TAG»
Use of probiotics in gastrointestinal disorders: what to recommend?
Elizabeth C. Verna
Division of Digestive and Liver Diseases, Columbia University College of Physicians and Surgeons, New York, NY, USA
Susan Lucak
Division of Digestive and Liver Diseases, Columbia University College of Physicians and Surgeons, 622 West 168th Street, PH 20, New York, NY 10032, USA. Email: sll1@columbia.edu
Abstract
Perturbation of bacterial microflora of the gastrointestinal (GI) tract may play an important role in the pathophysiology of some GI disorders. Probiotics have been used as a treatment modality for over a century. They may restore normal bacterial microflora and effect the functioning of the GI tract by a variety of mechanisms. Probiotics are not currently regulated and only few randomized controlled trials exist investigating their efficacy in different GI disorders. They are available in a variety of formulations and delivery systems making interpretation and comparison of studies even more difficult. The efficacy of probiotics, either as a single strain or a combination of probiotics, has been tested in antibiotic-associated diarrhea, Clostridium difficile colitis, infectious diarrhea, ulcerative colitis, Crohn’s disease, pouchitis, and irritable bowel syndrome, among other disorders. Results of the studies are reviewed in this article and recommendations for probiotic use in these disorders are made. Although probiotics appear to be generally safe in an outpatient setting, the situation may be different in immunocompromised, hospitalized patients who may be at a greater risk of developing probiotic sepsis. No studies exist addressing the issue of safety specifically. Many questions regarding use of probiotics in GI disorders remain to be answered in future studies, such as most optimal doses, duration of treatment, physiological and immunological effects, efficacy of specific probiotics in specific disease states, and safety in debilitated patients.
Keywords: antibiotic-associated diarrhea, Clostridium difficile colitis, Crohn’s disease, irritable bowel syndrome, pouchitis, probiotics, review, ulcerative colitis
Introduction
Probiotics are being used with increasing frequency as a treatment for several medical conditions, such as allergic diseases (atopic dermatitis, possibly allergic rhinitis), bacterial vaginosis, urinary tract infections, and prevention of dental caries or respiratory infections. Probiotics are used as a treatment for a variety of gastrointestinal (GI) disorders. In this review, the historical perspectives, proposed mechanisms of action, formulations and delivery systems, safety, and specific GI disorders for which probiotics have been used are discussed.
Historical perspectives
Probiotics have been used therapeutically for many centuries in different parts of the world for their contribution to longevity and digestive health. The World Health Organization has defined probiotics as ‘live organisms which when administered in adequate amounts confer a health benefit on the host’. Categories of probiotics in use today include: bacteria such as lactic-acid bacteria (LAB) and Escherichia coli strains (such as E. coli Nissle 1917), as well as yeast species including most prominently Saccharomyces boulardii among others (Table 1). Prebiotics such as lactulose, inulin, psyllium, and other oligosaccharides (found in onions, garlic, asparagus, leeks, artichoke, bananas, tomatoes, wheat, oats, soy beans, and other plants) are nondigestible food ingredients that stimulate the growth or activity of bacteria in the GI tract which are beneficial to the health of the body [Grajek et al. 2005]. Synbiotics are a combination of a prebiotic and a probiotic, such as inulin and Lactobacillus rhamnosus GG or Bifidobacter longum. Antibiotics, in contrast, are compounds that kill or inhibit the growth of bacteria.
Probably the first person of Western medicine to publish on the topic of probiotics in the early 20th century was the Russian Nobel Prize winner Ilya Metchnikoff, when he described longevity in people in Eastern Europe who lived largely on milk fermented by LAB. He theorized that proteolytic microbes in the colon produced toxic substances responsible for the aging process and proposed that consumption of fermented milk would coat the colon with LABs, decreasing intestinal pH, suppressing proteolytic bacteria and thus leading to slowing of the aging process [Gordon, 2008]. Metchnikoff and his followers ingested milk fermented with this ‘Bulgarian Bacillus’ and reported health benefits [Vaughan, 1965].
In 1917, during World War I, Alfred Nissle isolated a strain of E. coli from the feces of a soldier who did not develop enterocolitis during a severe outbreak of shigellosis. Nissle used the E. coli strain with considerable success in acute cases of infectious intestinal diseases such as salmonellosis and shigellosis [Nissle, 1959]. E. coli Nissle 1917 is still in use today and is one of the few examples of a non-LAB probiotic.
Researchers and clinicians have studied and used probiotics in a variety of medical conditions. In the last decade, over 5000 articles were published in the medical literature. Furthermore, the use of probiotics has surged dramatically as a result of direct-to-consumer marketing as probiotics are not regulated.
Mechanisms of action
The GI tract plays an important role as an interface between the host and the environment. It is colonized by about 10 trillion microbes of many different species, amounting to 1–2kg in weight [O’Hara and Shanahan, 2006]. Only a minority (300–500) of these species can be cultured in vitro and studied [O’Hara and Shanahan, 2006]. Intestinal epithelial cells have the capacity to distinguish pathogenic from nonpathogenic bacteria on the basis of their invasiveness and the presence of flagella, although the exact mechanisms that allow them to do this have not been elucidated fully [Borchers et al. 2009].
The precise mechanism(s) of action of probiotics has not thus far been clarified. Potential mechanisms to consider include: (1) modulation of GI immunity by altering inflammatory cytokine profiles and downregulating proinflammatory cascades or inducing regulatory mechanisms in a strain-specific manner; (2) displacement of gas-producing, bile salt-deconjugating bacterial species and thus possibly inhibiting pathogenic bacterial adherence; (3) alteration of bacterial flora by acidification of the colon by nutrient fermentation; (4) enhancement of epithelial barrier function; (5) induction of µ-opioid and cannabinoid receptors in intestinal epithelial cells; (6) reduction of visceral hypersensitivity, spinal afferent traffic, and stress response [Borchers et al. 2009; Lin et al. 2008; Vanderpool et al. 2008; Lawton et al. 2007; Quigley and Flourie, 2007; Rousseaux et al. 2007; Yan et al. 2007; Focareta et al. 2006; Makras et al. 2006; Roselli et al. 2006; Candela et al. 2005; Collado et al. 2005, 2007; Cotter et al. 2005; Matsumoto et al. 2005; Paton et al. 2005; Sherman et al. 2005; Smits et al. 2005; Sturm et al. 2005; Hart et al. 2004; Mukai et al. 2004; Pathmakanthan et al. 2004; Servin, 2004; McCarthy et al. 2003; Pena and Versalovic, 2003; Borruel et al. 2002].
Probiotics are available in a wide variety of formulations ranging from tablets and powders to yogurts, milk, and juices. Physicians tend to recommend tablets and powders; other formulations are heavily promoted by direct-to-consumer marketing.
To qualify as a probiotic, certain criteria need to be met: a bacterial strain must be fully identified, be safe for ingestion, adhere to the luminal mucosa, colonize the gut, and possess documented health benefits (Table 2). A probiotic should be delivered in a formulation that is stable when stored. The colony number of bacteria and viability need to be reliable and they must survive the acid and bilious environment in the upper GI tract before they reach the small intestine and colon. Since the quality and content of probiotics have not been regulated, it is difficult to accurately assess their efficacy and safety.
Many questions related to efficacy, viability, most optimal dose, and method of delivery remain:
- The optimal number of colony forming units (CFUs) for each bacterial strain delivered remains unknown. Doses in human trials are based on those used in animal studies despite the differences in intestinal surface area. Dose–response studies are generally lacking. Commercially available probiotic formulations typically have at least 106CFUs, but they may range up to 1012CFUs.
- Very few studies have actually documented survival of an administered probiotic as it transits the gut, by means of fecal recovery studies. One probiotic may not necessarily be translatable to other probiotic(s): for example, different Bifidobacterium species have different tolerances to acid and growth requirements and will have different fecal recovery rates [Matto et al. 2004; Takahashi et al. 2004].
- The method of delivery, i.e. yogurt versus milk, may have an impact on the viability and number of bacterial colonies. Furthermore, only one strain of B. longum could survive in fermented milk for 2 weeks [Takahashi et al. 2004].
- Probiotics may produce their effects with viable as well as nonviable bacteria, suggesting that metabolic or secreted factors or structural or cellular components may mediate their immunomodulatory activities [Borchers et al. 2009]. Furthermore, several experiments indicate that the ability to induce secretion of various cytokines is mediated by and large by cell wall components [Borchers et al. 2009].
- Different probiotic species and genuses may have different immunological and physiological effects in different disease states. Wagner and colleagues showed that different Lactobacillus species have different efficacy in preventing fungal sepsis in mice [Wagner et al. 1997].
- The composition of colonic bacterial microflora appears to change with aging (age >60 years). It is unknown whether elderly patients should be treated with different probiotics than young patients [Enck et al. 2009].
- Combination probiotics may interact and have an impact on host intestinal flora differently than single probiotic preparations.
- Optimal duration of probiotic treatment and durability of response are unknown. How long a given probiotic will take to colonize, alter the microflora, and have an impact on immune function remains uncertain. There is significant heterogeneity in treatment duration in the human studies, likely contributing to the differences in reported results.
It is not, therefore, possible to extrapolate the results of one study with one species of a probiotic, one dose, and one formulation in one disease state to other probiotic(s) as a whole. It is for these reasons that clinical trials using probiotics have yielded very inconsistent and difficult to interpret data.
Safety
Probiotics have been consumed by humans in one form or another for over 100 years, with a good safety record generally. A Finnish epidemiological study has shown no increase in Lactobacillus infections in healthy individuals in areas with documented large rises of use of Lactobacillus-containing products [Saxelin et al. 1996]. Probiotic supplementation has been studied in healthy volunteers, and the data suggest that several probiotic strains may enhance nonspecific immune responses, but the effects on adaptive cellular and humoral immune responses appear to be negligible [Borchers et al. 2009].
Questions and concerns have been raised, however, about the safety of probiotic administration in the setting of a severe illness. Probiotic sepsis is the most feared complication related to probiotic administration [Boyle et al. 2006]. Lactobacillus is a rare but documented cause of endocarditis in adults [Cannon et al. 2005]. There are several reports in the literature of bacteremia in adults and children in the setting of probiotic administration [De Groote et al. 2005; Land et al. 2005; Kunz et al. 2004, 2005; Mackay et al. 1999; Rautio et al. 1999]. In addition, several cases of Saccharomyces boulardii fungemia have been reported in the literature [Cherifi et al. 2004; Henry et al. 2004; Cassone et al. 2003; Lestin et al. 2003; Riquelme et al. 2003; Lherm et al. 2002; Cesaro et al. 2000; Hennequin et al. 2000; Perapoch et al. 2000; Rijnders et al. 2000; Niault et al. 1999; Bassetti et al. 1998; Fredenucci et al. 1998; Pletincx et al. 1995], including two series in which the fungi spread to neighboring patients who were not taking the probiotic [Cassone et al. 2003; Perapoch et al. 2000]. This spread was thought to be due to contamination of central catheters in patients who had intestinal surgery (jejunostomy) or chronic illnesses (valvular heart disease), and who were immunocompromised. Only one case of probiotic sepsis was thought to have been directly fatal [Lestin et al. 2003]. A randomized, double-blind, placebo-controlled trial was performed using probiotic prophylaxis (six different strains of viable bacteria: L. acidophilus, L. casei, L. salivarius, L. lactis, B. bifidum, and B. lactis) in a total daily dose of 1010 bacteria orally twice daily for 28 days in patients hospitalized with severe acute pancreatitis. This showed no decrease in infectious complications but increased mortality (16%) in the probiotics group in comparison with the placebo group (6%, relative risk [RR] 2.53, 95% confidence interval [CI] 1.22–5.25). Nine of the 152 patients in the probiotics group developed bowel ischemia, eight of whom died, compared with none in the placebo group [Besselink et al. 2008].
Other safety concerns relate to the unpredictability of immune modulation through change in intestinal flora in certain disease states. For example, worsening of Crohn’s disease (CD) in patients taking some probiotic formulations [Rolfe et al. 2006] or exacerbation of indomethacin-induced enteropathy in animal models by Lactobacillus GG [Kamil et al. 2007]. As rare as these complications appear to be, probiotic safety profile needs to be specifically studied, particularly in hospitalized patients. There are no formal clinical trials assessing the safety of probiotics as there are safety data on regulated medications. At this time, we can only rely on case reports, which is without a doubt suboptimal.
Gastrointestinal disorders for which studies of probiotics have shown some benefit
Ingested probiotics are thought to alter deranged bowel flora or to change patients’ tolerance to their own commensal flora and thus play a role in the pathogenesis of many GI disorders. The following disorders have been most commonly studied with regard to probiotic interventions.
Antibiotic-associated diarrhea (AAD) may be prevented by coadministration of probiotics, as suggested by several randomized controlled trials (RCTs). Several comprehensive meta-analyses, recently published, all show that probiotics significantly decreased incidence of AAD (RR 0.39–0.43) [McFarland, 2006; Szajewska and Mrukowicz, 2005; Cremonini et al. 2002; D’Souza et al. 2002]. The effects were similar across all categories and formulations of probiotics and treatment durations. The most commonly used probiotics were S. boulardii, LABs, and several combinations of LABs, given in doses from 107 to 1011, for durations of 5–49 days, generally paralleling the duration of antibiotic therapy. One of the meta-analyses found that S. boulardii, L. rhamnosus, and multiple mixtures of two different probiotics were the most protective against AAD [McFarland, 2006]. Other specific preparations have been studied to a lesser extent and that may be why their efficacy has been found to be less significant. One randomized, double-blind, placebo-controlled trial in individuals over the age of 50 using combination L. casei, L. bulgaricus, and S. thermophilus twice daily during a course of antibiotics and for 1 week after the completion of antibiotic therapy showed reduction in the incidence of AAD [Hickson et al. 2007].
Clostridium difficile colitis (CDC) has not consistently been shown to be prevented by probiotic cotreatment in a number of studies [Pillai and Nelson, 2008; McFarland, 2006]. In the largest study to date S. boulardii, at a dose of 2×1010 per day, in combination with vancomycin and metronidazole was associated with a significant decrease in risk of CDC recurrence [McFarland et al. 1994]. In addition, the study by Hickson and colleagues above showed efficacy in preventing CDC [Hickson et al. 2007]. Other studies, however, have not confirmed this benefit. A recent Cochrane review, limited by the small number of quality studies, concluded that there was insufficient evidence to recommend probiotic use even in combination with vancomycin and metronidazole [Pillai and Nelson, 2008]. In practice, however, many clinicians tend to recommend probiotics after antibiotic treatment, particularly in patients who have had CDC relapse.
Infectious diarrhea in both adults and children may be shortened by the use of probiotics [Allen et al. 2004]. The duration of symptoms is decreased by about 30 hours as suggested by a systematic review of trials in active infectious diarrhea. In this Cochrane review, 23 studies including almost 2000 participants (352 of which were adults), it was concluded that probiotics reduced the risk of persistent diarrhea compared with placebo or no probiotics at 3 days with a RR of 0.66 (95% CI 0.55–0.77) [Allen et al. 2004]. The majority (18 out of 23 studies) of the probiotics tested were LABs with two studies using S. boulardii.
Inflammatory bowel disease (IBD) treatment with antibiotics is fraught by variable efficacy. Recent discovery of genetic polymorphisms (NOD2/CARD15) in CD which play a role in bacterial peptidoglycan recognition [Inohara et al. 2003; Kobayashi et al. 2005], may be responsive to alterations in enteric flora and thus be important in the pathogenesis and maintenance of IBD [Sartor, 2004].
- Ulcerative colitis (UC). Several trials have been published examining probiotics in the induction and remission of UC, however, only few of these are RCTs. Most are with different probiotic formulations and overall have been performed in a relatively small number of patients (Table 3). For induction of remission, the first and largest controlled trial to date published by Remnacken showed no additional efficacy of E. coli Nissle 1917 than steroids, mesalazine, and antibiotics [Rembacken et al. 1999]. Three additional trials, all small in number of patients and of short duration of therapy and with variable standard of care, showed improvement in various measures of disease activity and even cytokine profiles [Furrie et al. 2005; Kato et al. 2004; Tursi et al. 2004]. Mallon and colleagues performed a Cochrane database systematic review, but no formal meta-analysis was possible due to differences in probiotics, outcomes and methodology, and concluded that probiotics when combined with other therapies did not improve remission rates [Mallon et al. 2007]. However, this analysis showed a reduction in disease activity in mild to moderately severe UC. A second systematic review published recently also suggested a similar efficacy profile between probiotics and anti-inflammatory agents [Zigra et al. 2007]. With regard to maintenance of UC remission, probiotics have been tested in a larger number of patients (Table 3). One trial by Kruis and colleagues tested E. coli Nissle 1917 and found no difference in relapse rates in patients on a probiotic versus mesalamine [Kruis et al. 2004]. A trial by Zocco and colleagues also found no difference in relapse rates at 6 or 12 months when comparing Lactobacillus GG with mesalamine with a combination of the two [Zocco et al. 2006]. Those patients who took the probiotic did appear to have a longer time to relapse. All of these studies support the idea that probiotics may be as effective as mesalamine in maintaining remission in the short-term trials.
- Crohn’s disease. The literature on the induction and maintenance of remission in CD is heterogeneous and difficult to interpret (Table 4). Partly, this is due to the unclear definition of extent of inflammatory involvement in patients who were studied and a small number of patients included in the trials. Furthermore, very few studies examined the additive effect probiotics may have on active CD. In one study with only 11 patients, probiotics provided no additional benefit to steroids and antibiotics in inducing remission [Schultz et al. 2004]. An open-label study with 10 patients who were refractory to prednisolone and aminosalicylates, were tried on a combination of probiotics (B. breve, B. longum, and L. casei) and a prebiotic (psyllium) simultaneously. A complete response was found in 6 of 10 patients without any adverse consequences [Fujimori et al. 2007]. More controlled studies have been performed on the maintenance of remission in adults with CD (Table 4), but in general these studies fail to show any benefit of probiotic administration [Schultz et al. 2004; Guslandi et al. 2000; Malchow, 1997]. Data are even more robust on the prevention of relapse following surgical intervention, but again probiotics fail to prevent endoscopic or clinical recurrence (Table 4) [Chermesh et al. 2007; Van Gossum et al. 2007; Marteau et al. 2006; Prantera et al. 2002]. Several meta-analyses and systematic reviews have shown that probiotics were ineffective in maintenance of remission in CD [Rahimi et al. 2008; Rolfe et al. 2006].
Table 5. Controlled trials of probiotics for the prophylaxis and remission of pouchitis.
Irritable bowel syndrome (IBS) is a multisymptom GI disorder with unclear etiology and pathogenesis. Changes in GI microflora in IBS patients have been reported by a number of investigators [Kassinen et al. 2007; Shanahan, 2007]. Recently, reports on variable prevalences of small intestinal bacterial overgrowth (SIBO) in IBS have been published [Posserud et al. 2007; Lin, 2004]. IBS symptoms such as bloating or flatulence have been attributed to possible alterations in the intestinal microflora and probiotics have been used empirically to treat these difficult symptoms [Kim et al. 2003, 2005]. Postinfectious IBS may begin after a bout of acute gastroenteritis suggesting that altered microflora or induction of an altered inflammatory or immune state in the bowel may lead to altered bowel function and IBS symptoms [Collins et al. 2009]. An increase in lymphocytes and an increase in pro-inflammatory cytokines have been described [Spiller et al. 2000].
Recently Brenner and colleagues analyzed 16 RCTs in IBS patients who were defined either by Rome II or Manning Criteria and who received either single or a combination probiotics versus placebo [Brenner et al. 2009]. Bifidobacterium infantis 35624 demonstrated efficacy in two appropriately designed RCTs. Both global as well as individual IBS symptoms (abdominal pain, bloating, incomplete evacuation, intestinal gas, straining, and bowel function) were significantly improved without evidence to suggest an increase in adverse events. No other probiotic, including isolated Lactobacillus species, showed significant improvement in IBS symptoms in appropriately designed RCTs [Brenner et al. 2009].
Another systematic review of the literature evaluating efficacy of probiotics in the treatment of IBS revealed that probiotics had a statistically significant effect in reducing IBS symptoms with a number needed to treat (NNT) of 4 (95% CI 3–12.5) [Moayyedi et al. 2010]. Almost all probiotic combinations contained both Bifidobacteria and Lactobacilli; the latter had no effect as assessed by continuous data meta-analysis. This raises the possibility that Bifidobacter may be the active treatment in probiotic combinations.
Elevated levels of interleukin 6 (IL-6), IL-6R, IL-1Beta and tumor necrosis factor alpha (TNF-α) [Dinan et al. 2006; Liebregts et al. 2007] and a lower IL-10/IL-12 ratio [O’Mahony et al. 2005] have been reported in IBS patients in comparison to controls, suggesting that IBS may be associated with increased pro-inflammatory cytokine secretion. However, plasma cytokine levels may not necessarily reflect the expression or levels of cytokines in the mucosa of the bowel wall, but may come from activated immune cells in the spleen or liver [Nance and Sanders, 2007]. The imbalance between IL-10 and IL-12, observed in peripheral blood mononuclear cells, was confirmed at the mucosal level in a recent study by Macsharry and colleagues suggesting that this finding may be an underlying phenotype in IBS and a potential biomarker for a subset of IBS patients [Macsharry et al. 2008]. B. infantis was shown to increase IL-10/IL-12 ratio in IBS patients [O’Mahony et al. 2005] suggesting a possible mechanism by which this probiotic may exert its effect.
The effect of probiotics on other GI disorders have also been studied, including lactose intolerance, Helicobacter pylori infection, microscopic colitis, prevention and treatment of diverticulitis, and even colon cancer prevention. The studies have been small and meta-analyses are too variable to draw firm conclusions of benefit.
Summary and conclusions
Probiotics are a therapeutic class being increasingly used for a variety of GI disorders. Probiotics appear to alter intestinal microflora and may exert their effect(s) by a variety of mechanisms. Many species of probiotics exist and it is generally accepted that all probiotics are not created equal. Efficacy may be due to a single strain or multiple strains or a combination of different probiotics. There is good evidence to support the efficacy of S. boulardii and LABs and the combination of the two for AAD, VSL#3 for pouchitis, and B. infantis 35624 for IBS. Probiotics decrease the duration of symptoms in acute infectious diarrhea. Probiotics, including E. coli Nissle 1917, LGG, and VSL#3 are as effective as standard therapy (mesalamine) in inducing or maintaining remission in UC or CD. When added to standard therapy, probiotics do not provide additional benefit compared with standard therapy alone. Most probiotics tested to date are not more effective than placebo in inducing or maintaining IBD remission.
Probiotics have been shown to be safe in immunocompetent hosts in an outpatient setting. However, administration of probiotics to immunocompromised, chronically ill, hospitalized patients with GI disorders, and indwelling catheters may predispose them to probiotic sepsis. Specifically, in GI disorders in which gut permeability and gut immunity may be compromised, adding probiotics may increase translocation of bacteria into the bloodstream. Until further studies become available on safety of probiotics in hospitalized patients, we caution their use in this setting.
Future studies should address many of the remaining questions related to the basic knowledge of probiotics, such as the composition of human intestinal flora, viability and fecal recovery rates, physiological and immunological effects. Furthermore, most optimal doses, duration of treatment, comparison of different strains and different probiotics, single versus combination probiotics, combination of probiotics with prebiotics, efficacy of various probiotics in different disease states, and safety of probiotics in debilitated patients or in patients with compromised gut epithelial integrity need to be evaluated.
Conflict of interest statement
The authors declare that there is no conflict of interest.
References
Articles from Therapeutic Advances in Gastroenterology are provided here courtesy of
SAGE Publications
SAGE Publications