Fecal Microbiota Transplantation: Therapeutic Potential for a Multitude of Diseases beyond Clostridium difficile

ABSTRACT

The human intestinal tract contains trillions of bacteria, collectively called the gut microbiota. Recent insights have linked the gut microbiota to a plethora of diseases, including Clostridium difficile infection (CDI), inflammatory bowel disease (IBD), and metabolic diseases such as obesity, type 2 diabetes (T2D), and nonalcoholic steatohepatitis (NASH). Fecal microbiota transplantation (FMT) is currently tested as a therapeutic option in various diseases and can also help to dissect association from causality with respect to gut microbiota and disease. In CDI, FMT has been shown to be superior to antibiotic treatment. For IBD, T2D, and NASH, several placebo-controlled randomized controlled trials are under way. Moreover, techniques and standardization are developing. With the extension of FMT as a treatment modality in diseases other than CDI, a whole new treatment option may be emerging. Moreover, correlating alterations in specific strains to disease outcome may prove pivotal in finding new bacterial targets. Thus, although causality of the gut microbiota in various diseases still needs to be proven, FMT may prove to be a powerful tool providing us with diagnostic and therapeutic leads.

INTRODUCTION

The human intestinal tract contains trillions of bacteria, collectively called the gut microbiota. The majority of bacteria belong to the Gram-negative phyla Bacteroidetes and Proteobacteria and the Gram-positive phyla Firmicutes and Actinobacteria (1). In humans, the diversity of the gut microbiota and the abundance of species increase rapidly after birth and, after 2 to 4 years, remain relatively stable throughout adult life (2). Nevertheless, shifts in gut microbiota composition may occur, especially after use of antibiotics. Even a short course of antibiotics can result in perturbations that last for several years (3, 4).

The gut microbiota has several beneficial functions. For example, a stable gut community offers resistance against colonization by pathogenic bacteria. When the intestinal microbiota composition is disturbed, a decrease in bacterial diversity and subsequent decrease in colonization resistance are thought to allow pathogenic bacteria, which are normally found in low numbers, to expand and cause disease. The most illustrative example of this is Clostridium difficile infection (CDI) after use of antibiotics (5).

In order to restore a balanced gut microbiota composition, fecal microbiota transplantation (FMT), the transfer of fecal material containing bacteria from a healthy donor into a diseased patient, has been developed. Having been used for centuries and previously referred to as “fecal bacteriotherapy,” “fecal transfusion,” “stool transplantation,” or “fecal enema,” FMT has emerged as a therapeutic option for a wide range of diseases. Moreover, FMT may be used as a research tool to search for novel therapeutic targets. In this chapter, we focus on the therapeutic potential of FMT.

HISTORY OF FMT

The use of FMT for human disease goes back many centuries. During the Dong Jin Dynasty, in the 4th century AD, the traditional Chinese medicine doctor Ge Hong first described the oral ingestion of human fecal suspension in his handbook “Zhou Hou Bei Ji Fang” (“Handy Therapy for Emergencies”). It was used to successfully treat patients who had food poisoning or severe diarrhea. In the 16th century, in the traditional Chinese medicine book “Ben Cao Gang Mu” (“Compendium of Materia Medica”), another well-known Chinese doctor, Li Shizhen, documented treatment of severe diarrhea, fever, pain, vomiting, and constipation using several different preparations, including dry feces, fermented and fresh fecal suspensions, and infant feces. For aesthetic considerations, these treatments were aptly given original names, such as “yellow soup” and “golden syrup” (6). During World War II, German soldiers in Africa were recommended by Bedouins to treat bacterial dysentery with “consumption of fresh, warm camel feces” (7).

The first description of FMT in modern medicine dates back to 1958. Dr. Ben Eiseman, an American surgeon, used fecal enemas to treat four patients who had developed fulminant pseudomembranous enterocolitis after antibiotic use; the treatment resulted in a rapid resolution of symptoms (8). Although not known as a cause at that time, it is likely that these patients were suffering from CDI. In 2013, the first randomized controlled trial (RCT) using FMT in CDI patients was published (9). Since then, FMT has been investigated as a possible therapy in a variety of diseases.

PRACTICAL FMT GUIDELINES

In the past years, FMT has become a booming practice, ranging from highly organized stool banking programs to individual treatments with patient-identified donors, and even to harmful do-it-yourself practices. Most published protocols and recommendations regarding FMT methodology are based on opinions, common sense, and anecdotal experiences. Thus, regarding donor selection, preparation of fecal samples, and administration of the solution, large differences in FMT methods still exist among centers worldwide. In 2011, the FMT Workgroup described a general protocol for FMT (10). Moreover, recently a European consensus report on clinical indications, applications, and methodological aspects of FMT was published (11). Here, we provide practical guidelines regarding FMT for CDI and research purposes based on these protocols. These steps are based on what has been described but not necessarily rigorously tested.

In all cases of planned FMT, local ethical approval and patient and donor informed consent should be obtained. Moreover, FMT should be performed only in specialized centers with extensive prior experience. In the case of CDI, a fecal transplantation expert and an infectious disease expert should be consulted to consider other treatment options and determine patient eligibility.

Donor Selection

It is important to carefully screen potential fecal donors for pathogens to prevent disease transmission to the recipient. Structured questionnaires can be used to estimate the risk of recently acquired pathogens that might not be detected in laboratory analysis. These questionnaires should at least contain questions regarding travel history, sexual behavior, medical history, use of medication, recreational drug use, and risk factors for communicable diseases such as recent tattoos or piercings. Donors that have had antibiotic exposure in the past 3 to 6 months, gastrointestinal (GI) illness, diarrhea, or medication use should be excluded. Apart from questionnaires, potential donors should undergo blood and stool testing for infectious diseases. In Table 1, we list the pathogens that should be tested.

TABLE 1.

Recommended analyses for screening of fecal donors^a

History

Antibiotic use in past 3 months

Atopy, allergies, or autoimmune diseases

Blood transfusions

Gastrointestinal illness (IBD, IBS, colorectal polyps, or cancer)

Incarceration, tattoos, or body piercings in past 6 months

Infectious diseases (HBV, HCV, HIV, HTLV, malaria, trypanosomiasis, tuberculosis, rotavirus, Giardia lamblia)

Metabolic disorders (morbid obesity, diabetes)

Neurologic disorders

Previous reception of tissue/organ transplant

Previous reception of blood products

Risky sexual behavior

Travel history

Use of any medication or illegal drugs

Stool testing

Adenovirus 40/41

Adenovirus non-41/41

Aeromonas spp.

Astrovirus

Blastocystis hominis

Clostridium difficile

Cryptosporidium spp.

Dientamoeba fragilis

ESBL-producing Enterobacteriaceae

Entamoeba histolytica

Enterovirus

Giardia lamblia

Helicobacter pylori

Isospora spp.

Microscopy for parasites, cysts, and ova

Microsporidium spp.

Norovirus type I and II

Pathogenic Campylobacter spp.

Plesiomonas shigelloides

Rotavirus

Salmonella spp.

Sapovirus

Shiga toxin-producing Escherichia coli

Shigella spp.

Vancomycin-resistant Enterococcus spp.

Yersinia spp.

Serologic testing

CMV

EBV

Entamoeba histolytica

Hepatitis A, B, C, E virus

HIV

HTLV

Strongyloides spp.

Treponema pallidum

^a A questionnaire and extensive screening of donor blood and feces are recommended. In the case of any positive result, a consultation with the clinical microbiologist is required to determine the eligibility of the donor. IBD, inflammatory bowel disease; IBS, irritable bowel syndrome; CMV, cytomegalovirus; EBV, Epstein-Barr virus; HTLV, human T-lymphotropic virus.

It is unpractical and expensive to screen donors shortly before every donation. Therefore, we propose to screen donors once every 6 months in the absence of overseas travel and illness. After traveling or illness, donors should be rescreened before being allowed to donate.

Donors should preferably be aged less than 60 years. However, this recommendation should not be mandatory, in order not to foreclose the use of intimate healthy partners merely because of age. It is unclear whether the use of feces from related donors improves the therapeutic potential of FMT, although this choice may be driven by specific needs, such as when there is a potential advantage of choosing unrelated donors to treat conditions with a pathogenic genetic basis. Two recent meta-analyses investigating the effect of FMT on CDI showed differential effects on resolution compared to nonrelated donor FMT (12, 13). For practical reasons (e.g., when using stools from donor banks), it may be more convenient to use donor feces from nonrelated donors. However, if a patient wishes to receive FMT from a related donor, there is evidence to abstain from this. In clinical studies, investigators should choose either related or nonrelated donors in all participants to improve homogeneity in the findings.

Preparation of the Fecal Sample

During preparation of the fecal sample, several precautions must be taken to ensure the safety of staff. For example, specimens must be contained in an airtight container and further processed under a hood, as stool is a level 2 biohazard. Preferably, processing should occur in an anaerobic chamber to protect obligate anaerobic bacteria. However, many studies showed beneficial effects of FMT despite processing the sample in an aerobic chamber, so this is not a necessity.

A stool sample weighing at least 30 g should be mixed with up to 500 ml of diluent, usually sterile saline. However, it is worth noting that bacterial counts are more relevant to measuring dosing than stool weight, as they can vary ∼10-fold even for the same donors on different days, so that 30 g on one day may actually be equivalent to 3 g (or 300 g) on another day. Homogenization can be done manually or using a dedicated blender. The sample should be filtered or centrifuged to remove large particles such as undigested food. The final amount of bacterial solution that is administered should be 200 to 500 ml.

The use of ethanol in the preparation of the suspension is still controversial. Despite encouraging phase 1 findings (14), an ethanol-treated mix of sporulating bacteria failed to treat recurrent CDI in a phase 2 study (15). It is plausible that ethanol, apart from killing pathogens, can also severely alter the composition of the commensal microbiota, possibly eliminating critical elements such as bacteriophages, fungi, and nonsporulating bacterial components.

It is expected that storing the sample decreases the viability and diversity of the bacteria. After obtaining the sample, it can be stored at 4°C until further processing. Storing the sample at 4°C for 8 h was shown to reduce diversity by 10% (16). Thus, we recommend to minimize the time from production of the sample to administration of the solution and preferably to keep it under 6 h.

As an alternative to fresh feces, frozen stool samples can be used. In CDI, a randomized, open-label, controlled study showed that using frozen FMT was safe and effective (17). Later, a double-blind RCT showed that frozen FMT was not inferior to fresh FMT in patients with CDI (18).

The FMT approach based on frozen feces is essential for the development of a stool bank and is the optimal way to standardize the FMT process and to allow the availability of stool on demand. Moreover, a frozen stool bank allows fecal donors to be recruited and thoroughly screened ahead of time, in a methodical manner, without time pressure, and with the potential advantage to reduce the costs associated with donor screening, as one donor can serve for multiple FMT donations. Finally, stool banks may improve the accessibility of FMT to centers that otherwise would be unable to provide the service due to inadequate resources to conduct donor recruitment/screening and FMT processing.

Indeed, several countries have created a stool bank in which frozen stool samples are stored for later use in FMT. Practical guidelines that can be used in establishing a frozen stool bank have been published elsewhere (19).

Administration of the Fecal Solution

The fecal solution can be administered via an upper GI (i.e., duodenal) or a lower GI (i.e., colonic) route. A nasogastric tube is not recommended due to risk of fecal aspiration if the recipient has gastric reflux or vomits. Colonic administration preferably occurs with the use of a colonoscope, although sigmoidoscopy and enemas have also been used in several studies. However, these approaches expose only part of the colon to the bacterial solution.

It is unclear which route is to be preferred. In CDI, both duodenal and colonic administration routes have provided excellent results (9, 12, 20). However, several systematic reviews and meta-analyses reported that colonoscopy achieved higher resolution rates of recurrent CDI and a similar safety profile compared to other routes of delivery (20–22). An individual patient data meta-analysis of FMT for CDI in nonrandomized trials showed failure rates of 5.6% and 17.9% after 1 and 3 months, respectively, in the upper GI group compared to 4.9% and 8.5% in the lower GI group (23). However, these results still need to be validated in a randomized controlled setting.

Another option is to apply the suspension by one or more enemas, especially in cases where gastroduodenoscopy or colonoscopy are contraindicated. Moreover, enema administration is widely available, does not require costly devices, and is less invasive than other routes. In the case of administration via enema, patients should be instructed to hold the infusate for at least 30 min and to remain supine to minimize the urge to defecate. The procedure could be repeated.

As the differences in efficacy are not sufficiently known, the choice between different administration modalities may depend on other factors. For example, in patients with GI motility disorders, the lower GI approach may be preferred, while in patients with toxic megacolon or severe colitis, the upper GI approach may be better suited due to the risk of colonic perforation.

Before administration, bowel preparation with laxatives may be considered, especially in the case of a lower GI approach. Bowel cleansing reduced the total bacterial load 31-fold on average (24) and may improve chances of successful engraftment of donor strains. We recommend using 1 to 2 liters of an osmotic laxative and to administer the bacterial solution at least 1 h after intake of the laxative. If a nasoduodenal tube is in place, laxatives can be injected directly into the duodenum. Recipients may also orally ingest the laxative the night before FMT.

In some trials, proton pump inhibitors were given to recipients prior to administration of the bacterial solution. However, there is no evidence supporting the use of proton pump inhibitors in FMT, and thus we do not recommend it. In the case of CDI, antibiotics should not be given on the day of FMT, but pretreatment with antibiotics for 3 to 5 days prior to FMT can be considered.

Safety

When adhering to screening guidelines, FMT has been shown to be a relatively safe procedure, even in immunocompromised patients (25). To date, no cases of transmitted infectious diseases due to FMT have been reported. Minor short-term adverse events such as diarrhea, flatulence, abdominal discomfort, and cramping are common after FMT. Most of these adverse events are self-limiting and disappear within 2 days after FMT (9).

In the upper GI approach, there is a risk of fecal aspiration if the solution is administered in the stomach. One case of fatal aspiration pneumonia due to sedation during colonoscopy was published (26). Moreover, when using the lower GI approach, there is a risk of colonic perforation.

Regarding very-long-term risks (more than 5 years), little evidence is available (27). In theory, FMT could increase the risk of diseases associated with the gut microbiota, including obesity, inflammatory bowel disease (IBD), and colorectal cancer. Moreover, there is a theoretical possibility to transmit potentially harmful microbiota traits that cannot be apparent for decades. However, for recurrent CDI this does not really matter, as these patients are most often elderly and FMT could be a life-saving treatment. For other indications, long-term prospective studies are necessary to assess the risks. Moreover, as FMT has not undergone the traditional regulatory approval process of pharmaceutical products with sequential testing leading to large phase 3 trials assessing efficacy and safety prior to clinical utilization, the implementation of registry data collections should be encouraged to effectively deal with the issue of long-term monitoring of patients for adverse events. For example, the American Gastroenterological Association established a National Institutes of Health-funded registry to track patient outcomes associated with FMT (28).

FMT IN C. DIFFICILE INFECTION

CDI (formerly known as C. difficile-associated disease) is one of the most important health care-related infections. While C. difficile can be cultured from feces of 3% of healthy adults, in hospitalized patients colonization rates may be 20 to 30% (29). In the United States in 2001 to 2010, incidence rates of CDI were up to 8 per 1,000 discharges, with a mortality rate of up to 8.0% (30), which resulted in a financial impact of $3.2 billion/year (31).

CDI is driven mainly by production of toxins by C. difficile, although disease severity is not associated with fecal toxin concentrations (32). The clinical course of the disease is dependent on the host immune response and the virulence of the invading strain. The increasing prevalence and severity of CDI since the 2000s is most likely due to an increase in virulent strains, most notably ribotype 027, which is associated with higher mortality and transmissibility than other strains (31, 33). The colonic microbiota of patients suffering from CDI is deficient in members of Firmicutes and Bacteroidetes, whereas Proteobacteria species are enriched (34, 35).

Historically, the first-line treatment for patients with CDI has been metronidazole or vancomycin (recently, the more expensive fidaxomicin has also been approved for the treatment of CDI by the FDA). However, although most infections initially respond to this treatment, recurrence rates range from 20% after an initial episode to 60% after recurrent CDI (36, 37). The underlying mechanisms involved in recurrence are reexposure or reactivation of spores in patients who have an impaired immune response to infection and an impaired colonic epithelial barrier function.

Previously, recurrences were usually treated with an additional course of metronidazole or vancomycin. Starting with Eiseman in 1958 (8), more than 500 patient cases showing successful treatment with FMT have now been described. RCTs have confirmed the efficacy of FMT in recurrent CDI, with an overall cure rate of 85 to 90%, compared to a cure rate of 30% with standard therapy (9, 17).

Of particular interest is the C. difficile ribotype 027, also referred to as the North American pulsed-field type 1 and restriction endonuclease type BI or NAP1/BI/027 strain. This hypervirulent strain has spread globally in the past decades, from North America to Europe, Australia, and Southeast Asia, and is associated with regional and national outbreaks with high rates of disease recurrence and mortality (38–41). Large differences in prevalence exist between different regions and countries (42). In the 1990s, strains belonging to ribotype 027 were infrequently isolated from CDI patients. However, in a large prospective point-prevalence study examining stool samples from CDI patients in 482 hospitals across 20 European countries in 2012–2013, ribotype 027 was the most prevalent strain, accounting for 19% of cases (43). Most ribotype 027 strains were found to be localized mainly to four countries (Germany, Hungary, Poland, and Romania), highlighting national differences in prevalence.

The problem of C. difficile ribotype 027 is further increased because of its insensitivity to several antibiotics. For example, the strain has reduced sensitivity to both vancomycin and metronidazole and is resistant to moxifloxacin (44). Fidaxomicin, a high-cost drug that is effective against other strains, has poor activity against C. difficile ribotype 027 and may not be cost-effective (45, 46). Given these issues, FMT may be an especially effective treatment option for patients infected with this strain.

FMT is now recommended as one of the treatment options in recurrent CDI by both the European Society for Microbiology and Infectious Disease (ESCMID) and the American College of Gastroenterology (ACG) (46, 47). Moreover, given its efficacy in recurrent CDI, FMT has been tested in primary infection and immunocompromised patients (26, 48). More easily applicable therapies are also being investigated. Treatment with a frozen capsulized microbiota may be equally effective in CDI (17, 18, 49). Other research is focusing on mining for specific strains that are involved in protection against CDI. Thus, building on the success of FMT in CDI, further research may provide interesting new therapies that are more easily applicable and less invasive than current strategies. A recently published trial provided evidence that this strategy might be feasible (14). However, it has to be kept in mind that CDI is from a pathological viewpoint a relatively simple disease for which the causality of the gut microbiota is clear. Thus, while manipulation of the gut microbiota may be expected to be a successful treatment option in CDI, for other diseases this may not always be the case (Fig. 1).

FIGURE 1.

Causality of the gut microbiota in different diseases. The gut microbiota may have various levels of causality in diffent diseases. Diseases that have a more complex pathophysiology may respond less to fecal microbiota transplantation. ESBL, extended spectrum beta-lactamase producer; VRE, vancomycin-resistant enterococci; IBD, inflammatory bowel disease; IBS, irritable bowel syndrome; T2D, type 2 diabetes. Modified from Molecular Metabolism with permission of the publisher.

FMT IN IBD

IBD encompasses several separate entities that include ulcerative colitis (UC) and Crohn’s disease (CD). While both UC and CD are characterized by chronic relapsing inflammation of the bowel, there are important pathophysiological differences. UC generally affects the large intestine and is restricted to the epithelial lining of the gut, while CD can affect the entire gastrointestinal tract and a transmural pattern of inflammation is often seen. Nevertheless, there is overlap between the pathophysiology and clinical presentation, and a definitive distinction is not always possible.

Both UC and CD have been linked to gut microbiota dysbiosis. UC is characterized by a decrease in Firmicutes and Bacteroidetes, with an increase in Proteobacteria and Actinobacteria (50). However, it is unclear to what extent these shifts represent a cause or a consequence, as IBD is a complex disease involving genetic, dietary, immune system, and gut microbiota factors. Possibly, the interaction between the gut epithelium and the mucosal microbiota plays a larger role in the pathogenesis of the disease than the luminal microbiota. Indeed, so far, no intestinal bacterium has been found to be able to induce IBD. Moreover, FMT has rendered conflicting results in ulcerative colitis (51, 52). Nevertheless, animal models have shown a role for the gut microbiota in the pathogenesis of IBD, and the use of FMT in IBD is currently ongoing in several randomized studies, which will show whether and to what extent FMT can improve disease symptoms in (subsets of) patients with IBD. These studies will be pivotal in showing which bacterial strains should be developed as therapeutic targets for IBD.

The first report of infusion of donor feces for ulcerative colitis dates from 1989, when the patient-physician Bennet treated himself with multiple fecal enemas, resulting in long-term remission of his IBD flares (53). Since then, several FMT studies have been published with ranges of effects in patients with CD and UC. Importantly, the vast majority of trials on FMT in IBD are one-armed cohort studies or case series. In 2012, a meta-analysis of FMT found a 63% remission rate in UC patients, while 76% of IBD patients experienced a reduction in symptoms and 76% were able to stop taking drugs for their disease. Another meta-analysis of only cohort studies performed in 2014 found a clinical remission rate of 36% in IBD patients and 61% in CD patients. A subgroup analysis demonstrated a pooled estimate of clinical remission of 22% for UC (54). Of note, no RCTs were included in either of these meta-analyses.

Two RCTs investigating the effect of FMT in UC patients have been performed so far. Moayyedi et al. (52) randomized 70 patients with active UC, defined as a Mayo Clinic score of ≥4 with an endoscopic Mayo score of ≥1 to receive allogenic FMT or placebo (water) via retention enema. Patients received 50 ml of fecal suspension from anonymous donors (all but one case) or a partner (one case) or water weekly for 6 weeks. Notably, the fecal suspension was either administered immediately (n = 16) or stored at −20°C and later thawed (n = 22). There was a significant difference in the primary endpoint of a full Mayo score of <3 and endoscopic Mayo score of 0 at 7 weeks after the first FMT, with nine patients (24%) in the FMT group achieving the endpoint versus two patients (5%) in the control group. FMT also resulted in a greater improvement in microbial diversity compared with placebo treatment.

Interestingly, the authors used only six donors to minimize variability of the intervention. Seven of nine patients in remission after FMT received fecal material from a single donor, while this donor was used in 18 cases. Three donors did not induce remission in any patient, although it is not described how many patients received FMT from these donors. The difference in treatment success between the best donor (7 of 18, 39%) and the other donors (2 of 20, 10%) was not statistically significant (P = 0.06). Thus, although there is a trend for superiority of one donor compared to the others, even fecal suspensions from this donor had various results in the patients.

In the other RCT, Rossen et al. (51) randomized 50 patients with mild to moderately active UC with a patient-reported simple clinical colitis activity index (SCCAI) score of ≥4 and ≤11 and an endoscopic Mayo score of ≥1 in a double-blind fashion to receive allogenic or autologous FMT. Patients received two duodenal FMTs of 500 ml fresh (≤6 h after production) fecal suspension 3 weeks apart after pretreatment with bowel lavage. Fifteen donors were used, of whom one was related to the recipient (partner). In the intention-to-treat analysis, 7 of 23 patients (30.4%) who received allogenic FMT achieved the composite primary endpoint of a SCCAI score of ≤2 in combination with ≥1-point improvement on the combined Mayo endoscopic score of the sigmoid and rectum 12 weeks after the first FMT, compared to 5 of 25 (20.0%) in the autologous group. This difference was not significant (P = 0.51). In the per-protocol analysis, 7 of 17 allogenic FMT recipients (41.2%) compared to 5 of 25 controls (25.0%) achieved the primary endpoint (P = 0.29).

The authors could not identify “superdonors” or “poor donors.” For example, one donor donated 12 times for eight patients, of whom four patients responded to treatment. Although microbial diversity as measured by the Shannon diversity index was not different between donors and patients at baseline, an improvement in diversity was seen in responders in both the allogenic FMT group (P = 0.06) and the autologous group (P = 0.01). Redundancy analysis showed that at 12 weeks, responders had a significantly higher microbiota composition similarity to that of their donors than did nonresponders.

These RCTs were rather different in their methodology, making it hard to interpret the different findings. Specifically, one study used weekly retention enema for 6 weeks, while the other used two duodenal infusions in 3 weeks. However, considering that neither study could identify a “superdonor,” the response to FMT may be determined by engraftment of donor bacterial strains. Indeed, analyses of microbiota engraftment after FMT in metabolic syndrome patients revealed that same-donor recipients had various degrees of microbiota transfer, indicating differences in donor-recipient “compatibility” (55).

Regarding the effect of FMT on UC activity, a 2016 meta-analysis pooled the data from several nonrandomized cohort studies with these two RCTs and showed an overall clinical response rate of 65% and an overall clinical remission rate of 42% (56). However, it is important to note that the cohort studies were generally of very poor quality.

Thus, FMT offers promising therapeutic potential in UC, although more high-quality studies are needed. Of particular interest is determining the factors that influence response versus nonresponse.

Treating CD with FMT is more challenging than treating UC, as the former disease usually involves the small intestine rather than the large intestine. For example, FMT via enema does not reach the terminal ileum, the most frequently affected region in CD. Moreover, if multiple FMTs are needed, FMT via colonoscopy or duodenoscopy may not be feasible or well tolerated. Thus, the use of capsulized microbiota may be an interesting approach in this patient group. However, so far, no RCT investigating FMT in CD patients has been published.

Of note, use of FMT in IBD patients should be considered with caution, as in a small group of CD patients, adverse effects (transient fever, abdominal pains, bloating, and no clinical improvement with only transient effects on the host’s fecal microbial composition) were reported upon three consecutive FMTs.

In conclusion, FMT is a promising therapy for IBD patients that may lead to major improvements in patient care. However, at present there is limited high-quality evidence, especially regarding CD. Moreover, there are some concerns regarding the safety of FMT in this vulnerable group of patients. Several randomized studies are ongoing to determine the safety, relevance, and best way of using FMT in this group. Until these trials have shown consistent results, clinicians should be cautious in recommending FMT to their IBD patients.

FMT IN METABOLIC DISEASE

Metabolic diseases, such as obesity, insulin resistance, metabolic syndrome, nonalcoholic fatty liver disease, and nonalcoholic steatohepatitis (NASH), constitute a major disease burden. For example, in 2014, more than 1.9 billion adults were overweight (body mass index [BMI] ≥ 25 kg/m²), of whom over 600 million were obese (BMI ≥ 30 kg/m²) (57). In the United States, one in three adults is obese (58). Obesity is associated with a wide range of pathologies, including cardiovascular disease, hypertension, type 2 diabetes mellitus (T2D), hyperlipidemia, stroke, several types of cancer, sleep apnea, liver and gallbladder disease, osteoarthritis, and gynecological problems (59). Thus, it is unsurprising that obesity and even overweight result in a significant increase in all-cause mortality (60).

Since the development of high-throughput next-generation sequencing techniques, there has been increasing interest in the role of the gut microbiota in metabolic diseases. As the gut microbiota is involved in the digestion and absorption of nutrients from the diet, direct energy absorption may be a major way in which the gut microbiota contributes to the development of obesity. Indeed, the gut microbiota converts indigestible fibers into energy-rich substrates such as short-chain fatty acids (SCFAs). Moreover, the gut microbiota promotes absorption of monosaccharides. Germfree mice remain significantly leaner than conventional mice, despite a higher food intake (61), while colonization of germfree mice increases total body fat content without changes in intake (62).

Several pathways are involved in signaling between the gut microbiota and the host and may regulate metabolic homeostasis. For example, SCFAs directly affect lipogenesis (63, 64), gut barrier function (65–67), gut motility (68), and immune responses (69–71). Moreover, the gut microbiota has been suggested to be involved in the secretion of several intestinal hormones that regulate metabolism. For example, SCFAs bind to G-protein-coupled receptor 41, mediating secretion of peptide YY, a hormone that reduces appetite (72, 73).

A third mechanism by which the gut microbiota is involved in the development of metabolic diseases is bacterial translocation. In this process, whole bacteria, bacterium-derived vesicles, or bacterial metabolites (such as lipopolysaccharide [LPS], a major component of the cell wall of Gram-negative bacteria that is highly immunogenic) translocate from the intestinal lumen into the systemic circulation and visceral adipose tissue. Indeed, both obesity and insulin resistance are associated with a low-grade inflammatory state characterized by endotoxemia (i.e., the presence of LPS in the plasma) (74). In particular, inflammation of visceral adipose tissue (which is in close proximity to the gut microbiota), which is characterized by infiltration of macrophages (75, 76) and increased production of cytokines (77), is a crucial driver in the development of insulin resistance.

While bacterial translocation is generally accepted to be a crucial factor in the development of insulin resistance in obesity, the exact mechanisms underlying bacterial translocation have not been elucidated. However, a high-fat diet seems to be an important trigger. For example, mice fed a high-fat diet for 4 weeks developed peripheral insulin resistance along with increased levels of circulating LPS and bacterial DNA, while a chronic infusion of LPS resulted in the same phenotype (74, 78). In humans, even a single high-fat meal has been shown to result in increases in LPS (79–81). The gut barrier dysfunction that leads to this “dietary endotoxemia” may be a direct consequence of the gut microbiota. For example, alterations in gut microbiota composition with subsequent translocation of bacterially derived antigens have been associated with nonalcoholic fatty liver disease and NASH (82–84).

So far, only one RCT studying the effect of fecal transplantation on insulin resistance has been performed (85). This trial showed an improvement of insulin sensitivity after lean-donor FMT. Interestingly, even in this small trial there was a clear distinction between responders and nonresponders, which was unrelated to the insulin sensitivity of the donor. An analysis of gut microbiota composition in patients included in several FMT studies showed an important diversity in donor bacterial engraftment (55). These results suggest that the effects of FMT may be dependent on donor-recipient compatibility and warrant further research.

The mechanisms underlying the favorable effects of FMT on insulin sensitivity are not yet clear. Future research should focus on the effect of FMT on bacterial translocation. Moreover, other methods of manipulation of the gut microbiota may be used. For example, vancomycin was shown to affect insulin sensitivity in humans (86), further suggesting an important causal role for the gut microbiota in the development of insulin resistance. Especially interesting in this regard is the role of gut microbiota manipulation on dietary bacterial translocation, which warrants further study.

Although the effect of FMT on insulin sensitivity is remarkable, it is unlikely to be used as a future therapy, as it is an invasive procedure. Moreover, gut microbiota composition has been shown to return to the original recipient composition, even when there is significant engraftment of donor bacteria (55). Instead, future research should aim to find specific causal agents involved in (the prevention of) bacterial translocation, ideally in human intervention studies.

FMT IN OTHER DISEASES

Irritable Bowel Syndrome

IBS is a chronic noninflammatory gastrointestinal disorder characterized by abdominal pain with diarrhea and/or constipation. In developed countries, the prevalence of IBS may be as high as 10 to 15% (87, 88). Given that IBS has a profound impact on quality of life and health care costs (89, 90), finding novel therapies is urgent.

Although the pathogenesis of IBS is not fully understood, there have been several associations with the gut microbiota. For example, IBS patients have a reduced variability of the gut microbiota compared to healthy controls (91). Moreover, IBS patients had lower levels of bifidobacteria and lactobacilli, while Enterobacteriaceae were more prevalent (92, 93). Other studies reported reduced populations of Bacteroidetes (94, 95) and lactobacilli and Collinsella (96) in IBS patients. Additionally, the microbiota of IBS patients may produce a lower amount of butyrate and higher amounts of sulfides and hydrogen than healthy microbiotas, which could promote IBS symptoms (20). Moreover, IBS patients have altered protein and carbohydrate energy metabolism within the gut (12) and increased amounts of acetic and propionic acids, which are associated with both the severity of abdominal pain and bloating symptoms (21).

A meta-analysis showed a positive effect of probiotics in IBS patients (97). However, high-quality data using the same probiotic in a randomized setting are scarce. Instead, FMT might offer a better alternative, as healthy donor feces represent a natural stable gut flora. So far, FMT has been used to treat IBS only in case series. One case series on FMT within this patient group showed that of 45 IBS patients with chronic, severe constipation who underwent FMT, a total of 60% obtained improved defecation and absence of bloating and abdominal pain during a follow-up of 9 to 19 months (98, 99). So far, no RCT evaluating FMT as a treatment option for IBS has been published. However, at the moment, one RCT is recruiting subjects (100). Thus, while there is no high-quality evidence so far, FMT may represent a viable therapeutic option for this patient group.

FMT in Intestinal Colonization by Multidrug-Resistant Pathogens

The increase in antimicrobial resistance, in part due to excessive use of antibiotics, is a major public health problem (101). Antimicrobial therapy for systemic disease, often given orally, has a major impact on the gut microbiota. A healthy gut microbiota may be able to prevent intestinal colonization by gut pathogens such as extended spectrum beta-lactamase (ESBL) producers. However, exposure to antibiotic leads to destruction of the gut microbiota, resulting in reduced colonization resistance.

FMT may offer a new way of treating intestinal colonization by multidrug-resistant pathogens. Indeed, animal studies have shown promising results (102, 103). For example, FMT resulted in decolonization of vancomycin-resistant enterococci (VRE) (104). In humans undergoing hematopoietic stem cell transplantation, patients who did not acquire VRE had higher levels of Barnesiella species in their feces samples than those who did (105). Thus, the composition of the gut microbiota may be important in maintaining colonization resistance. However, in humans, no prospective randomized intervention study has been performed on this subject. Our group did publish the first case report of a patient with an ESBL-producing Escherichia coli, which was eradicated by a single FMT (106). Moreover, five of five patients cleared methicillin-resistant Staphylococcus aureus (MRSA) after three jejunal administrations of FMT (107), while a study using RBX2660, an experimental microbiota suspension, for FMT in CDI patients showed that 8 of 11 VRE-positive patients became negative after treatment.

A search performed in December 2016 on www.clinicaltrials.gov revealed five trials that were under way to evaluate the effectivity of FMT against multidrug-resistant pathogens (NCT02312986, NCT02543866, NCT02461199, NCT02390622, NCT02472600). Thus, FMT offers a promising option in fighting colonization by multidrug-resistant pathogens, and important studies will be performed in the near future.

CONCLUSION

FMT has undergone a revival and is currently being tested as therapeutic agent in a variety of diseases. For example, a search performed in December 2016 on www.clinicaltrials.gov for intervention studies containing the terms “fecal” and “transplantation” revealed 103 active studies, of which 16 involved children aged 0 to 17 with a variety of indications. While there is no doubt about the efficacy of FMT in CDI, in other diseases the evidence is not as clear. In IBD and IBS, FMT has shown promising results that require further investigation. In particular, prospective RCTs are necessary to determine the effects of FMT in different patient groups.

Apart from CDI, IBD, and IBS, FMT may be useful in several other pathologies. While high-quality, placebo-controlled data so far are lacking, treatment by FMT could prove highly valuable in specific subsets of patients. Moreover, as there has been increasing interest in the relationship between gut microbiota and several autoimmune diseases, such as type 1 diabetes, idiopathic thrombocytopenic purpura (108), and even multiple sclerosis (109), FMT may even offer a future therapeutic strategy in these diseases. In conclusion, FMT is an exciting novel therapeutic option for a variety of diseases. However, high-quality evidence is lacking in most cases. Future therapy should resolve questions about the efficacy of FMT in diseases other than CDI.