Altered Intestinal Flora Composition and IBS

The effect of gut microflora on gastrointestinal physiology has been most clearly demonstrated in animal experiments under controlled conditions not feasible in human studies. For example, germ-free rats had delayed gastric emptying and intestinal transit, and a prolonged interdigestive migrating motor complex (MMC) as compared with rats with conventional flora. Moreover, introduction of normal gut flora to these germ-free rats normalized their motility. Of interest, when germ-free rats were mono-associated with either Lactobacillus acidophilus or Bifidobacterium bifidum , their small intestine transit accelerated and their MMC frequency increased. Hooper and Gordon profiled gene expression patterns in germ-free mice, and showed reduction of several enteric neuron and intestinal smooth muscle genes. Subsequent mono-association with Bacteroides thetaiotaomicron , a highly adapted and abundant commensal of the human and murine colon, restored the normal expression pattern. These experiments, which involve profound changes in the gut flora of rodents, imply a critical role of the resident flora in establishing and maintaining normal intestinal function, and suggest that changes in the gut microflora can lead to significant alterations in gastrointestinal function.

Changes in gut flora of patients with gastrointestinal disorders, including IBS, have been sought for decades. Efforts have been hampered by (1) disease heterogeneity and multifactorial pathophysiology; (2) studies not controlling for diet and medication use that can influence flora composition; (3) potential fluctuations in stability of gut flora and topographic/geographic variability, both in “normal” and affected subjects; and (4) inherent limitations in methodologies to assess gut flora composition. The last challenge in this area will be overcome by evolving technology.

Although culture of the bowel flora has been the mainstay of evaluating intestinal bacterial composition, the majority of intestinal flora are nonculturable, based on fastidious requirements and limited understanding of the vast expanse of human colonizers. DNA-based strategies such as high-throughput pyro-sequencing are considered more sensitive and accurate, but are still costly and technology intensive. Despite these limitations, culture studies have consistently demonstrated a paucity of Lactobacillus and Bifidobacterium species in the feces of IBS patients compared with controls ( Table 1 ), with the exception of Tana and colleagues who noted increased Lactobacillus . Although the influence of Lactobacillus , Bifidobacterium , and other so-called beneficial bacteria have been studied extensively based on their effects on the epithelium, host immune response, and other factors, this is beyond the scope of this article. However, one finding is notable. Balb/c mice infected with a probiotic L acidophilus strain had elevated expression of several intestinal pain receptors that led to decreased visceral sensitivity.

While these results sparked the use of these specific probiotics in IBS, there were inherent problems with this initial research. The results are difficult to interpret because of the failure of these studies to control for diet. A common finding in the literature related to IBS is the association between IBS and lactose intolerance. The reason for this remains unclear, yet it is recognized that more than 60% of IBS sufferers have dairy intolerance on this basis. Because dairy products are the prebiotic for Lactobacillus and Bifidobacterium species, not accounting for diet, leaves the finding of reduced counts of these organisms possibly secondary to intrinsic diet issues in IBS subjects. The ideal study of this topic would be to put IBS and controls on an identical diet for 2 weeks followed by stool evaluation. This lack of control may explain the overall failure of Lactobacillus -based treatment in IBS, as discussed later.

Recently, more sophisticated techniques have been used to examine subjects with IBS and their fecal content. In a recent study, molecular techniques were used to determine shifts in flora between IBS and controls. In addition to finding differences categorically, subjects with constipation predominant IBS (C-IBS) also appeared to have unique differences in contrast to diarrhea predominant IBS (D-IBS). Specifically, a lack of Lactobacillus and Collinsella species were seen in IBS. Of note, C-IBS subjects had an abundance of Ruminococcus . In D-IBS, a decrease in Bifidobacterium was seen. Even in this sophisticated study, however, diet was not controlled, making interpretation an ongoing issue.

Though not specifically a chronic change in intestinal microflora, acute changes may have an impact on IBS and its development. This process involves the association between IBS and acute gastroenteritis. While this is discussed in detail in this article, animal models used to study PI-IBS further suggest a link between altered gut microflora and IBS. The most characterized postinfectious model of IBS used the organism Trichinella spiralis . This parasitic mouse infection model was found to produce reduced gut motility and increased visceral sensitivity to colorectal distention, and has been likened to IBS. However, the stool flora have not been characterized in this model.

Small Intestinal Bacterial Overgrowth and IBS

SIBO is a situation whereby coliform bacterial counts in the small bowel become excessive. Symptoms of SIBO are similar to IBS. In the last decade, growing data have linked SIBO and IBS. Whereas the initial criticism of the work was a consequence of inaccuracies of breath testing as a means of diagnosing SIBO, recent work has begun to confirm the results of breath testing in IBS, supported by small bowel culture.

As early as 2000, work began to emerge suggesting that subjects with IBS have bacterial overgrowth, based on the lactulose breath test. In this initial study, SIBO was suspect in 76% of IBS subjects and although based on a prospective database, appeared to improve after antibiotic therapy using an open-label approach. In the first follow-up study to this work, a higher rate of positive lactulose breath test results (up to 84%) were identifed. This rate was noted to be far greater than in healthy control subjects. After this work was published there was a high degree of skepticism, due to the complexities of the breath-testing techniques. Now 10 years later, meta-analyses have been conducted that support the breath test findings in IBS compared with controls. In the first of 2 meta-analyses, Ford and colleagues demonstrated that IBS subjects appear to have a higher prevalence of abnormal breath test results in IBS, but only using the most conservative interpretation of the test compared with controls. The second meta-analysis used a different approach based on simply combining the results of studies using breath testing in IBS versus controls in general. This study demonstrated that IBS patients have a greater likelihood of a positive test compared with controls. When only the best studies were used (age- and sex-matched studies), the odds ratio of a positive test in IBS was 9.64 (confidence interval = 4.26–21.82) compared with controls.

Further validation of the SIBO concept in IBS is based on culture and antibiotic trials. In the largest published study of small bowel culture in IBS, aspirates of jejunal fluid in IBS were found to harbor a greater number of coliform bacteria compared with healthy controls (using >5000 coliforms/mL) ( P <.001). Studies of antibiotic response also support SIBO in IBS ( Table 2 ). Controlled trials in IBS Pimentel and colleagues, and functional bloating demonstrate successful treatment of IBS with antibiotics based on this excessive flora. Using breath testing as an outcome measure, antibiotic therapy led to improvement of SIBO, with a 75% improvement in IBS symptoms observed if normalization of the breath test is seen with antibiotics. Another controlled trial demonstrated improvement in IBS symptoms that were sustained for a full 10 weeks of follow-up after cessation of antibiotics. Taken together, these findings strongly support a role for the gut microbiome and perhaps SIBO in the pathophysiology of symptoms in a subset of IBS sufferers.

Evidence suggests that SIBO in IBS may be caused by a deficiency in phase III of interdigestive motor activity. During the fasting state, the small bowel cycles through 3 phases of activity, phases I to III. Phase III is a high-amplitude multiphasic motor event, and an absence or reduced frequency of these contractions is known to induce SIBO. The authors recently demonstrated that IBS patients with SIBO have significantly reduced phase III frequency, suggesting that attenuated gut motility may underlie the development of SIBO in IBS. Although the physiologic basis for this reduced phase III frequency remains unknown, the interstitial cells of Cajal (ICC) are known to be required for normal intestinal motility (including phase III), and their loss interferes with electrical pacemaker activity, slow-wave propagation, and motor neurotransmission in the gut, suggesting that altered ICC function may contribute to altered gut motility in IBS. It has yet to be conclusively demonstrated that changes in ICC are involved in IBS.

Postinfectious IBS

Over the last decade, it has been established that intestinal pathogens play a significant role in the development of IBS. Numerous studies have shown that IBS can be precipitated by an episode of acute gastroenteritis, and that up to 57% of subjects who otherwise had normal bowel function may continue to have altered bowel function for at least 6 years after recovering from the initial acute illness. Based on 2 recent meta-analyses of this research, approximately 10% of subjects who have documented acute gastroenteritis develop IBS, with a summary odds ratio of 6 to 7 for PI-IBS. As gastroenteritis is extremely common, so-called PI-IBS may in fact constitute a large proportion of IBS cases. Thus, reducing risk factors for IBS development after acute gastroenteritis may have an impact on the incidence of IBS.

Although the mechanisms of PI-IBS remain unclear, investigators have identified certain risk factors for the development of IBS after gastroenteritis. The 2 most significant of these are duration/severity of gastroenteritis and female sex. Stress, manifest as recent traumatic life events, and a neurotic personality trait were also predictors of PI-IBS. Evidence of low-grade inflammation is evident in PI-IBS patients. Rectal biopsies demonstrate mildly elevated intraepithelial lymphocytes and enteroendocrine cells that persisted 12 months after Campylobacter jejuni infection. Increased rectal lymphocytes also occur in general IBS patients, but to a lesser degree. Elevated expression of proinflammatory cytokine interleukin (IL)-1β was detected in C jejuni PI-IBS rectal biopsies and in Shigella PI-IBS rectosigmoid and terminal ileum biopsies. Thus, acute gastroenteritis may increase the risk of developing IBS in a susceptible individual through persistent low-grade activation of the gut immune system, or possibly through establishment of an intestinal dysbiosis, defined as an alteration of the composition of the gut flora. Animal infection models of PI-IBS will play a key role in characterizing the mechanistic pathways and underlying alterations in this process.

The discovery of PI-IBS has led to the development of several animal models. The most comprehensively studied model of PI-IBS developed by Canadian researchers uses Trichinella spiralis . This model has now been well characterized. As T spiralis is not a common pathogen in humans and is thus a rare cause of human IBS in the Western hemisphere, other models such as post– C jejuni are being developed. One rat model of C jejuni infection recapitulates many features of human PI-IBS including altered stool form, bacterial overgrowth, and increased rectal lymphocytes, observed 3 months after clearance of the initial acute infection. In fact, some of the first descriptions of PI-IBS in humans stem from C jejuni gastroenteritis.

Antibiotic Treatment of IBS: Support for a Gut Flora Hypothesis

Although antibiotics will be discussed as a therapeutic approach to IBS in a later article, the role of antibiotics is important here because its benefit supports the concept of altered gut microflora. There are now 5 randomized controlled studies examining the effect of antibiotics in IBS, all of which have demonstrated significant improvement in primary outcome measures ; these are summarized in Table 2 . The first of these studies used neomycin. While neomycin demonstrated successful improvement in the primary outcome measure of the study, an important component of the result was that the antibiotic most improved the symptoms when subjects had normalization of their breath test findings. In fact, subjects with complete normalization of their breath test had a near 75% improvement in IBS. This finding is supported by another controlled study by Sharara and colleagues, wherein subjects who were deemed responders to a course of the nonabsorbed antibiotic, rifaximin, had a greater reduction in breath hydrogen, indicating a reduction in intestinal flora. However, the most convincing of all these studies are the 2 latest phase 3 trials (TARGET 1 and TARGET 2). In these studies, rifaximin was effective in improving IBS based on abdominal pain, stool consistency, bloating, and the primary outcome measure of global relief.

Although there is some remaining debate as to why antibiotics help improve IBS symptoms, these antibiotic approaches have provided support for the role of altered gut microflora in IBS.

Probiotics in IBS

If alterations in gut microbiota account for a large fraction of IBS, it seems reasonable that probiotics should restore a “healthy” gut microbiota and alleviate IBS symptoms. Unfortunately, the numerous controlled trials of probiotics in IBS have shown mixed results at best. These studies used a variety of probiotic species and strains, with heterogeneity of dosing regimens and clinical end points (reviewed by Parkes and colleagues ). The data are strongest for Bifidobacterium and Lactobacillus strains. Bifidobacterium infantis 35624, Lactobacillus salivarius UCC4331, or placebo was given to 77 patients and after 8 weeks, the B infantis group had a significant reduction in composite IBS symptom scores and abdominal pain scores versus placebo ( P <.05). In addition, a decrease in the ratio of IL-10/IL-12 cytokine expression in peripheral mononuclear cells suggested an additional anti-inflammatory effect that was not characterized further. No significant benefit was noted with the Lactobacillus strain. A larger multicenter study of 362 women with IBS randomized them to receive B infantis (at a dose of 10 ⁶ , 10 ⁸ , or 10 ¹⁰ CFU daily) or placebo for 4 weeks followed by a 2-week washout. Only the middle dose led to statistically significant but modest improvements in abdominal pain, bloating/distention, and IBS composite scores at the end of treatment. The unexpected dose response may have reflected poor capsule dissolution and subsequent lack of bioavailability of the higher-dose probiotic following ingestion. VSL#3 (a probiotic mixture of 8 species of Bifidobacterium, Lactobacillus, and Streptococcus ) or placebo was given to 25 D-IBS patients for 8 weeks. No difference in the primary end point of global symptom relief or gastrointestinal transit was seen. However, the investigators observed a significant reduction in the secondary end point of abdominal bloating. A larger follow-up study did not demonstrate a significant benefit for bloating, but noted a significant decrease in flatulence during the treatment period.

More recently, 3 meta-analyses or systematic reviews identified a small overall beneficial effect of probiotics over placebo. The investigators found the studies to be heterogeneous, and that funnel plot asymmetry suggested publication bias. However, several of the trials were of good quality, and these tended to show more modest treatment effects. The clinical trials, as well as animal and translational studies with probiotics, indicate that the myriad species and strains of probiotics are clearly unique, with different biochemical and physiologic effects and possibly highly specific interactions with the mucosal immune and neuroendocrine systems. Therefore, the benefits of one probiotic cannot be extrapolated to another strain without thorough studies. Fortunately, the safety profile and adverse event rate of probiotics has been good. Large, well-designed controlled trials are clearly needed to guide the future use of probiotics in treating IBS.