consumed in sour milk on a daily basis, citing that these compounds inhibited other bacteria from producing toxins capable of producing “intestinal putrefaction” [23–25]. Despite these observations and those made by several other physicians and scientists of the potential beneficial effect of certain strains of bacteria, it was not until 1965 that the word “probiotic” was first used. This term was first used to describe certain products produced by one strain of bacteria in culture that would promote the growth of another strain of bacteria [26]. This definition has been refined several times since its inception to incorporate the concept that these compounds should consist of living bacteria, with the most widely accepted definition currently being “live microorganisms, which when consumed in adequate amounts, confer a health effect on the host” [27–29].

A number of commensal bacteria have been assessed as potential probiotics. While the initial commentary of prominent luminaries such as Metchnikoff and Tissier promoted the development of many compounds claimed to be probiotics, it has only been over the past 20 years that researchers have attempted to define and purify specific strains of bacteria and test the efficacy of these agents experimentally [24]. The majority of probiotic species used today are lactic-acid-producing strains such as Lactobacillus species, Bifidobacteria species, Enterococcus, Lactococcus, Leuconostoc mesenteroides, Pediococcus acidilactici, Sporolactobacillus inulinus, and Streptococcus thermophilus. Additionally, Escherichia coli Nissle 1917 and Saccharomyces boulardii, a yeast, have also been used as probiotics.

In addition to the individual strains noted above, there have been several studies looking at combinations of several bacteria, with the most studied combination being VSL#3. VSL#3 consists of 8 bacterial strains, including L. plantarum, L. casei, L.acidophilus, L. bulgaricus, B. infantis, B. longum, B. breve, and Streptococcus thermophilus [30].

Probiotic Mechanisms of Action

There are a number of potential mechanisms of action for probiotic bacteria, and different strains have been shown to use different combinations of mechanisms. It is likely that each strain or combination of strains has multiple effects on the epithelial barrier, host immune system, and other bacterial populations within the gut.

Probiotics have been shown to have significant effects on the composition of a host’s microbiome. These commensal organisms are capable of inhibiting the growth of or killing pathogenic bacteria via the production of antimicrobial peptides known as bacteriocins [31]. Both Lactobacillus and Bifidobacteria species have demonstrated direct effects against Salmonella typhimurium via this mechanism [32, 33]. Furthermore, as probiotic commensal populations expand, they can compete with pathogenic strains for various epithelial and mucin binding sites, preventing detrimental local mucosal surface colonization by more harmful, invasive bacteria while promoting the growth of beneficial strains [34, 35]. Lactobacillus species have been shown to increase the biodiversity of not only other Lactobacilli strains in patients with UC but also increase Bifidobacteria strains in neonates 5 days after birth when given to mothers prior to birth [36, 37]. VSL#3 has been shown to increase biodiversity in a DSS-based murine model, independent of its effects on mucosal inflammation or mucin production. This probiotic has also been shown to increase bacterial biodiversity while decreasing fungal biodiversity in patients with UC [38, 39]. It remains unclear whether these modulations in microbiome composition consistently translate into clinically meaningful changes, however [40].

Another potential mechanism of action of probiotic bacteria is direct modification of mucosal immunity via promoting barrier formation, upregulating defensin production, stimulating IgA production, and modulating local cytokine production. Both Lactobacillus and VSL#3 have been shown to promote mucous secretion, which functions as a protective layer against bacterial infiltration [41, 42]. Lactobacillus, E. coli Nissle, and VSL#3 are capable of upregulating genes responsible for defensin production [43–45]. These small peptides have direct antibacterial properties.

Several strains of commensal bacteria have exhibited the ability to directly modulate cytokine production. One study of patients with UC treated with 5-aminosalicylate (5-ASA) and Lactobacillus versus 5-ASA alone showed that 6 weeks of probiotic treatment reduced levels of IL-6, TNF-α, NF-κB, and leukocyte recruitment compared to controls [46]. Bifidobacterium has demonstrated similar effects in UC, reducing TNF-α, IL-8, and NF-κB+ mononuclear cells in colonic biopsies of inflamed mucosa [47].

Probiotics in Ulcerative Colitis

Given the multitude of potential modulatory effects that probiotics have demonstrated in vitro and in murine models, there have been several recent studies ascertaining the clinical effects of these bacteria in patients with IBD. The most extensive research has been done in pouchitis, but there is an expanding literature on the use of these agents in UC as well. Several studies have assessed different probiotic formulations, as well as different doses for these products.

Several agents have been assessed for their ability to induce remission in active UC. VSL#3, one of the more extensively studied probiotics in UC, has demonstrated the potential to induce remission and when used in combination with balsalazide was superior compared to balsalazide or mesalazine alone [48, 49]. In 2009, Sood et al. assessed the efficacy of VSL#3 in inducing remission in patients with mild to moderately active UC in a randomized, placebo-controlled trial [50]. Seventy-seven patients were randomized to the treatment arm, receiving 3,600 billion colony forming units (CFU) per day, compared to placebo. At 12 weeks, 42.9 % of the patients receiving VSL#3 were in remission, compared to 15.4 % of placebo-receiving patients (P < 0.001). 32.5 % of patients in the treatment arm had a UCDAI decrease >50 %, compared to 10 % in the placebo arm. This study did have significant loss to follow-up, and the placebo arm had more patients on azathioprine which may indicate increased severity of disease in the placebo group [51]. In 2010, Tursi et al. conducted a randomized, placebo-controlled trial, using VSL#3 3600 billion CFU per day in 2 divided doses, with a primary end point of reduction of UCDAI >50 % at 8 weeks [52]. 57.7 % (41/71) of the patients in the treatment arm met this primary end point, compared to 39.7 % (29/73) in the placebo group (p = 0.031). Induction of remission at 8 weeks was also assessed, though there was no significant difference between groups for this end point (43.7 % vs. 31.5 %, p = 0.132).

Several studies have also examined the role of Bifidobacterium and other probiotics in inducing remission. Kato et al. examined the efficacy of Bifidobacteria in the induction of remission in 20 patients with mild to moderately active UC, randomizing them to a Bifidobacteria-fermented milk preparation versus placebo. Significant clinical improvement occurred in both the treatment group and placebo group, though significant endoscopic improvement occurred only in the treatment arm [53]. Furrie et al. also examined the efficacy of Bifidobacteria in combination with a prebiotic compared to placebo in 18 patients with active UC for 1 month. While reductions in sigmoidoscopic scores were appreciated in the treatment arm, they were not statistically significant [54]. A newer agent, BIO-THREE, which contains Streptococcus faecalis, Clostridium butyricum, and Bacillus mesentericus, was recently assessed in a small case series of 20 UC patients, demonstrating induction of remission in 9 of 20 patients and improvement in UCDAI in an additional 2 of 20 patients [55]. While promising, this agent will need further evaluation in prospective, randomized, placebo-controlled studies.

There have also been multiple systematic reviews of studies of probiotics for induction of remission. A Cochrane analysis in 2007 showed no evidence that probiotics were superior to ASA compounds or placebo for induction of remission. This pooled analysis was conducted prior to the Tursi and Sood studies, however [56]. A recent meta-analysis by Sang et al., published in 2010, assessed 13 randomized controlled trials involving several different preparations of probiotics, including E. coli Nissle, Bifidobacterium, Lactobacillus GG, VSL#3, and a combination product containing both Bifidobacterium and a prebiotic oligosaccharide called synbiotic [57]. The authors found no significant improvement in remission rate (OR 1.35, 95 % CI 0.98–1.85), though there was significant heterogeneity. The authors then performed an analysis stratified by probiotic, and neither E. coli, Bifidobacterium, nor VSL#3 demonstrated statistically significant improvement. However, This analysis also did not include the most recent study of VSL#3 by Tursi et al. The authors concluded that there was no significant benefit for using probiotics in inducing remission.

There have also been several studies assessing probiotics for the maintenance of remission as well. Kruis et al. examined the efficacy of E. coli Nissile 1917 for maintenance of remission compared to mesalazine and found them to have equivalent efficacy [58]. Kruis later assessed E. coli Nissile 1917 versus mesalazine in a larger randomized controlled trial of 327 patients with UC over 12 months. At the end of the study period, intention to treat analysis demonstrated that 45.1 % of patients receiving E. coli relapsed, compared to 37.0 % in the mesalazine group, with significant equivalence between the two groups [59]. Rembacken also examined the ability of E. coli to maintain remission in those who had successfully entered remission on prior E. coli therapy versus those who had entered remission on mesalazine. The probiotic preparation maintained remission in 67 % of patients, compared to 73 % who maintained remission with mesalazine [60]. These results were not significantly different. Lactobacillus GG has also been assessed in comparison to 5-ASA in prevention of relapse, without significant difference between groups [61]. Bifidobacterium was also assessed for maintenance of remission in two small studies published in 2004, with both studies demonstrating potential benefit for these agents [62, 63]. Given the small sample size, further research regarding this agent is required.

A recent systematic review has attempted to synthesize these results for maintenance of remission. Sang et al. assessed eight randomized controlled trials involving several different probiotics [57]. The authors found no significant improvement in prevention of relapse (OR 0.69, 95 % CI 0.47–1.01). Yet as with the induction trials, there was profound heterogeneity in both of these analyses. When assessing only placebo-controlled studies, there was a significant benefit of probiotic therapy in maintaining remission, with a remission ratio of 0.25 (95 % CI 0.12–0.51); there was no significant heterogeneity in this subgroup. Non-placebo-controlled trials did not demonstrate statistically significant benefit. Overall, the authors concluded that probiotics potentially provide benefit in the maintenance of remission. The degree of heterogeneity and variety of methods, agents, controls, and study duration in this meta-analysis make interpretation difficult.

In summary, with growing information on the impact of intestinal flora in disease activity and laboratory data demonstrating the beneficial effects of several specific strains of commensal bacteria, probiotics represent a potential adjunct to the current armamentarium in IBD. The data on efficacy of inducing and maintaining remission in UC has been mixed, although several recent placebo-controlled trials of VSL#3 have demonstrated benefit. A meta-analysis for induction of remission, which does not include two of the most recent trials and includes a heterogenous pool of 11 studies, has not confirmed these results. However, Another meta-analysis for maintenance of remission did appreciate a benefit when isolating only placebo-controlled trials. Further randomized controlled trials are still needed to confirm the efficacy of various probiotic preparations such as VSL#3 and Bifidobacterium in ulcerative colitis. Table 18.1 lists the probiotic clinical trials for induction and maintenance of remission in ulcerative colitis.

Table 18.1 Probiotic clinical trials for induction of remission and maintenance of remission in ulcerative colitis
Full size table

Probiotics in Pouchitis

There have been several studies assessing the efficacy of probiotics in pouchitis. Pouchitis occurs in up to 45 % of patients after proctocolectomy and is thought to be secondary to alterations of the luminal flora in the pouch [64]. This hypothesized pathophysiology has made pouchitis an attractive candidate for probiotic therapy. In 2000, Gionchetti et al. published a randomized, double-blind, placebo-controlled trial assessing VSL#3 at 900 billion CFUs twice daily in the maintenance of remission of pouchitis in UC [30]. Forty patients in clinical and endoscopic remission were enrolled in the trial and followed for 9 months. Fifteen percent of patients (3 of 20) in the VSL#3 arm and 100 % (20/20) of the placebo arm relapsed. Gionchetti also demonstrated that the same dose of VSL#3 was capable of preventing onset of pouchitis after surgery in 2003 [65]. In 2004, Mimura et al. were able to demonstrate that a once daily dose of 600 billion CFUs of VSL#3 maintained remission in 85 % of pouchitis patients, compared to 5 % in the placebo arm [66]. A meta-analysis published in 2007 assessed 5 randomized controlled trials of probiotics in pouchitis. This study demonstrated an overall OR of 0.04 (95 % CI 0.0–0.14, p < 0.0001). There was significant heterogeneity between trials and variability in probiotics used, with one trial using Lactobacillus rhamnosus GG, while the other four used VSL#3 [67].

Prebiotics and Synbiotics

The term “prebiotic” was coined by Glenn R. Gibson and Marcel B. Roberfroid in 1995 as “nondigestible food ingredients that beneficially affect the host by selectively stimulating the growth and/or activity of one or a limited number of bacterial species already resident in the colon, and thus attempt to improve host health” [68, 69]. This definition was refined in 2007 by Roberfroid to “a selectively fermented ingredient that allows specific changes, both in the composition and/or activity in the gastrointestinal microflora that confers benefits upon host well-being and health” [70]. Combining a prebiotic and probiotic in the same preparation is considered a “synbiotic” [69]. Such combinations are thought to enhance colonization, survival, and function of the probiotic species.

Prebiotics typically consist of oligosaccharides and polysaccharides that cannot be digested by the human host but can be digested by specific bacteria in the gut, providing them with a selective advantage. To be considered a prebiotic, a compound must be completely resistant to the host digestive tract, including gastric acid, host hydrolytic enzymes, and direct absorption. The compound must then be fermentable by host bacteria, resulting in stimulation of specific commensal bacteria. Two compounds that have been extensively researched and meet these criteria are inulin and trans-galactooligosaccharides (TOS) [70]. The bacterial “targets” of these agents are typically the same bacteria delivered in common probiotic formulation. When added to both pure strains of colonic flora and cultured human feces, inulin has been shown to selectively promote the growth of Bifidobacterium and may even inhibit the growth of other species such as C. perfringens and E. coli in mixed culture [71]. Furthermore, the end products of fermentation of these sugars include short chain fatty acids (SCFAs), which are an energy source of colonic enterocytes [69].

Research into the effects of both prebiotics and synbiotics in human disease, particularly with regard to disorders of the gastrointestinal tract, has begun. Inulin, oligofructose, and TOS have been assessed in the management of constipation, which is thought to be secondary to dysbiosis. A review by Macfarlane published in 2007 assessed 7 trials of various types and doses of prebiotics, with only two demonstrating a statistically significant improvement in stool output [72]. Further research has demonstrated a potential role for fructooligosaccharides (FOS) in a placebo-controlled, randomized trial, though results did not reach statistical significance [73]. Additional research demonstrated potential improvement in some symptoms in IBS with administration of TOS as well, though further research is required [74].

With regard to ulcerative colitis, there have been several animal models suggesting efficacy, but there is limited human data. The effects of a wide range of agents, including FOS, inulin, lactulose, or combinations thereof, have demonstrated efficacy in increasing the quantity of Bifidobacterium and Lactobacillus species in several animal models of colitis, as well as modulating inflammatory markers [72, 75–79]. Controlled trials in humans are limited, however. Furrie et al. conducted a small randomized, placebo-controlled trial in 18 patients of a 1-month course of a synbiotic containing Bifidobacterium longum and a combination of inulin and oligofructose [54]. Patients were assessed before and after therapy via clinical index, endoscopic score, and several immunologic markers such as defensin excretion, TNF-α, IL-1α, and IL-10. After therapy with the synbiotic, all patients had a significant reduction in defensins, TNF-α, and IL-1α. There was also a 42-fold increase in concentration of Bifidobacterium on mucosal biopsies, determined via rRNA, after therapy. Histologically, there was also reduced inflammation in those in the treatment arm as well as reduced clinical symptoms. Statistical significance was not reported for these outcomes, however. Fujimori et al. conducted a 3-armed trial of a synbiotic (Bifidobacterium and psyllium) versus probiotic alone versus prebiotic alone in 120 patients with mild UC or in remission for 4 weeks [80]. Only the synbiotic group appreciated an improvement in IBDQ, a validated questionnaire of IBD symptoms and quality of life.

In summary, prebiotics and synbiotics represent a new method of modifying the microbiome, promoting the growth of potentially beneficial commensal and probiotic strains. There is a small but growing body of literature of the effect of these oligosaccharides on microbiome composition and their ability to modulate inflammation. There are also several small, randomized controlled trials, but much more research is needed to assess the efficacy of these agents.

Antibiotics

As previously noted, the first publications of antibacterial agents being used to treat IBD were published in the 1940s [1–3]. However, the role of antibiotics in the pathogenesis and treatment of IBD has become considerably more complex since these early studies. Recent research has demonstrated that antibiotic exposure has been shown to have a significant and long-lasting effect on microbiome composition in neonates and infants [81, 82]. Amoxicillin can markedly reduce Lactobacillus species in the gut after administration, and this has been shown to have a significant effect on developmental gene expression in enterocytes [83]. Antibiotic-related dysbiosis has also been shown to create a permissive environment for several invasive, pathogenic strains of bacteria, including Clostridium difficile, Clostridium perfringens, Salmonella species, and E. coli O157:H7 [84–86]. Promotion of these species may exacerbate IBD-related inflammation.

There is intriguing new data that suggests antibiotic exposure may increase the risk of later developing IBD. Margolis et al. performed a retrospective study in The Health Improvement Network database in the UK, assessing 94,487 patients with acne for exposure to tetracycline antibiotics. Tetracyclines are frequently used in the treatment of acne, and this class includes drugs such as minocycline, tetracycline, oxytetracycline, and doxycycline. The authors detected an increased risk of developing IBD with exposure to any of these antibiotics, with a hazard ratio (HR) of 1.39 (95 % CI 1.02–1.90). When stratified by IBD subtype and antibiotic, tetracycline/oxytetracycline remained associated with an increased risk of CD, while no antibiotics maintained significance for UC [87]. Further epidemiologic and animal-based research is needed to explore this potential relationship between antibiotic exposure and risk for developing IBD.

Once IBD has developed, antibiotic exposure may actually have a protective effect. A recent population-based cohort study using the General Practice Research Database (GPRD) in the UK assessed this effect [88]. The authors studied 1,205 patients with CD and 2,230 patients with UC, with a median of approximately 4 years’ follow-up time for each group. In this cohort, exposure to antibiotics was associated with an overall reduced risk of disease flare for CD, with an OR of 0.78 (95 % CI 0.64–0.96), but this association was not present for UC. This protective effect was strongest with more recent exposure, suggesting that the acute changes in the microbiome may be responsible.

Antibiotics in the Management of Ulcerative Colitis

There have been a multitude of studies looking at the role of antibiotics in the treatment of IBD. The two most commonly used classes of antibiotics are the fluoroquinolones, such as ciprofloxacin and levofloxacin, and the nitroimidazoles, including metronidazole. The combination of these two classes of antibiotics provides broad-spectrum coverage against most enteric bacteria, with the fluoroquinolone providing coverage against gram-negative and gram-positive aerobes and metronidazole covering gram-negative and gram-positive anaerobes [4]. Both classes are typically well tolerated, although fluoroquinolones can cause nausea, vomiting, abdominal pain, diarrhea, lightheadedness, photosensitivity, and an increased risk of tendon rupture. Side effects due to metronidazole include dysgeusia, resulting in a metallic taste. It has also been associated with nausea, vomiting, diarrhea, abdominal cramping, and a disulfiram-like reaction when combined with alcohol. Another common, though more serious, complication of metronidazole is peripheral neuropathy. The risk of this side effect appears to increase with prolonged exposure and increasing dose. It typically resolves upon cessation of the drug but may persist. A newer agent that has been assessed in several recent trials is rifaximin. This nonabsorbable rifamycin derivative is a nonabsorbable antibiotic with broad-spectrum coverage against gram-positive and gram-negative aerobes and anaerobes and is also well tolerated.

There appears to be an established role for antibiotic therapy in pouchitis. A small, randomized controlled trial by Madden et al. examined the benefit of metronidazole versus placebo in pouchitis in 1994. The authors appreciated a statistically significant decrease in the number of bowel movements, but no significant endoscopic or histological changes [89]. Another study compared metronidazole and budesonide enemas for a total of 6 weeks in active pouchitis, and a clinical improvement was appreciated in both groups, but there was no difference between the two groups [90]. Shen et al. performed a randomized trial in 2001 comparing ciprofloxacin to metronidazole, demonstrating a greater reduction in Pouchitis Disease Activity Index (PDAI) in the ciprofloxacin group. Ciprofloxacin was also better tolerated, with 33 % of patients experiencing adverse effects in the metronidazole group [91]. Another study looking at flora changes related to pouchitis suggested that more complete eradication of pathogenic C. perfringens and E. coli strains with ciprofloxacin may be responsible for the observed improvement in efficacy compared to metronidazole [92]. Mimura et al. also performed an open-label trial assessing the efficacy of combining both metronidazole and ciprofloxacin for refractory pouchitis [93]. Eighty-two percent (44 of 36) of their patients entered remission, with a significant decrease in median PDAI from 12 to 3 after therapy. The therapy was well tolerated. Rifaximin has also been assessed in open-label trials, either alone or in combination with other antibiotics [4, 94]. A recent case series demonstrated a reduction in PCDAI in 16 of 18 patients, with 6 patients entering remission. However, in a recent small, randomized, double-blind, placebo-controlled trial by Isaacs et al., rifaximin provided no benefit over placebo [95]. Based on this evidence, the American College of Gastroenterology currently recommends either metronidazole or ciprofloxacin for the treatment of pouchitis [96].

The data are less clear regarding the potential benefit of antibiotics in inducing remission in UC. As is the case with pouchitis, multiple antibiotics and combinations of antibiotics have been assessed for efficacy in active UC, with most studies focusing on ciprofloxacin or combination therapy. A number of other agents, such as tobramycin, oral vancomycin, or rifaximin, have been assessed, with mixed results. With regard to ciprofloxacin, there have been several placebo-controlled trials. In 1997, Mantzaris et al. performed a randomized, placebo-controlled trial of a 2-week course of oral ciprofloxacin versus placebo in 70 patients with mild to moderately active UC in addition to 5-ASA and prednisolone [97]. No significant difference in response was appreciated between groups. Mantzaris also conducted a study in 55 patients with severe UC, examining the effects of IV ciprofloxacin versus placebo in addition to IV steroids and parenteral nutrition. IV ciprofloxacin provided no additional benefit [98]. In one of the few positive studies, Turunen et al. assessed 6 months of ciprofloxacin versus placebo in conjunction with 5-ASA and steroids for maintenance of remission of subjects with moderate to severe active UC. At 6 months, 79 % of patients in the ciprofloxacin group had maintained an initial response, compared to 56 % in the placebo group (p = 0.02) [99].

Several other antibiotic-based therapies have been assessed in randomized controlled trials of induction of remission in UC. Burke et al. performed a randomized controlled trial of oral tobramycin versus placebo in mild to severely active UC, in conjunction with steroid therapy, with 31 of 42 (74 %) patients achieving complete symptomatic remission compared to 18 of 42 (43 %) in the placebo arm (p = 0.008) [100]. Several combinations of antibiotics have been assessed as well. Mantzaris et al. assessed the combination of IV metronidazole and tobramycin versus placebo, along with parenteral nutrition and steroids, in 39 patients with acute severe UC. Sixty-three percent in the treatment arm and 65 % in the placebo arm noted significant improvement [101]. Rifaximin was assessed by Gionchetti et al. in a small placebo-controlled trial. Rifaximin 400 mg twice daily demonstrated significant decreases in clinical activity, with 9 of 14 patients receiving rifaximin demonstrating benefit compared to 5 of 12 receiving placebo [102].

Recent studies of antibiotic therapy have considered targeting specific organisms. In 2005, Okhusa et al. published a randomized controlled trial of a regimen specifically targeting Fusobacterium varium, containing amoxicillin, tetracycline, and metronidazole (ATM) versus placebo for 2 weeks in 20 patients with mild to moderately active UC [103]. At 3–5 months, the authors appreciated a statistically significant reduction in endoscopic score, but not histology or symptom index, in the treatment group. At 12–14 months after therapy, there were significant reductions in endoscopic score, symptom index, and histological grading. The same group performed a placebo-controlled, randomized trial of 2 weeks of oral ATM in 206 patients with mild to severe chronic relapsing UC [104]. The authors appreciated a greater clinical and endoscopic response at 3 months in the treatment group compared to placebo, though remission rates were not significantly different. Interestingly, this 2-week course of antibiotics improved clinical, endoscopic, and remission rates at 12 months in the treatment arm compared to placebo. While such targeted approaches are compelling, further research is required to ascertain the exact effects such broad-spectrum therapies are having on the microbiome and clinical outcomes of UC patients.

There have also been several meta-analyses of the efficacy of antibiotics in UC. Rahimi et al. published a meta-analysis including ten randomized, placebo-controlled trials of antibiotics in addition to steroids for induction of remission in active UC [105]. Disease severity was not reported. Antibiotics assessed included vancomycin, metronidazole, tobramycin, ciprofloxacin, rifaximin. Two studies of the studies in the meta-analysis evaluated combinations of antibiotics for treatment of UC. Overall, there did appear to be a statistically significant benefit for antibiotic use in active UC, with an OR of 2.14 (95 % CI 1.48–3.09), without significant heterogeneity or detected publication bias. Khan et al. performed another meta-analysis in 2011, examining 9 trials including 662 patients for induction of remission of active UC. Of note, 7 of these trials were also included in the meta-analysis conducted by Rahimi et al. Those studies that commented on UC disease severity were typically moderate to severe. In this study, an overall benefit for antibiotic therapy was appreciated as well, with an Odds RatioR of not being in remission of 0.64 (95 % CI 0.43–0.96). They did detect moderate heterogeneity as well as possible publication bias. Of note, the authors rigorously reviewed the quality of these studies as well and found that only one of the 7 trials included for UC had a low risk of bias [106].

There is currently limited data regarding the role of antibiotics in the maintenance of remission in UC. Lobo et al. performed a long-term follow-up of active UC patients who had received tobramycin and entered remission. In this trial, antibiotics were only given for a single 1-week period during the induction of the remission phase [107]. When examining those who were in remission, there was no significant difference between groups regarding relapse rates at 1 and 2 years. The previously mentioned trial by Turunen et al. which examined the benefits of ciprofloxacin for 6 months in active UC also reported the rates of relapse in those that responded from 6 months to 1 year after cessation of study medication, demonstrating similar failure rates in the treatment arm (9 of 30, 30 %) as in the placebo arm (7 of 25, 28 %) [99]. Both of these trials demonstrate failure rates after antibiotic cessation, however; a paucity of data regarding continued therapy may represent wariness in using long-term antibiotic therapy.

In summary, there is ample evidence demonstrating that antibiotic therapy has a significant impact on the microbiome of the intestinal track. Furthermore, there is a growing body of literature suggesting that antibiotic exposure may have an effect on the risk of developing IBD, and once diagnosed with IBD, antibiotic exposure may significantly impact the course of disease. It also appears that there is benefit to antibiotic therapy for treatment of IBD-related complications and for pouchitis. However, there is currently mixed evidence with regard to the efficacy of antibiotic therapy for the treatment of UC. There is considerable variation in efficacy in treatment based on the specific antibiotic, number of antibiotics used, and duration of treatment. Based on these data, more research is required before antibiotic therapy can be formally recommended for the management of ulcerative colitis in the absence of peritonitis, abscess, or toxic megacolon. Table 18.2 shows antibiotic clinical trials for induction and maintenance of remission in UC.

Table 18.2 Antibiotic clinical trials for induction and maintenance of remission in ulcerative colitis
Full size table

Summary

There is growing evidence that dysbiosis plays a significant role in the pathogenesis of ulcerative colitis. As such, efforts to modify the composition of a patient’s microbiome represent an attractive adjunct to the current therapeutic options in UC. In this chapter we explored the evidence for several different classes of agents designed to alter the microbial composition of a patient’s enteric flora. Probiotics, which are living organisms ingested by a patient, can introduce commensal organisms into a patient’s GI tract. Several different agents exist in this class, and there appears to be evidence of possible benefit in UC. Prebiotics promote the growth of beneficial commensal bacteria. The effects of these agents require significantly more research before specific recommendations can be made. Antibiotics are pharmaceutical agents designed to kill or halt the growth of existing bacteria and have a long history of efficacy in treating intraperitoneal complications of CD and UC. There is conflicting data supporting their use in UC. As such, they are not currently used in UC, except empirically in the setting of severe, fulminant colitis.