To review the current clinical literature regarding the use of fecal microbiota transplantation (FMT) for severe and recurrent Clostridium difficile disease (CDAD).
To review the current clinical literature regarding the use of fecal microbiota transplantation (FMT) for severe and recurrent Clostridium difficile disease (CDAD).
Clostridium difficile (C. difficile) is a gram positive, spore forming bacteria, and an important nosocomial pathogen causing healthcare associated diarrhoea in hospitalized patients in developed and developing countries. During the past several years, CDAD has become more frequent, severe, refractory, and more likely to relapse. It has become apparent that C. difficile is no longer just a nosocomial infection, with a rising rate of infection in populations not previously affected. Standard treatment regimens and new medications exist, but recurrence rates are high.
Using PubMed, we conducted a Boolean search with the following medical subject headings (MeSH): Clostridium difficile infection and fecal transplantation or recurrent C. difficile infection. We restricted the search to human studies, published in English, between 2011 through June 1, 2013.
There were 104 publications identified. Of those related to FMT, there were 20 clinical reviews, 6 case reports, 3 clinical trials (one, a randomized control trial), and 1 meta-analysis. Since 1958 there have been 36 published reports of FMT for C. difficile infection (CDI) representing 583 patients. Success rates were higher when FMT was administered via colonoscopy (representing the majority of patients, 79.2%). The overall success rate for FMT, regardless of administration method, was 80–98%.
Fecal microbiota transplantation attempts to restore the normal microbiome of the colon, and has achieved a cure rate reaching more than 90%. Mounting evidence supports the utility of FMT for severe and recurrent cases of CDI. Barriers that will need to be addressed are patient perceptions and fears, standard protocol development, and further clinical trials.
Clostridium difficile is a Gram-positive, spore forming bacteria, and an important nosocomial pathogen causing healthcare associated diarrhoea in hospitalised patients in developed and developing countries. Clostridium difficile infection (CDI) causes millions of human infections each year . Severe CDI is defined as peripheral leucocytosis of 15,000 cells/ml or more and/or serum creatinine concentration increase by 1.5 fold or more above baseline, and associated with: hypotension, shock, sepsis, ileus, fulminant colitis, megacolon, perforation or death . Infection is thought to be mediated via enterotoxins (e.g., A, B, and binary toxin). During the past several years, C. difficile-associated disease (CDAD) has become more frequent, severe, refractory and more likely to relapse .
One of the major reasons for increase in severity and incidence of C. difficile is the emergence of a new strain B1/NAP1/027; although more than 90% respond to treatment, there is high recurrence rate reaching 20–60% within a few weeks of completion of treatment . Factors that increase risk for recurrence include: inadequate antitoxin antibody response, persistent disruption of the colonic micobiota, advance age, continuation of non-C. difficile antimicrobial therapy, prolonged hospital stay, concomitant use of antacid medication and long-term dialysis .
In recent years, it has become apparent that C. difficile is no longer just a nosocomial infection, with a rising rate of infection in populations not previously affected . Clostridium difficile is now seen in young, healthy individuals without previous exposure to hospitals or antibiotics, peri-partum women, patients with inflammatory bowel disease, cirrhosis, severe comorbid illnesses and the immune-compromised [4, 5].
Standard recommendations for treatment of mild to moderate CDI include oral metronidazole (500 mg three times a days for 10–14 days), or oral vancomycin (125 mg four times a day for 10–14 days). In patients with severe disease (e.g. toxic megacolon), it is recommended to use intravenous metronidazole with high dose vancomycin (i.e., 250–500 mg four times a day) either orally or through a nasogastric (NG) tube in case of difficulty with oral intake . In recent years, there has been an increase in metronidazole failure, reaching as high as 35% . Risk factors for metronidazole failure include recent cephalosporin use, C. difficile on admission, and transfer from another hospital. Interestingly, no association between the virulent strain B1/NAP1/027 and metronidazole failure has been observed . Recurrence can occur within days up to 2 months (for late recurrent cases) after treatment has been discontinued. Recurrence may be related to failure to produce IgG antibodies against toxin A.
In addition to metronidazole and vancomycin, there are new drugs including nitazoxanide, an antiparasitic and antibiotic agent used to treat Giardia intestinalis and cryptosporidium; rifaxamin, a gastrointestinal-selective antibiotic with broad spectrum against most Gram-negative and Gram-positive bacteria and anaerobes; and fidaxomicin, a macrocyclic antibiotic with narrow spectrum of activity against Gram-positive aerobic and anaerobic bacteria including C. difficile [2-4]. Other options include probiotics, which attempt to restore the balance of normal colonic flora (e.g. Saccharomyces boulardii), cholestyramine, intravenous immunoglobulin, monoclonal antibodies and vaccines [2, 4]. Surgical treatment with colectomy for refractory cases, or in patients with fulminant colitis, has been tried with generally poor outcomes . Lastly, and the focus of this review, is faecal microbiota transplantation (FMT). FMT attempts to restore the normal microbiome of the colon, and has achieved a cure rate reaching more than 90% .
Microbial communities populate all surfaces of the human body, but are present at their greatest density in the distal gut, reaching to 1014 bacterial cells, outnumbering the human somatic cells by approximately one order of magnitude. Recent research has offered new insights into the microbial diversity of gut microbiota, with the dominant organisms belonging to seven main bacterial divisions, i.e. Firmicutes, Bacteroides, Proteobacteria (notably, Helicobacter pylori in the stomach), Fusobacteria, Verrucomicrobia, Cyanobacteria and Actinobacteria. Of these, the Firmicutes (mostly members of the Clostridia class) and Bacteroides are the most abundant, comprising approximately 70% of all gut bacteria.
The distal gut, the primary site for CDI, is responsible for multiple physiological functions, including various aspects of energy metabolism, development and modulation of the immune system, pathogen resistance and clearance and control of epithelial proliferation and differentiation. Defence against invading pathogens occurs through competition for nutrients and adhesion sites, and via production of bacteriocins and bacterial-derived immunomodulatory molecules [6, 7]. The intestinal microbiota is a rich source of protective probiotics which produce novel antimicrobial and antipathogenic bacteriocin such as thuricin CD, produced by Bacillus thuringiensis, found to possess potent activity against C. difficile .
It is believed that gut microbiota play an important role as a regulator of heath and disease. A century ago Elie Metchnikoff, a Russian biologist, noted the large consumption of fermented milk in certain eastern European rural populations famed for their purported longevity. He then introduced sour milk into his own diet and noticed a subsequent improvement in his own health, thus forming the foundation for probiotics. He hypothesised that toxins produced by putrefactive microbes in the colon accelerate senescence, and that useful microbes could be used to replace harmful ones .
Studies have attempted to analyse stool before and after faecal transplantation. Tvede and Rask-Madsen observed an absence of bacteroides before FMT and during vancomycin therapy while patient were symptomatic. After bacteriotherapy, bacteroides were regularly cultured . Gustafsson et al. studied stool short chain fatty acids, the end products of intestinal bacterial metabolism of ingested complex carbohydrates. All short chain fatty acids were found to be reduced in the patient group compared with the healthy adults, and following treatment with faecal transplant . They also observed a reduction in butyrate and acetate, which are the main nutrient sources for the colon. More recently, two studies have shown a significant change in microbiota following faecal transplantation. Khoruts et al. demonstrated a reduction in Bacteroidetes and Firmicutes in a patient with C. difficile diarrhoea  . Following faecal transplantation, there was a rapid change in the patient's microbiota to a composition that was similar to that of the healthy donor for up to 4 weeks, the follow-up period. Grehan et al. undertook analysis of stool from 10 patients who underwent faecal transplantation . They observed a dramatic change in the recipient's microbiota to a composition similar to their donor's microbiota up to 24 weeks after transplantation. These patients received members of the Bifidobacterium genus, Bacteroides, Clostridium coccoides and the Clostridium leptum subgroup [6, 9-11]. These studies support the role of FMT in its ability to replenish the depleted flora but also introduce species whose bacteriocins may eradicate susceptible pathogens, and presumably restore multiple functions of microbiota .
The concept of FMT is not new. This idea was possibly first used in veterinary medicine by Italian anatomist Fabricius Aquapendente in the 17th century . However, there is evidence of much earlier use of human faecal transplantation during the Dong-jin dynasty in the 4th century in China . Ge Hong, a well-known traditional Chinese medicine doctor, described the use of human faecal suspension by mouth for patients who had food poisoning or severe diarrhoea. This yielded positive results and was considered a medical miracle that brought patients back from the brink of death [12, 13]. The first recorded use as a treatment for antibiotic-induced staphylococcal pseudomembranous colitis (PMC) in humans, dates back more than 50 years. In 1958, Eiseman and colleagues reported the successful treatment of four patients with severe PMC using a faecal enema. They were unaware that they were treating CDI because C. difficile was not recognised as a cause of PMC until 1978 . The first use of FMT for confirmed recurrent CDI was reported in 1983 by Schwan et al. in a 65-year old woman who thereafter had prompt and complete normalisation of bowel function  . Until 1989, the retention enema had been the most common technique for FMT, however alternative methods of faecal infusion subsequently were developed including NG or nasoduodenal tube, upper endoscopy and colonoscopy . Current indications for FMT are listed in Table 1 [14, 15]. Brandt et al. has even advocated for FMT as a primary treatment for CDI, noting superior outcome and less cost than antibiotic therapies . Others have presented guidelines, recommending FMT for the third recurrence of CDI . Notably, all studies to date have shown FMT to be safe and efficacious. Kassam et al. compared 11 studies involving FMT . FMT appeared effective for the treatment of recurrent CDI. They noted no reported adverse events with a range of follow up from weeks to years. In a recent randomised controlled trial, van Nood et al. compared vancomycin followed by bowel lavage followed by nasoduodenal FMT, with a standard vancomycin regimen, and a standard vancomycin regimen along with bowel lavage . They observed that the FMT arm was more effective for the treatment of recurrent CDI than vancomycin.
|Recurrent or relapsing CDI defined as a least three episodes of mild to moderate CDI and failure of a 6–8 week taper with vancomycin with or without an alternative antibiotic|
|At least two episodes of severe CDI resulting in hospitalisation and associated with significant morbidity|
|Moderate CDI not responding to standard therapy for at least 1 week|
|Severe CDI with no response to standard therapy after 48 h|
In the majority of reported cases, faeces donors were immediate family members or relatives; however, in few reports, donors were unrelated healthy volunteers or housemates. Earlier reports did not employ rigorous screening protocols, whereas more recently, there has been a trend towards standard protocol development. While exclusion criteria have been proposed (Table 2), suggested screening investigations include full blood count, liver function tests, screening for viral hepatitis, screening for human immunodeficiency virus, human T-cell lymphotrophic virus, Cytomegalovirus, Ebstein-Barr virus, Treponema pallidum, selective stool culture (e.g., for Salmonella, Shigella, Escherichia coli O157:H7, Yersinia enterocolitica, and Campylobacter), stool investigation for C. difficile toxin A and B, and microscopy for ova cysts and parasites [1, 3, 8]. The best route of avoiding iatrogenic complications and exposure is to complete a comprehensive screening whenever possible .
|Risk of infectious agent|
|High risk sexual behaviours|
|Use of illicit drugs|
|History of incarceration|
|Travel within the last 6 months to diarrhoea endemic area|
|Gastrointestinal comorbidities, including history of major surgery.|
|Factors affecting intestinal microbiota|
|Antibiotics within the preceding 3 months|
|Recent ingestion of a potential allergen|
The University of Minnesota Fairview Medical Center has moved away from the approach of using directly identified individualised donors. They have created a standardised laboratory process and utilise banked frozen faecal material. When 12 patients were treated for CDI with fresh donor material (10 identified donors and two standardised donors) compared with 33 patients treated with standardised frozen material, there were no significant differences in infection clearance for fresh vs. frozen samples, or in patient identified donors vs. standardised donors, and no adverse events were reported for either group .
Taking a different approach, clinicians often elect to utilise donations from individuals living in the same household, hypothesising that in close living arrangements, and particularly with intimate partners, potential pathogens would likely already have been widely shared by both parties. Considering the virulence of C. difficile, and the spore's ability to survive in the environment, utilising a donor from the same household as the infected patient might theoretically be a risk factor; however, the data thus far have demonstrated that transmitting donated stool containing C. difficile is not necessarily correlated with treatment success or failure .
Transplantation of fresh donated faeces is recommended to take place within 24 h and ideally within 6 h, although frozen stool samples from standardised donors have been thawed and colonoscopically administered 1–8 weeks after passage for treatment of recurrent CDI with similar success rats to fresh stool . The donor is instructed to take a single dose of osmotic laxative (e.g. citrate of magnesium or milk of magnesia) in the evening before the procedure to ensure the donor can reliably produce stool on the day of donation. The patient is instructed to stop vancomycin or metronidazole 1–3 days before the procedure. The use of bowel lavage was not included in all of the reviewed protocols. Presumed reasoning behind lavage is to enhance FMT success by flushing out the residual faeces, antibiotics, C. difficile bacteria, toxins and spores, prior to the administration of the donated flora. Clone library sequencing has shown that colonic mucosa associated microbiota composition is altered by standard bowel preparation lavage, enhancing the potential for FMT to provide a fresh start in repopulating the colonic habitat of the recipient. Polyethylene glycol electrolyte lavage was standard prior to colonoscopic FMT administration in almost all of the reviewed studies; however, a recent analysis noted that patients who received both bowel lavage and an antibiotic before the faecal transplant had the greatest rate of relapse (12%) [1, 3, 20].
It is recommended that a large volume donation suspension be attempted, since resolutions seem to be greatest (97%) when more than 500 ml is transferred compare to 200 ml (80%) . Relapse rates up to four times higher have been reported when less than 50 g of stool is donated. Approximately six to eight tablespoons of donor stool is added to a 1 l bottle of sterile water, saline or cow's milk, and shaken vigorously to homogenise. It is then filtered through gauze if necessary. The faecal suspension is then drawn up into 60 ml slip-tip syringes. There are also different posttransplant protocols – some centres recommend taking two tablets of loperamide immediately after lower gastrointestinal tract FMT, and again approximately 6 h later to maximise retention time of the donated microbiota. It is recommended to encourage the patient to retain stool (if possible) for 30–45 min after procedure. Patients should not resume vancomycin after procedure, and should be instructed to report any symptoms of recurrent infection. Stool should be tested only if symptoms recur.
Methodology of delivery can be tailored according to the clinician's evaluation of the most appropriate avenue considering the patient circumstances, the available physician and staff skill sets and equipment accessibility at the transplantation site . There is no general agreement on the best approach to delivering faecal microbiota or optimal volume because of absence of enough clinical studies and experience. Comparing the results of different methods of administration, there is some advantages and disadvantages for each method . In a meta-analysis, a total of 182 patients from 12 published studies were identified; 148 patients received FMT via colonoscopy and 34 patients received FMT via NG tube . The median age in the colonoscopy group as compared with the NG tube group was 72 and 82, respectively. There were differences regarding pre-FMT treatment for CDI; 134/148 (90.5%) received lavage with and without antibiotic in the colonoscopy group and 34/34 patients (100%) received antibiotic without lavage in NG tube group. While a higher stool volume was used for the colonoscopy group, the treatment efficacy did not differ significantly, with 93.2% success for the colonoscopy group as compared with 85.3% success for the NG tube group, p = 0.162. Recurrence of CDI after FMT was also similar in both the colonoscopy group (5.4%) vs. the NG tube group (5.9%). Despite procedural differences, FMT via colonoscopy or NG tube appeared to be highly effective and safe for the management of recurrent CDI. There were no deaths because of CDI in the NG tube group; however, four patients in the colonoscopy group died from CDI despite FMT (three with severe CDI). No significant adverse events associated with either NG tube or colonoscopy guided FMT procedures were observed .
Upper GI tract administration includes NG tubes, duodenal tube and endoscopy. In 2011 about 23% of all FMT procedures were provided by means of NG tube or gastroscopy, with a success rate of 76% . Most cases received one infusion. Upper GI tract FMT requires smaller volumes of suspension to be transplanted, partly related to concerns over aspiration. The NG tube is placed in the gastric antrum (this can be verified by X-ray), and a syringe can flush the suspension through the tubal system . Immediately after instillation, the tube is removed and the patient is allowed to go home and continue with their usual diet.
Use of a NG tube requires less patient preparation, clinical time, patient inconvenience and cost than retention enema or colonoscopy. NG tube insertion is also technically easier to perform, providing more extensive exposure of the gastrointestinal tract to donor faecal microbiota . Upper GI tract FMT is more appropriate for the paediatric population as well as those with severe comorbidities. However, this approach may carry the risk of vomiting and aspiration of feculent material, or predispose patient to bacterial overgrowth – particularly in elderly patients who are achlorhydric or have intestinal motility disorder . Disadvantages of this approach include the inability to place the sample directly at the affected sites in the colon, potential degradation of the sample by gastric and pancreatico-biliary secretions and slightly lower cure rate .
Lower GI tract delivery methods include colonoscopy and retention enemas. Colonoscopy is currently considered the first-line approach. After giving informed consent, the patient undergoes standard colonoscopy under sedation. Colonoscopic examination is performed and biopsy specimens are obtained if necessary. Approximately 20 ml of stool suspension is drawn up in a syringe and injected through the biopsy channel of the colonoscope every 5–10 cm as the scope is withdrawn, for a total volume of 250–500 ml. Another approach is to disperse the entire suspension at the most proximal aspect of the colon. The patient should be advised to refrain from defecating for at least 30–45 min after FMT [3, 22, 23]. This approach allows for direct visualisation of the entire colon, allowing instillation of the stool suspension in certain areas where C. difficile may predominate or hide (e.g. diverticuli), and allows delivery of a large volume of faecal suspensions throughout the entire colon with better retention. Bowel preparation itself may also decrease the patient's C. difficile organism and spore concentration. One disadvantage to this route of administration is the risk of colon perforation, especially in toxic colitis distension [1, 23].
No randomised controlled studies have been performed to assess stool transplantation in a systematic manner [24, 25]. Additionally, the FMT procedure has not been standard across studies. Of the studies reviewed, data on stool weight were not reported in 43%, data about pre-FMT treatment were not reported for 21%, and volume and number of suspensions given were not reported in more than 10%. Most patients received treatment with antibiotics and lavage prior to the procedure making it difficult to estimate the effect of FMT alone. Additionally, a relapsing infection could have been misclassified because of the lack of strain data. Finally, the small number of studies and sample size, along with the heterogeneity of the populations treated, affect outcome generalisability .
While FMT appears to be safe and generally well tolerated, it is important to understand patient perceptions of this modality. In a survey done which included hypothetical case scenarios to assess patient perceptions of the aesthetics of FMT and their willingness to consider it as a treatment option, 400 surveys were distributed, and 192/400 (48%) were returned complete . Seventy per cent of respondents were women, 59% were more than 49 years of age. When provided efficacy data, 162/192 (85%) chose to receive FMT and 29/192 (15%) chose antibiotics alone. When aware of the faecal nature of FMT, 16 respondents changed their choice from FMT to antibiotics alone, but there was no significant change in the total number choosing FMT 154/192 (81%, p > 0.001). Respondents rated all aspects of FMT at least somewhat unappealing, preferring to receive FMT in the hospital (48%) or physician office (39%), and 77% were willing to pay out of pocket for FMT . In the future, to gain greater appeal, centralised facilities may be capable of filtering and processing the donor material and shipping it in frozen and in lyophilised form as powder, including the encapsulated form used in carriers such as yogurt or favoured beverages [6, 8].
Mounting evidence supports the utility of FMT for severe and recurrent cases of CDI. Barriers that will need to be addressed are patient perceptions and fears, standard protocol development and further clinical trials . However, the unmet medical need for an efficacious treatment modality against the most severe and difficult-to-treat CDI warrants efforts to continue to research and develop this treatment option.