Systematic review: faecal microbiota transplantation therapy for digestive and nondigestive disorders in adults and children

Authors

  • S. Sha,

    1. State Key Laboratory of Cancer Biology & Xijing Hospital of Digestive Diseases, Fourth Military Medical University, Xi'an, Shaanxi Province, China
    Search for more papers by this author
    • These authors contributed equally to this work.
  • J. Liang,

    1. State Key Laboratory of Cancer Biology & Xijing Hospital of Digestive Diseases, Fourth Military Medical University, Xi'an, Shaanxi Province, China
    Search for more papers by this author
    • These authors contributed equally to this work.
  • M. Chen,

    1. State Key Laboratory of Cancer Biology & Xijing Hospital of Digestive Diseases, Fourth Military Medical University, Xi'an, Shaanxi Province, China
    Search for more papers by this author
    • These authors contributed equally to this work.
  • B. Xu,

    1. State Key Laboratory of Cancer Biology & Xijing Hospital of Digestive Diseases, Fourth Military Medical University, Xi'an, Shaanxi Province, China
    Search for more papers by this author
  • C. Liang,

    1. State Key Laboratory of Cancer Biology & Xijing Hospital of Digestive Diseases, Fourth Military Medical University, Xi'an, Shaanxi Province, China
    Search for more papers by this author
  • N. Wei,

    1. State Key Laboratory of Cancer Biology & Xijing Hospital of Digestive Diseases, Fourth Military Medical University, Xi'an, Shaanxi Province, China
    Search for more papers by this author
  • K. Wu

    Corresponding author
    1. State Key Laboratory of Cancer Biology & Xijing Hospital of Digestive Diseases, Fourth Military Medical University, Xi'an, Shaanxi Province, China
    • Correspondence to:

      Dr Kaichun Wu, State Key Laboratory of Cancer Biology & Xijing Hospital of Digestive Diseases, the Fourth Military Medical University, 17 Changle Western Road, Xi'an, Shaanxi Province 710032, China.

      E-mail: kaicwu@fmmu.edu.cn

    Search for more papers by this author

  • This commissioned systematic review was subject to full peer-review and the authors received an honorarium from Wiley, on behalf of AP&T.

Summary

Background

There has been growing interest in the use of faecal microbiota transplantation (FMT) for the treatment of gastrointestinal and nongastrointestinal diseases.

Aim

To review systematically the reported efficacy and safety of FMT in the management of gastrointestinal and nongastrointestinal disorders in adults and children.

Methods

The systematic review followed Cochrane and PRISMA recommendations. Available articles were identified using three electronic databases in addition to hand searching and contacting experts. Inclusion criteria were any reports of FMT therapy written in English.

Results

A total of 844 patients who had undergone FMT were identified from 67 published studies. The most common indications were refractory/relapsing Clostridium difficile infection (CDI) (76.3%) and inflammatory bowel disease (IBD) (13.2%). There has been only one placebo-controlled trial, a successful trial in 43 patients with recurrent CDI. Seven publications report FMT in paediatric patients with a total of 11 treated, 3 with chronic constipation and the remainder with recurrent CDI or ulcerative colitis (UC). 90.7% of patients with refractory/relapsing CDI were cured and 78.4% of patients with IBD were in remission after FMT. FMT therapy could also be effective in treatment of some nongastrointestinal disorders such as chronic fatigue syndrome. The only reported serious adverse event attributed to the therapy was a case of suspected peritonitis.

Conclusions

Although more controlled trials are needed, faecal microbiota transplantation therapy shows promise in both adults and children with gastrointestinal diseases such as CDI and IBD.

Introduction

Humans have been proposed to be ‘meta-organisms’ consisting of 10-fold greater numbers of bacterial than human cells. It is calculated that the intestinal bacteria typically encode 100-fold more genes than are present in the human genome.[1] The majority of microbes reside in the gut, where they have a profound influence on human physiology and nutrition, and are crucial for human health.[2, 3] Alteration of the intestinal microbiota has been associated not only with digestive tract dysfunction, but also with diseases beyond the gut including diabetes and metabolic syndrome.[4-10]

Given the role of the gastrointestinal microbiota in driving these disorders, it follows that manipulation of the microbiota represents a promising therapeutic strategy for conditions where the microbiota is known to be altered. Faecal microbiota transplantation (FMT) therapy involves infusing intestinal microorganisms (present in a suspension of healthy donor stool) into the intestine of a sick patient to restore the balance of the normal intestinal microbiota. Brandt et al.[11] noted that this idea was possibly first used in veterinary medicine by the Italian anatomist Fabricius Aquapendente in the 17th century. Zhang et al.[12] reported even earlier literary evidence of FMT therapy during the Dong-jin dynasty in the 4th century in China. In the recent literature, it was first reported by doctors in Denver, who administered faeces by enema to their patients with fulminant, life-threatening pseudomembranous enterocolitis in 1958.[13] Over the subsequent decades since then, leading experts from US, Canada and Australia have recently showed that FMT therapy is safe and effective for recurrent Clostridium difficile infection (CDI).[14, 15] Recently, there has been a proliferation of reports of FMT as a therapy, not only in CDI but also for other diverse indications including inflammatory bowel diseases (IBD), irritable bowel syndrome (IBS), metabolic syndrome, neurodevelopmental disorders, auto immune diseases, allergic diseases, pulmonary diseases, constipation and even colorectal carcinoma prevention.[16, 17]

FMT has been reported from many countries and the methods used, screening of donors and patient groups treated with this therapy have varied greatly. Several studies reviewed the use of FMT therapy in the management of IBD and other gastrointestinal disease in recent years,[15, 18] but none of these has tried to compare the effectiveness of FMT therapy between adults and children. Because the intestinal microbiota alters with lifestyle, diet and age in different individuals,[19] we review here the use of FMT therapy for diseases in adults and children since the first report from 1958.

Methods

Search strategy

The systematic review was undertaken in line with the recommendations of the guidelines of the Cochrane Handbook for Systematic Reviews of Interventions (http://handbook.cochrane.org/). This systematic review adheres to the relevant criteria of the PRISMA statement (Preferred Reporting Items for Systematic reviews and Meta-Analyses)[20] and the meta-analysis of observational studies in Epidemiology (MOOSE) statement.[21, 22] A comprehensive search strategy was developed by a research librarian to identify any potentially relevant studies on FMT therapy. The search terms were derived from the previously published study.[18] The paediatric patients were defined as aged from 0 to 18 years. Electronic databases searched included the Cochrane Library, PubMed and EMBASE. The search also included international gastroenterology conference proceedings and the bibliographies of key reviews. In addition, hand searching of the reference lists of relevant reviews and included studies was undertaken to identify further relevant references. All studies reported in English were included because there is a lack of controlled trials in this area. The last search was run on 27 May 2013.

Study selection

Two investigators (S. S. and C. M.) independently assessed the title, abstract and key words of every record for eligible publications. Disagreements were resolved by consensus. The inclusion criteria were patients treated with FMT therapy. Primary studies were included if they (i) included patients of any age with FMT therapy, (ii) compared FMT with standard care or reported safety and efficacy/effectiveness outcomes of FMT in a series of patients without a control group, (iii) certain publication types (i.e. letters to the editor, abstracts and proceedings from scientific meetings) were also included, if a full set of data were available from the authors, (iv) were English full text studies published between 1958 and 2013. Studies were excluded if they (i) focused on the treatment using antibiotics or other probiotics, (ii) did not report original data of the FMT therapy procedure, (iii) reports describing the use of a cultured bacterial suspension rather than human faeces, (iv) were interviews and reviews.

Data extraction and synthesis

Once eligibility was determined, the next step was retrieval of the full text of potentially relevant trials. Two reviewers (S. S. and C. M.) independently abstracted data from selected publications using a standardised pre-tested form. If they had different opinions, these were resolved by discussion with the third reviewer (L. J.). When multiple publications related to the same patient group,[11, 23-35] the most complete data were analysed.[11, 24, 26, 28, 30, 32, 35] Individual patient information was extracted from each study. The following information was retrieved: (i) characteristics of patients (including number of patients, age, gender and disease types); (ii) characteristics of intervention (including delivery method, route of instillation, relationship between the recipients and the donors); (iii) type of outcome measures (included duration of follow-up, change in frequency and severity of symptoms, resolution, treatment failure, relapse, death et al.); (iv) safety outcomes included any adverse events associated with the FMT procedure. Study time, country and study design were also abstracted. Data extracted from each study were synthesised using a narrative approach.

Results

Description of included studies

A total of 16 610 nonduplicated articles were identified in the search. The titles and abstracts were reviewed and only 98 were deemed potentially eligible. Following review of the full article, 67 fulfilled the inclusion criteria (Figure 1). The majority of reports were journal articles (52.2%); followed by abstracts (37.3%) and letters (10.4%). 62.7% (42/67) were case series, the others were case reports. Twenty eight of these were published since 2012, demonstrating the recently renewed interest in this area (Tables 1-4).

Table 1. Summary of FMT for treating CDI
Characteristics of studiesCharacteristics of patientsCharacteristics of interventionCharacteristics of outcomes
Author (Year)Type (country)Number (Male/Female)Age median (range)Faeces suspension/volume infused (route of FMT)FrequencyDonor relationshipOutcomeSuccess rateAdverse events
  1. FMT, faeces microbiota transplantation; CDI, Clostridium difficile infection; PEG, percutaneous endoscopic gastroscopy.

  2. a

    Two deaths: 1 after development of peritonitis (3 days post-FMT); 1 of pneumonia (14 days post-FMT).

  3. b

    Transient and included sore throat, flatulence, rectal discomfort, nausea, abdominal cramping, bloating, headache and abdominal pain.

  4. c

    Developed constipation in the initial months after the procedure.

  5. d

    Approximately a third noted some irregularity of bowel movements and excessive flatulence, but resolved in clinic follow-up.

  6. e

    The patient died because of ongoing renal failure and acute respiratory failure.

Eiseman[13] (1958)Journal article (US)4 (3/1)56 (45–68)/(retention enemas)1–3Healthy adultsSevere diarrhoea ceased within 48 h4/4 (100%)None
Cutolo LC[76] (1959)Journal article (US)1 (1/0)65Two ounces of faeces and 10 ounces of yogurt in a quart of normal saline) (Cantor tube)3/day for 7 daysHealthy humansThe first formed stool was passed 36 h later, after 48 h the diarrhoea ceased. A response was obtained and both clinical and laboratory evidence revealed the elimination of the staphylococcus from the intestine1/1 (100%)Died of upper GI haemorrhage (not related to the FMT therapy)
Fenton[77] (1974)Journal article (Canada)1 (0/1)57-/500 mL saline (enema)1/Marked improvement occurred and continued to complete resolution of the clinical and sigmoidoscopic abnormalities within 4 days.1/1 (100%)None
Schwan[24] (1984)Journal article (Sweden)1 (0/1)67450 mL mixture contain faeces and saline (retention enemas)2 (over three consecutive days)HusbandPrompt and complete normalisation of the bowel function with disappearance of IBS symptoms. Stools of normal consistency, colour and smell have thereafter been passed daily or every other day. Weight gain 6 kg1/1 (100%)None
Tvede[78] (1989)Journal article (Denmark)2 (1/1)60 (59–60)50 g faeces suspended in 500 mL saline (Enema)1–2Husband and daughterPatient 1: complete clinical recovery with eradication of C. difficile and its toxin after 1 infusion; Patient 2 had 2 infusions but did not respond (but did respond to a cultured bacterial mixture)1/2 (50%)None
Borody[44] (1989)Letter (Australia)1 (0/1)35Retention enemas (/)//The patient appear to be ‘cured’ by resolution of symptoms dramaticly.1/1 (100%)None
Patterson-1[79] (1994)Letter (Australia)1 (0/1)39Single daily retention enemas for 3 days (400 mL consisting of 200 mL stool mixed with 200 mL saline (Rectal tube)3HusbandNo recurrence of diarrhoea in 2 years1/1 (100%)None
Patterson-2[79] (1994)Letter (Australia)6 (-/-)56 (30–80)Single daily retention enemas for 3 days (400 mL consisting of 200 mL stool mixed with 200 mL saline) (Rectal tube)3RelativeAll patients experienced rapid resolution of disabling persistent C. difficile infection without relapse6/6 (100%)None
Persky[80] (2000)Journal article (US)1 (0/1)60(500 mL containing stool mixed in saline) (colonoscopy)1HusbandImmediate and complete resolution of diarrhoea with normal bowel movements that was maintained long term. Repeat C. difficile toxin assay negative1/1 (100%)None
Faust[81] (2002)Abstract (Canada)6 (1/5)53 (37–74)/(/)1Family members (spouse 4, brother 1, son 1)All patients responded promptly and continued to be asymptomatic; 4/6 patients C. difficile toxin negative6/6 (100%)None
Aas[63] (2003)Journal article (US)18 (5/13)73 (51–88)30 g faeces homogenised in 50–70 mL saline (Single 25 mL dose of stool suspension) (Nasogastric tube)1Family members or clinic staff15/18 patients asymptomatic for duration of follow-up. 13 patients were C. difficile negative (2 patients not tested); 2 died of unrelated illnesses; 1 treatment failure (C. difficile positive)15/16 (94%) (who survived) a
Broody (CD SYNDROMES)[61] (2003)Abstract (Australia)24 (11/13)19–59200–300 g stool diluted in 200–300 mL saline (The suspension was infused into the colon) (colonoscope and/or rectal enema and/or nasojejunal tube (Combination of colonoscopy and rectal enema was the most common (46%) delivery method))Daily for 1 (3/24, 13%), 5 (11/24, 46%) or 10 days (10/24, 42%)Relatives or unrelated healthy individualsEradication of C. difficile was confirmed by negative Cd toxin and culture results in 20/24 patients (83%, P < 0.0001) post-treatment. 2/24 unsuccessful, 2/24 nonresponse, (including 1/24 recurrent)20/24 (83%) b
Wettstein[38] (2007)Abstract (Australia)16 (5/11)11–87200–300 g/200–300 mL Saline with added psyllium [Colonoscopy (day 1) Enema (between 5, 10, or 24 days)]5–24Relatives or unrelated healthy individualsEradication of C. difficile was confirmed by negative Cd toxin A or B and culture results in 15/16 patients (93.5%) 4–6 weeks post-treatmentResolution in 15/16None
You[82] (2008)Letter (US)1 (1/0)6945 g/300 mL normal saline (Retention enema)1DaughterThe patient's blood pressure stabilised, the leucocyte count normalised, and oliguria resolved and both vasopressors and continuous venovenous hemofiltration was discontinued. The patient's bowel function returned, and abdominal distention decreased1/1 (100%)None
Keller[83] (2009)Abstract (Netherlands)11 (-/-)/>100 g/300–400 mL saline (infusion of suspension of donor faeces) (in jejunum (nasoduodenal tube) or in coecum and colon ascendens (via colonoscope))1/Successfully treated 11 patients with multiple recurrences of CDI11/11 (100%)None
Maccanochie[84] (2009)Journal article (UK)15 (1/14)81.5 (68–95)30 g faeces in 150 mL saline (30 mL of faecal fluid was administered) (Nasogastric tube)1 (14/15) or 2 (1/15)Healthy related volunteersPatients were symptom free. 2 no responses; 2 relapsed (1 responded to the 2nd FMT)12/15 (80%)None
Rubin[85] (2009)Letter (US)16 (-/-)70–99~30 g or 2cm[3]/50–70 mL saline (30–60 mL suspense) (Nasogastric tube)1Family member14/16 ambulatory patients are able to return to normal diet and activities immediately after the procedure14/16 (87.5%)None
Arkkila, P. E.[86] (2010)Abstract (/)37 (-/-)69 (24–90)20–30 mL mixed with 100–200 mL of water (colonoscopy)1–2Related to the recipient34/37 (92%) patients were cured patients had relapse after 5–12 months after receiving new antibiotic treatment and they got successful faeces reinfusion thereafter. One noncured patient died after 1 month due to the toxic megacolon34/37 (92%)None
Silverman[26] (2010)Journal article (Canada)7 (4/3)65 (30–88)50 mL faeces in 200 mL saline (250 mL of faecal fluid was administered) (Retention enema)1Family memberAll of the patients were successfully cured7/7 (100%)None
Khoruts[60] (2010)Journal article (US)1 (0/1)6125 g/300 mL saline (250 mL of faecal fluid was administered) (colonoscope)1HusbandAt 1 month after bacteriotherapy, stool studies were culture negative for C. difficile1/1 (100%) c
Rohlke[87] (2010)Journal article (US)19 (2/17)49 (29–82)-/350 mL saline (200–300 mL of faecal fluid was administered) (colonoscopy)1 (majority) or 2 (1 patient)Intimate domestic partners, family members and close friends.All of the patients in this study maintained prolonged periods free of symptoms and are considered ‘cured’ after treatment with Faeces Flora Reconstitution19/19 (100%)/
Yoon[88] (2010)Journal article (US)12 (3/9)66 (30–86)Unknown volume faeces in 1000 mL saline (250–400 mL of faecal fluid was administered) (colonoscopy)1Family member or partnerAbsence of diarrhoea, cramps, and fever. All patients experienced a durable clinical response to faeces transplantation12/12 (100%)None
Garborg[89] (2010)Journal article (Norway)40 (19/21)75 (53–94)50–100/250 saline (200 mL of faecal fluid was administered) [Gastroscopy (38) or colonoscopy (2)]1 (34/40) or 2 (6/40)Close relatives or other household membersA total of 33/40 patients (4 patients responded to the 2nd FMT) were successfully treated33/40 (83%)None (5 unrelated deaths 3 weeks to 2 months post-FMT)
Kelly[90] (2010)Abstract (US)12 (1/11)55.6 (19–80)6–8 tablespoons of donor stool was added to 1 L of sterile water (740 mL (range 500–960 mL) of faeces suspension was delivered) (colonoscope)1Partner or family memberTen have remained symptom free. Two had diarrhoea after the procedure, but both were C. difficile negative. One responded to treatment with a fibre supplement and the other resumed vancomycin. None have had a documented recurrence of CDI to date12/12 (100%)None
Mellow[31] (2010)Journal article (US)13 (7/6)67 (32–87)-/300–600 mL saline (colonoscopy)1Healthy donorsA total of 12/13 patients were successfully treated with diarrhoea resolved. 3 patients have died (B strep pneumonia 1 month after FMT; superior mesenteric vein thrombosis 5 months after FMT; ovarian cancer 7 months after FB)12/13 (%)None (3 unrelated deaths 1–7 months post-FMT)
Miller[91] (2010)Abstract (US)2 (0/2)42 (34–50)Fresh stool was liquefied and delivered (colonoscopy)1Sister/husbandOne patient had an immediate improvement in symptoms and has been infection free for 9 months2/2 (100%)None
Russell[14] (2010)Journal article (US)1 (0/1)230 g/50–75 mL saline (25 mL mL of faecal fluid was administered) (temporary nasogastric tube)1FatherThe abdominal symptoms and diarrhoea resolved and did not recur during 6 months of monitoring. The stool tested negative for C difficile toxin 2 weeks, 3 months, and 6 months after the procedure1/1 (100%)None
Girotra[92] (2011)Abstract (US)3 (0/3)56.3 (31–85)/(colonoscopy and enteroscopy)1Healthy donorsAll three patients had complete resolution of symptoms within a few days after administration of faeces bacteriotherapy. complete resolution of symptoms3/3 (100%)None
Wilcox[93] (2011)Abstract (Portland)13 (4/9)69.4 (27–93)/(colonoscope into the patient's caecum)1Family members or friendsFMT appears to be effective, tolerated well by patients, and without complications13/13 (100%)None
Brandt[11] (2012)Journal article (US)77 (21/56)65 (22–87)300–700 mL (colonoscopic)1 or 2 (4 patients)Spouse/partner, 46; 1st degree relative, 19; friend, 9; other relative, 2; unknown, 1diarrhoea resolved in 82% and improved in 17% of patients within an average of 5 days after FMT. The primary cure rate was 91%. Seven patients either failed to respond or experienced early CDI recurrence (≤90 days)70/77 (90.9%)One patient was not treated and died in hospice care of unclear cause
Gallegos-Orozco[94] (2012)Journal article (US)1 (1/0)71-/400 mL saline (400 mL) of faecal fluid was administered (colonoscopic)1BrotherThe patient's diarrhoea and abdominal pain resolved1/1 (100%)None
Garg[39] (2012)Abstract (US)1 (1/0)20-month-old/(colonoscopy)1MotherThe patient has now remained symptom free with complete resolution of diarrhoea, haematochezia and had gained weight consistently for the past 3 months1/1 (100%)None
Hamilton[30] (2012)Journal article (US)29 (9/20)64.7 ± 3.350 g/250 mL saline (220–240 mL of faecal fluid was administered) (Colonoscopy (1 patient used upper push enteroscopy))1 (37/43) or 2 (6/43)Mothers, daughters, sons, wives, husbands and friendsOne other patients were treated with a second infusion, and all cleared the infection bringing the overall success rate to 100%.29/29 (100%) d
Jorup-Ronstrom[95] (2012)Journal article (Sweden)32 (12/20)75 (27–94)30–60 mL (Enema and colonoscopy)1–3A healthy adult22 patients were cured and 10 patients failed to fully respond to treatment. Later in the study, the culture was given by colonoscopy in patients who failed after the simple enema procedure. 4 out of 5 healed. 6/32 relapse26/32 (81.3%)None
Kahn[40] (2012)Letter (US)1 (1/0)16-months(140 mL of faecal fluid was administered) (colonoscopy)1MotherWithin 24 h the patient's symptoms resolved completely. One week post-FMT, Clostridium difficile toxin PCR was negative and 2 months post-FMT he remains asymptomatic1/1 (100%)None
Kurtz[96] (2012)Journal article (US)1 (0/1)78/(colonoscopy)1HusbandThe patient failed to respond to the stool transplant. Biopsies from a colonoscopy revealed concomitant cytomegalovirus (CMV) infection. The patient underwent an emergency total colectomy, and oral valganciclovir0/1 (0%)None
Mellow[97] (2012)Abstract (US)1 (1/0)39/(colonoscopy)1WifeOver the next several days, diarrhoea resolved and stool for C. difftoxin A&B was negative on day seven post-FMT. Diarrhoea did not recur after 1 year1/1 (100%)None
Neemann[98] (2012)Journal article (US)1 (0/1)2130 mL/- in saline (suspension) (nasojejunal instillation)1HusbandThe patient had no further diarrhoea or haematochezia. Her abdominal pain and distention had resolved completely, and she was eating a regular diet. Repeat C. difficile toxin and antigen EIA assays were negative at 3 and 11 days after faeces transfer, and the patient remained well at 2 months after treatment1/1 (100%)None
Kelly[71] (2012)Journal article (US)26 (2/24)59 (19–86)6–8 tablespoons/1 L sterile water (or saline) (500–960 mL was administered) (colonoscopy)1Intimate partner or adult family member24/26 patients have remained free of significant diarrhoea or CDI. One experienced loose stool and resumed vancomycin despite remaining C. difficile negative; she developed CDI recurrence 11 months post-FMT after taking cephalexin. Another had diarrhoea 2 months post-FMT. Stool was not tested for C. difficile; she received 1 week of vancomycin and CDI did not recur after this24/26 (92.3%)None
Kassam[99] (2012)Journal article (Canada)27 (14/13)69.4 (26–87)150 g/300 mL water (The supernatant component was administered rectally by enema) (retention enema)1 (22/27) or 2 (5/27)Two healthy volunteers22 resolved within 24 h of transplant. 5 patients underwent a second FMT because of ongoing diarrhoea; 3 had symptom resolution and 2 continued to experience diarrhoea despite 2 FMTs25/27 (93%)None
Matilla[100] (2012)Journal article (Finland)70 (28/42)73 (22–90)20–30 mL/100–200 mL water (100 mL suspension) (colonoscopy)1 (68/70) or 2 (2/70)Close relatives, household members, or healthy volunteerDuring the first 12 weeks of follow-up evaluation the transplantation resulted in the resolution of symptoms in 66 patients. During the 1-year follow-up period, 4 patients with an initial favourable response had a relapse after receiving antibiotics for unrelated causes. 2 were treated successfully with another faeces transplantation and 2 with antibiotics for CDI66/70 (94.3%)10 patients died of unrelated illnesses during the 1-year follow-up period
Rubin[41] (2013)Journal article (US)75 FMT courses (49/26) (74 patients)63 (6–94) two paediatric patients (age 6 and 8)~30 g (~3 cm[3])/50–70 mL saline (25 mL) (nasogastric tube, or gastroscope or through a PEG tube)1 or 2 (1 patient)Healthy close household memberFifty-nine FMT courses resulted in clinical resolution of diarrhoea for a primary cure rate of 79%. diarrhoea relapsed following 16 FMT courses; in 9 of these cases diarrhoea subsequently resolved after a single course of vancomycin. (paediatric patients: one experienced a clinical resolution following the FMT, while the other had a clinical relapse)59/75 (79.7%)None
Trubiano[101] (2013)Journal article (Australia)1 (0/1)7530 g/70 mL saline(30 mL suspension of donor faeces) (delivered into the jejunum through gastroscopy)1HusbandThe C. difficile-related symptoms were resolved. An abdominal computed tomography in the days before death showed no evidence of colitis, whereas stool cultures for C. difficile remained negative for culture and toxin on days 14, 20 and 30 after transplant1/1 (100%) e
Van Nood[37] (2013)Journal article (Netherlands)16 (8/8)73 ± 13Faeces were diluted with 500 mL of sterile saline (0.9%) (nasoduodenal tube)1(13/16) 2 (3/16)VolunteersThe infusion of donor faeces was significantly more effective for the treatment of recurrent C. difficile infection than the use of vancomycin. 13 (81%) had resolution of C. difficile-associated diarrhoea after the first infusion. The 3 remaining patients received a second infusion with faeces from a different donor, with resolution in 2 patients15/16 (94%)Mild diarrhoea and abdominal cramping on the infusion day
Table 2. Summary of FMT for treating IBD
Characteristics of studiesCharacteristics of patientsCharacteristics of interventionCharacteristics of outcomes
Type of IBDAuthor (Year)Type (country)Number (Male/Female)Age median (range)Faeces suspension/volume infused (route of FMT)FrequencyDonor relationshipOutcomeSuccess rateAdverse events
  1. FMT, faecal microbiota transplantation; IBD, inflammatory bowel disease; UC, ulcerative colitis; CD, Crohn's disease; /, not known.

  2. aTemporary increase in CRP, flatulence, vomiting, common cold, itchiness, erythema, paraesthesia on the hip, collapse and blisters on the tongue was reported.
UCBennet, J. D[102] (1989)Letter (US)1 (1/0)/Large-volume (Enema)1/Patient symptom-free (no bloody diarrhoea, cramping, tenesmus, skin lesions or arthritis) for the first time in 11 years without any medication1/1 (100%)None
UCBorody[44] (1989)Letter (Australia)1 (1/0)45Retention enemas (/)//The patient appear to be ‘cured’ by resolution of symptoms dramaticly.1/1 (100%)None
UCBroody[103] (2001)Journal article (Australia)3 (1/2)28, 38, 42Donor stool suspended in 200 mL saline with one tablespoon of psyllium husks (retention enemas)Single daily for 5 daysHealthy adultsDisappearance of symptoms within 1 month, maintained long term without medications.3/3 (100%)None
UCBroody[104] (2003)Journal article (Australia)6 (3/3)36 (25–53)200–300 g stool diluted in 200–300 mL saline (Retention enema)51 Female partner/1 Unrelated Male/1 Brother in law/3 brotherAll patients showed complete resolution of symptoms by 4 months and were able to cease all UC medications. No clinical, colonoscopic or histological evidence of UC at long-term follow-up.6/6 (100%)None
UC and ITPBorody[105] (ITP) (2011)Abstract (Australia)1 (0/1)39/(/)//In the months following FMT the patient experienced an abrupt rise in her platelet count. A marked reduction in UC symptoms, passing 2-3 semi-formed stools daily, without bleeding or urgency1/1 (100%)None
UCBorody, T.[106] (2012)Abstract (Australia)62 (40/22)45.4/(/)1/Overall, 91.9% of patients responded to FMT. Of these, 67.7% of patients (42/62) achieved complete clinical remission, and 24.2% of patients (15/62) achieved partial response. The remaining 8% (5/62) were treatment failures. Improvement in CRP and ESR correlated with clinical response observed in FMT patients57/62 (91.9)None
UCBrandt[15] (2012)Abstract (US)6 (2/4)44 (26–73)/(colonoscopy, self-administered faecal enemas)1First-degree relatives, spouses or otherwise relatedAll 6 of our patients reported improvement after FMT. Maximal benefit was seen in the subgroup of patients with concomitant CDI (n = 2) and newly diagnosed UC in the setting of antibiotic use (n = 1). FMT was not as effective in the 3 remaining patients, whose UC onset or worsening was not associated with CDI or antibiotic use6/6 (100%)None
UCKunde[43] (2012)Abstract (US)3 (-/-)>7 and <218 oz faecal enemas (enema)For 5 days over 1 week period/Two of three subjects achieved clinical response by reduction in paediatric ulcerative colitis activity index (PUCAI) of 15 points or more. One of them had complete resolution of disease activity3/3 (100%)Self-limiting symptoms from mild to moderate
UCAngelberger S[28] (2012)Abstract (Austria)5 (3/2)/23.8 g (16.7–25 g) (nasojejunal tube); 20 g (6 g–21.7 g) (enema) stool was administered (nasojejunal tube and enema)3 (daily for 3 days)Healthy adultsThe general well-being improved from poor to very well at week 12 in 3 patients. The FMT might be safe but activates a temporary systemic immune response3/5 (60%) a
UCKump[64] (2013)Abstract (Austria)6 (-/-)//(colonoscopy)1/Within the first 14 days all patients experienced a reduction in stool frequency. However, none of the 6 patients achieved a complete remission and only 2 of the 6 patients had a durable improvement in their clinical UC scores. Subsequently two patients underwent total colectomy and one additional patient was treated with ciclosporin2/6 (33.3%)1 patient had a self-limiting episode of fever after FMT
CDBorody[44] (1989)Letter (Australia)1 (1/0)31Retention enemas (/)//The patient appear to be ‘cured’ by resolution of symptoms dramaticly.1/1 (100%)None
CDGrehan[45] (2010)Journal article (Australia)1 (1/0)57NR/250 mL saline (200–400 mL of faecal fluid was administered) (Colonoscope, nasojejunal tube, enema)5–15 (9.1 ± 3.25)Unrelatedmale///
CDVermeire S[65] (2012)Abstract (/)4 (1/3)37.5 (29–50)-/- (200 g) (nasojejunal tube, colonoscopy)3Healthy donorNo change or ceased due to not working, none of the patients experienced clinical, biologic or endoscopic benefit0/4 (0%)Transient fever developed in 3 out of 4 patients
Chronic refractory pouchitisLandy[107] (2013)Abstract (UK)8 (-/-)/30 g/50 mL saline (-) (nasogastric administration)1/FMT via nasogastric administration was not effective in achieving clinical remission for chronic refractory pouchitis with no change in PDAI or Cleveland global quality of life score (CGQoL) identified at 4 weeks after FMT0/8 (0%)None
UC/CDBorody T(IBD)[59] (2011)Abstract (Australia)3 (1/2)33 (19–57)/(Enema)33–69/Severe mixed IBD were successful reversal.3/3 (100%)None
Table 3. Summary of FMT for treating CDI in IBD
Characteristics of studiesCharacteristics of patientsCharacteristics of interventionCharacteristics of outcomes
Author (Year)Type (country)Number (Male/Female)Age median (range)Type of IBDFaeces suspension/volume infused (route of FMT)FrequencyDonor relationshipOutcomeSuccess rateAdverse events
  1. FMT, faecal microbiota transplantation; CDI, Clostridium difficile infection; IBD, inflammatory bowel disease; CD, Crohn's disease; PCR, polymerase chain reaction; UC: ulcerative colitis.

  2. a

    Approximately a third noted some irregularity of bowel movements and excessive flatulence, but resolved in clinic follow-up.

You[108] (2011)Abstract (US)1 (0/1)33CD75 g/200 mL saline (200 mL of faecal fluid was administered) (Nasogastric tube)1HusbandRepeat Clostridium difficile PCR test was negative. The patient's abdominal pain markedly improved, and her haematochezia resolved1/1 (100%)None
Duplessis[109] (2012)Journal article (US)1 (0/1)33CD75 g/200 mL saline (200 mL of faecal fluid was administered) (Nasogastric tube)1HusbandBowel movements became less frequent, without haematochezia, and her abdominal pain improved and fever abated. Two days after stool transfer, a repeat C. difficile PCR was negative. She successfully resumed Crohn's colitis therapy.1/1 (100%)None
Neelakanta A[110] (2012)Abstract (US)1 (0/1)27CD/(Colonoscopy)1FriendAble to commence Adalimumab. Diarrhoea improved but did not completely resolve despite negative stool toxin studies for Clostridium difficile 2 weeks post-procedure.2/2 (100%)None
Hamilton[30] (2012)Journal article (US)14 (3/11)44.6 ± 5.84/14UC, 6/14 CD, 4/14 lymphocytic colitis)50 g/250 mL saline (220–240 mL of faecal fluid was administered) [Colonoscopy (1 patient used upper push enteroscopy)]1 or 2Mothers, daughters, sons, wives, husbands and friendsThree other patients were treated with a second infusion, and all cleared the infection bringing the overall success rate to 85.7%.12/14 (85.7%) a
Singh[42](2012)Abstract (US)1 (0/1)6UCFresh donor stool was mixed with saline then superfiltrated to 30 cc suspension (nasogastric tube)1MotherStool calprotectin decreased from 504 at baseline to 76 by week 12. C. diff antigen and toxin were cleared by 3 weeks. Microbiome analysis of the donor stool and the patient's stool before and at 3 periods after bacteriotherapy are pending and will be finalised shortly.1/1 (100%)None
Watson[62] (2012)Abstract (/)1 (1/0)78UC/(Colonoscopy)1WifeNo diarrhoea or abdominal pain.1/1 (100%)UC flare
Neelakanta A[110] (2012)Abstract (US)1 (1/0)39UC/(Colonoscopy)1Family memberHe has not had any relapse of CDI but still has evidence of active IBD in rectosigmoid on colonoscopy done a year after FMT.1/1 (100%)None
Zainah, H[111] (2012)Journal article (US)1 (1/0)51UC300 mL/- warm water (300 mL) (Colonoscopy)1WifeThe patient remained symptom free for 8 months and was able to stop oral vancomycin without CDI recurrence.1/1 (100%)None
Table 4. Summary of FMT for treating other diseases
Characteristics of studiesCharacteristics of patientsCharacteristics of interventionCharacteristics of outcomes
IndicationAuthor (Year)Type (country)Number (Male/Female)Age median (range)Faeces suspension/volume infused (route of FMT)FrequencyDonor relationshipOutcomeSuccess rateAdverse events
  1. FMT, faecal microbiota transplantation; MS, multiple sclerosis; IBS, irritable bowel syndrome; AAD, antibiotic-associated diarrhoea.

Chronic

constipation

Borody[44] (1989)Letter (Australia)1 (0/1)31/(Retention enemas)//The patient appear to be ‘cured’ by resolution of symptoms dramaticly.1/1 (100%)None
 Andrews PJ[46] (1992)Journal article (Australia)1 (0/1)43/(Enema)Twice in two consecutive daysHusbandResolution of abdominal bloating and an unexplained disappearance of reflux symptoms with a pronounced reduction in the frequency of tension headaches1/1 (100%)None
 Broody[47] (2001)Journal article (Australia)3 (1/2)16,17,68Donor stool suspended in 200 mL saline with one tablespoon of psyllium husks (retention enemas)Single daily for 5 daysHealthy adultsLong-term restoration of normal bowel function to 1–2/day without laxatives.3/3 (100%)None
 Grehan[45] (2010)Journal article (Australia)4 (1/3)36.8(22–50)NR/250 mL saline (200–400 mL of faecal fluid was administered) (Colonoscope, nasojejunal tube, enema)5–15 (9.1 ± 3.25)Unrelatedmale///
 Borody T[48] (MS) (2011)Abstract (Australia)3 (2/1)46 (29–80)/(/)5–10/Constipation was complete resolution. MS also progressively improved3/3 (100%)None

IBS or abdominal

pain

Grehan[45] (2010)Journal article (Australia)5 (4/1)46 (38–56)NR/250 mL saline (200–400 mL of faecal fluid was administered) (Colonoscope, nasojejunal tube, enema)5–15 (9.1 ± 3.25)Unrelatedmale///
 Borody[44] (1989)Letter (Australia)1 (0/1)21/(Retention enemas)//The patient appear to be ‘cured’ by resolution of symptoms dramaticly.1/1 (100%)None
AADGustafsson A[34] (1999)Journal article (Norway)32 (14/18)27–8920 mL containing 5–10 g homogenised faeces in ordinary pasteurised cow's milk (Enema)1 or 2 (1 patient)Healthy adult30 patients were regarded as clinically normalised within 1 week and, in some patients, even after 1–2 days. Two patients received additional treatment because of persistent symptoms; one was given a second enema (day 3) and metronidazole (day 8), and the other patient vancomycin (day 4), and the diarrhoea ceased after 3–4 days.30/32 (93.8%)None

Metabolic

syndrome

Vrize[54](2012)Journal article (Netherlands)18 (-/-)50 ± 3The faeces was covered with sterile saline (500 mL 0.9% NaCl) (gastroduodenal tube)1Lean healthy Caucasian malesInsulin sensitivity of recipients increased (median rate of glucose disappearance changed from 26.2–45.3 mol/kg/min; P < 0.05) along with levels of butyrate-producing intestinal microbiota.18/18 (100%)None
Figure 1.

Flow chart of studies of faecal microbiota transplantation therapy in adults and children.

Patient characteristics

Worldwide, at least 844 cases of FMT have been reported from 10 countries (Tables 1-4). The most common indication is refractory/relapsing CDI (644/844, 76.3%) (Table 1), others including IBD (111/844, 13.2%) (94 UC, 14 CD and 3 UC/CD) (Table 2), CDI in IBD (21/844, 2.5%) (Table 3), chronic constipation (12/844, 1.4%) (Table 4), no clear diagnoses but with symptoms including IBS, diarrhoea, or abdominal pain (6/844, 0.7%) (Table 4), antibiotic-associated diarrhoea (AAD) (32/844, 3.8%) (Table 4) and metabolic syndrome (18/844, 2.1%) (Table 4). Taken together, 307/739 (41.5%) of patients were male (eight studies did not report the gender ratio). Follow-up ranged from 1 day to 15 years after the transplantation. FMT has been described in patients as young as 16 months old to patients over 95 years of age. There have been seven publications on FMT for the treatment in children since 2001 and these included 11 patients aged between 16 months and 17 years. FMT was most frequently offered to patients with severe, intractable disease. In most cases, faecal microbiota replacement was used alone following the failure of other standard therapies or due to recurrence of the condition after withdrawal of conventional medications.

Outcomes treatment of CDI (C. difficile infection) in adults and children

In agreement with a previous systematic review,[36] we found FMT therapy to be a safe and effective procedure in adult patients with recurrent/refractory CDI, and the success rate of FMT in treating CDI is 90.8% (Table 5). Only one placebo-controlled trial was done for 43 patients with recurrent CDI.[37] In this trial, duodenal infusion of healthy donor faeces cured 15 of 16 recurrent CDI patients (94%) (Cure was defined as an absence of diarrhoea or persistent diarrhoea that could be explained by other causes with three consecutive negative stool tests for C. difficile toxin, within 10 weeks after FMT), whereas vancomycin resolved CDI in only 3 of 13 patients (31%). FMT was shown to be more effective (< 0.001) than the use of vancomycin for the treatment of CDI. At least 644 patients have received FMT for treatment of CDI, including six children.[14, 38-42] It also proved a feasible selection for paediatric patients in this study. Five of the six children experienced a clinical resolution following the FMT, while only one had a clinical relapse.[41]

Table 5. Characteristics and outcomes of FMT in adults and children
IndicationRefractory/relapsing CDI (%)Adults N = 639/833 (76.7%)Children N = 5/11 (45.5%)Total N = 644/844 (76.3%)
(a) Characteristics and outcomes of FMT in adults and children with refractory/relapsing CDI. [11, 13, 14, 24, 26, 30, 31, 37-41, 44, 60, 61, 63, 71, 76-101]
Donor relationshipClose relative or household member (%)382/639 (59.8)5/5 (100)387/644 (60.1)
 Healthy voluntary donors (%)183/639 (28.6)0183/644 (28.4)
 Spouse or partner (%)60/639 (9.4)060/644 (9.3)
 Unknown (%)14/639 (2.2)014/644 (2.2)
Route of faecal instillationColonoscopy (%)320/639 (50.1)2/5 (40)322/644 (50)
 Nasogastrictube, gastroscope or PEG tube (%)161/639 (25.2)3/5 (60)164/644 (25.5)
 Enema or retention enemas (%)52/639 (8.1)052/644 (8.1)
 Nasojejunal or nasoduodenal tube (%)17/639 (2.7)017/644 (2.6)
 Combined two or more of these above (%)72/639 (11.3)072/644 (11.2)
 Unstated (%)17/639 (2.7)017/644 (2.6)
Frequency1–241–2/
OutcomesSuccess rate after FMT (%)580/639 (90.8)4/5 (80)584/644 (90.7)
 No response/recurrence after FMT (%)62/639 (9.7)1/5 (20)63/644 (9.8)
 Adverse eventsReported in four articles[30, 37, 60, 61]Only one study reported some self-limiting symptoms[43]/
IndicationIBD (%)

Adults

N = 108/833 (13.0) (including 94 UC, 14 CD and 3 UC/CD)

Children

N = 3/11 (27.3) (including 3 UC)

Total

N = 111/844 (13.2)

(b) Characteristics and outcomes of FMT in adults and children with IBD. [15, 28, 43-45, 59, 64, 65, 102-107]
Donor relationshipClose relative or household member (%)10/108 (9.3)010/111 (9.0)
 Healthy voluntary donors (%)14/108 (13.0)014/111 (12.6)
 Spouse or partner (%)1/108 (1.0)01/111 (0.9)
 Unknown (%)83/108 (76.9)3/3 (100)86/111 (77.5)
Route of faecal instillationColonoscopy (%)6/108 (5.6)06/111 (5.4)
 Nasogastric tube, gastroscope or PEG tube (%)8/108 (7.4)08/111 (7.2)
 Enema or retention enemas (%)15/108 (13.9)3/3 (100)18/111 (16.2)
 Combined two or more of these above (%)16/108 (14.8)016/111 (14.4)
 Unstated (%)63/108 (58.3)063/111 (56.8)
Frequency1–695/
OutcomesSuccess rate after FMT (%)84/108 (77.8) (1 patients unstated) (including 80 UC, 1 CD and 3 UC/CD)3/3 (100) (including 3 UC)87/111 (78.4%) (including 83 UC, 1 CD and 3 UC/CD)
 No response/recurrence after FMT (%)22/108 (20.4)022/111 (19.8)
 Adverse eventsReported in two articles[64, 65]//
IndicationCDI in IBD (%)

Adults

N = 20/833 (2.4)

Children

N = 1/11 (9.1)

Total

N = 21/844 (2.5)

  1. FMT, faecal microbiota transplantation; CDI, Clostridium difficile infection; IBD, inflammatory bowel disease; AAD, antibiotic-associated diarrhoea; PEG, percutaneous endoscopic gastroscopy.

  2. The success means resolving C. difficile in the patients, as measured by negative stool sample enterotoxin for CDI; achieving clinical remission, disappearance of symptoms or reduction in disease activity index for IBD; resolving C. difficile and improved response to IBD medications for CDI in IBD; and resolution of symptoms for others.

(c) Characteristics and outcomes of FMT in adults and children with CDI in IBD. [30, 42, 62, 108-111]
Donor relationshipClose relative or household member (%)15/20 (75)1/1 (100)16/21 (76.2)
 Healthy voluntary donors (%)1/20 (5)01/21 (4.8)
 Spouse or partner (%)4/20 (20)04/21 (19.0)
Route of faecal instillationColonoscopy (%)18/20 (90)018/21 (85.7)
 Nasogastric tube, gastroscope or PEG tube (%)2/20 (10)1/1 (100)3/21 (14.3)
Frequency1–21/1/
OutcomesSuccess rate after FMT (%)18/20 (90.0)1/1 (100)19/21 (90.5)
 No response/recurrence after FMT (%)2/20 (10)02/21 (9.5)
 Adverse eventsReported in one articles[62]//

Outcomes treatment in IBD in adults and children

We found reports of 111 patients receiving FMT therapy for IBD (Table 5). The overall success rate of IBD (including UC and CD) in adults is 77.8%. FMT may improve UC, with a success rate close to 90%, as measured by disappearance of symptoms or reduction in ulcerative colitis activity index (UCAI). To date, there has been report of three paediatric patients receiving FMT for treatment of UC and C. difficile infections with UC, by nasogastric (NG) tube or enema.[42, 43] All of them demonstrated FMT as a feasible and well-tolerated therapy as it is usually given by duodenal endoscopic lavage during endoscopy or performed as an enema. There were no serious adverse events noted in those children.

FMT has so far proved disappointing in CD. In most patients, FMT therapy was not effective in achieving clinical remission or biologic/endoscopic benefit (Table 5). However, CDI in IBD patients has been improved by FMT therapy. Twenty-one patients have received FMT for treatment of CDI in patients with IBD, including a child.[42] FMT therapy resolved C. difficile in 90.5% patients, as measured by negative stool sample enterotoxin and improved response to IBD medications.

Outcomes treatment of other gastrointestinal diseases in adults and children

One study reported that 30 of 32 patients with AAD (who had received antibiotic drugs before and hospitalised owing to diarrhoea from 3 to 20 stools daily, C. difficile toxin positive or negative) were cured by FMT.[34] FMT also showed effective in treatment of IBS, although the number treated was small (the total number in this systematic review is 6), its successful rate was 100% in the studied.[44, 45] In addition, Borody et al. successfully treated at least 12 patients with chronic constipation in adults and children (Table 4).[44-48]

Outcomes treatment of nongastrointestinal diseases in adults and children

Several reports described patients with autoimmune and neurologic disorders who showed improvement with FMT therapy (Table 4). One report described the co-development of myoclonus dystonia and chronic diarrhoea in a patient who was subsequently treated with FMT therapy, resulting in 90% improvement of her myoclonus dystonia symptoms, allowing her to resume employment and execute fine motor tasks.[49] The onset of autism is often accompanied by gastrointestinal complaints and preceded by antibiotic use so it has been suggested that an altered microbiota might be involved in its pathogenesis.[50] For example, autistic patients were found to have greater numbers and different types of Clostridial species when compared with controls.[51] Improvement in autistic symptoms was reported in two children after FMT and in five children who received daily cultured Bacteroidetes and Clostridia for several weeks.[52] Neurologic improvement has also been reported in one patient with Parkinson's disease who received FMT for chronic constipation.[53] Vrieze et al.[54] reported the first application of FMT therapy in metabolic syndrome. In this study, they found FMT from lean donors significantly increased insulin sensitivity of 18 recipients with metabolic syndrome, which was most likely by increasing concentrations of butyrate-producing intestinal bacteria in the small and large intestine. These findings demonstrated FMT therapy for metabolic diseases warrants further investigation. FMT was also proved effect in patients with chronic fatigue syndrome (CFS). In patients with CFS followed up for up to 20 years, FMT has been reported to be effective with an initial success rate of 42/60 (70%) and sustained response in 35/60 (58%).[55] It is suggested that the pathophysiology of CFS could be in part explained by enteric derived toxin-releasing bacteria capable of producing systemic effects.[56]

Outcomes FMT therapy for diseases in children

We compared the characteristics with FMT therapy for diseases in adults and children (Table 5). Since the number of reports of FMT in children, we here review them all. The first case report of FMT in children was published in 2001 by Borody et al.[47] In the three enrolled chronic constipation patients, two were aged <18 with more than 3 years history. Transplantation of healthy donor stool was repeated daily for at least 5 days. After the first infusion, each patient defecated within hours, and long-term restoration of normal bowel function was observed. The remaining five reports of FMT therapy in children were treated for recurrent CDI and published between 2007 and 2013 in Australia and US. Russell et al.[14] reported a 2-year-old girl with culture-confirmed relapsing CDI despite intensive therapy with traditional and experimental agents including probiotics. Within 36 h of FMT administration, the symptoms resolved and did not recur during 6 months of monitoring. The patient's stool tested negative for C. difficile toxin after the procedure. Garg et al.[39] and Kahn et al.[40] also reported successful case reports of children infected with recurrent CDI.

Kunde et al.[43] have initiated the first phase I clinical trial for use of FMT in children with UC. They enrolled eight patients and completed data are available on three subjects who have completed the study. Two of three subjects achieved clinical response by reduction in paediatric ulcerative colitis activity index (PUCAI) of 15 points or more. One of them had complete resolution of disease activity. Singh et al[42] reported a 6-year-old female was with unremitting UC and C. difficile infection who was successfully treated by FMT therapy. Rubin et al.[41] report a case series that included two paediatric patients (aged 6 and 8). One experienced a clinical resolution following the FMT, while the other had a clinical relapse. No adverse events were noted. These cases demonstrated the therapeutic potential of FMT as a practical and effective therapy treatment in paediatric patients.

Donor selection and screening

Screening of faecal donors before transplantation may be just as necessary as screening organ donors to ensure that unwanted traits are not transferred with the transplant.[57] The source of donor faeces has generally been divided across four groups: (i) close relative or household member of the patient (54.5% and 54.5% in adults and children respectively), (ii) Unrelated (defined as healthy voluntary donors who were not related to the recipients) (25.3% and 18.2% in adults and children respectively), (iii) individuals who had intimate physical contact with the patients (spouse or partner) (7.9% in adults) or (iv) Unknown – the relation of the donor to the recipient was not specified in the study (12.2% and 27.3% in adults and children respectively) (Tables 1-5). The microbial composition and functional activities of the faecal enemas prepared from different individuals cannot readily be standardised. This problem could be overcome by the collection of multiple donations, which could be stored frozen, from a small number of thoroughly screened donors. The feasibility of this kind of procedure is supported by a recent paper by Hamilton et al.[29, 30]

Potential donors are excluded if they have known HIV infection, hepatitis B virus infection, or hepatitis C virus infection or known exposure to these viruses within the previous year. For the same reason, the donation criteria exclude people who participate in high-risk sexual behaviours or use illicit drugs, anyone who has had a tattoo or body piercing within the previous 6 months or has recently been incarcerated, and individuals who have travelled to areas of the world where endemic diarrhoea is prevalent. Potential donors should also be excluded if they have IBD, IBS, chronic constipation or chronic diarrhoea, or a history of gastrointestinal malignancy or known gastrointestinal polyposis. Potential donors are also excluded if they have received antibiotics in the preceding 3 months or are currently receiving major immunosuppressive medications or systemic anti-neoplastic agents. It is also prudent to exclude individuals with metabolic syndrome, systemic autoimmunity, atopic diseases or chronic pain syndrome.[58]

Route of administration

FMT therapy has been successfully performed via colonoscopy, nasogastric tube, enema and even two or more methods in the same recipient (Tables 1-5). The majority of adult patients received FMT therapy by colonoscopy (42.0%); followed by nasogastric tube, gastroscope or percutaneous endoscopic gastroscopy (PEG) tube (22.7%); enema or retention enemas (12.4%); combined two or more of these above(11.8%); administration methods unstated (8.6%) or nasojejunal (2.5%). However, the paediatric patients were treated mainly by enema (45.5%), by nasogastric tube or PEG tube (36.4%) or colonoscopy (18.2%). Normal saline was used to prepare most FMT suspensions, followed by water and milk. 83.3% (667/801) patients received only one therapy (number of therapies not reported in 43 patients), but if it did not work, another one or more transplantation could be administered up to as many as 69.[59] There are insufficient published data to establish a relationship between the frequency of FMT and its success rate.

In several reports, it is stated that faeces are processed and infused as quickly as possible (mostly less than 6 h before the transplantation.) following production by the donor. Almost all faecal preparations are processed in a normal aerobic environment. To overcome the barriers such as aesthetic concerns and costs of donor screening, Hamilton et al. standardised frozen preparation for transplantation of faecal microbiota and found that standardised frozen donor faecal bacterial preparations are also effective in treating recurrent CDI.[30] High-throughput 16S rRNA gene sequencing demonstrated that frozen faecal microbiota can be used to effectively treat recurrent CDI resulting in restoration of the structure of gut microbiota and clearing of Clostridum difficile.[29]

Effects on stool composition and immunological effects of FMT

FMT therapy had a dramatic impact on the composition of the patient's gut microbiota. A study showed a durable beneficial change in the patients' (Five IBS, Four constipation and one Crohn's disease) bacterial populations of the colon to represent those of the healthy donor's microbiota.[45] Similar results were found when treating recurrent CDI by Khoruts A. et al.[60], who also found the change in bacterial composition was accompanied by resolution of the patient's symptoms. The Nood et al.[37] comparative study also described a deep microbiota analysis. They showed not only low diversity in the CDI patients but also rapid and long-term maintenance of bacterial diversity after FMT, with an increase in Bacteroidetes species and clostridium clusters IV and XIVa and a decrease in Proteobacteria species. This suggests that the donor's bacteria quickly occupied their requisite niches resulting in restoration of both the structure and the function of the microbial communities present. Deep sequencing exploring the human microbial diversity in patients with CDI after FMT showed that the clinical cure of CDI is associated with an increase in diversity and richness.[25] A positive clinical response of FMT therapy for UC was observed in one (of five patients treated) whose microbiota had been effectively augmented by FMT. This augmentation was marked by successive colonisation of donor-derived phylotypes including the anti-inflammatory and/or short-chain fatty acids-producing Faecalibacterium prausnitzii, Rosebura faecis and Bacteroides ovatus. Disease severity (as measured by Mayo Score) was associated with an overrepresentation of Enterobacteriaceae and an underrepresentation of Lachnospiraceae.[27]

Adverse events

Adverse events were uncommon. Possible adverse events were reported in seven articles. Borody et al.[61] reported a transient adverse events included sore throat, flatulence, rectal discomfort, nausea, abdominal cramping, bloating, headache and abdominal pain when using FMT therapy to treat C. difficile syndromes. After treatment of FMT for recurrent CDI, a patient developed constipation in the initial months after the procedure.[60] Hamilton, M. J. et al.[30] reported approximately a third noted some irregularity of bowel movements and excessive flatulence, but resolved in clinic follow-up. One UC patient with inactive disease subsequently developed diarrhoea, abdominal cramping and active disease requiring medication after FMT therapy.[62] Aas[63] reported one CDI patient undergoing peritoneal dialysis for end-stage renal disease developed peritonitis after 2 days post-FMT and died shortly thereafter. Other adverse events included a temporary increase in CRP, flatulence, vomiting, common cold, itchiness, erythema, paraesthesia on the hip, collapse and blisters on the tongue;[28] transient fever developed in four out of nine patients;[64, 65] and mild diarrhoea and abdominal cramping on the infusion day.[37] Some studies report patient deaths due to the underlying disease, where the patient has not responded to the faecal transplantation.

Limitations of the data

Reporting on the efficacy of FMT is limited by the absence of well-designed randomised controlled trials, with evidence to date coming largely from small case series and case reports. Only one open-label, randomised, controlled trial has been reported to date[37], in patients with recurrent CDI. Twenty five of 67 articles were available in abstract form only. There were often absent or incomplete data for the FMT procedure, including preparation and administration of the faecal infusion and patient preparation. Most patients also received additional treatment or preparation with another procedure (e.g. vancomycin or bowel lavage) before FMT was performed, making it difficult to estimate the effect of FMT alone. Outcome variables such as symptom resolution were not well defined, making comparison between studies difficult. A final limitation is the heterogeneity of the populations treated. This analysis included patients from eight countries treated over a period of 55 years. The variability in outcomes across procedures may be partly explained by differences in the underlying populations studied and by the small number of patients in some categories.

Discussion

The potential to manipulate the microbiome, by application of FMT therapy, is now being extensively investigated, but most of the reports to date, often positive, have been uncontrolled case reports and small case series. A high percent of gastroenterologists[66] and patients[67, 68] indicated that they are willing to consider FMT as treatment and are eager for it to become available. Although there have been some concerns regarding safety and acceptability associated with FMT,[69] this approach has already been used for treating microbiome diseases such as Clostridium difficile associated pathologies, IBD and IBS.[11, 16] The most impressive demonstration of efficacy comes from the first randomised controlled trial of FMT in CDI reported by Nood et al.[37] where the efficacy of FMT is compared with conventional pharmacological therapy (standard vancomycin regimen).

Use of FMT has not been confined to gastrointestinal disease, the single blinded study of Vrie et al.[54] with an autologous transplantation for metabolic syndrome opens the door for FMT in treatment of nongastrointestinal disease. There are preliminary reports on the use of FMT therapy in a wide range of disorders including Parkinson's disease,[53] fibromyalgia, chronic fatigue syndrome,[56] multiple sclerosis,[48] myoclonus dystonia,[49] obesity,[70] insulin resistance and the metabolic syndrome,[54] and childhood regressive autism.[51]

We need to recognise FMT as a form of organ transplantation. In many aspects, FMT is simpler to perform than other organ transplants, without the need for immunological matching of donor and recipient or the need for immuno suppression after the procedure.[16] Most faecal donors have been healthy family members or spouses/significant partners who have common genetic and/or environmental factors. It can be performed by various routes including nasogastric (NG) tube, nasojejunal tube, upper tract endoscopy (EGD), colonoscopy or by retention enema, depending on which is deemed safest for the individual patient. As this procedure is increasingly used, several groups have proposed both guidelines for selection and testing of stool donors and protocols for performance of the procedure.[71, 72]

FMT therapy is a ‘low tech’ procedure which is easy to perform. The only significant reported adverse events following FMT therapy to date have been a solitary case of IBS and suspected peritonitis following the transplant. Mortality was reported for each study, and none was attributed to the FMT procedure. Death occurred generally in the critically ill and elderly patients with severe comorbidities. While further clinical studies and long-term follow-up of patients are required, FMT appears to be a viable, safe and inexpensive option for cases of recurrent and refractory disease.

Clinical improvement can be noticed within a few days following donor faeces infusion. C difficile carriage rates average 37% for infants 0–1 month of age and 30% between 1 and 6 months of age. In the 11 children patients in this review, the success rate of FMT therapy was 90.9%, which is higher than that in adults although the numbers are too small for statistical comparison. We believe though that this systematic review could provide support for the further use of FMT therapy in children, which might also help avoid the adverse effects of other drugs.

It is now well-appreciated that intestinal microbiota constitute a microbial organ that is integral to overall host physiology, including pivotal roles in metabolism and immune system function.[2] FMT aims to restore the intestinal microbiota of an individual and reestablishment of normal bowel function. Only patients with the most recalcitrant cases are likely to undergo FMT, usually out of desperation after multiple treatment approaches have failed.[72] Systematic review reveals that the reported efficacy of FMT in treating recurrent CDI is greater than 90%.[73] But it is aesthetically unappealing, it is logistically challenging (in terms of harvesting and processing suitable donor material) and there is a lack of efficacy data from randomised, controlled trials. It is not surprising perhaps that it has been very little used till recently.[74]

In this study, we describe the current knowledge of FMT both in adults and in children. Data suggest that FMT therapy may be a highly effective and safe therapy for CDI when standard treatments have failed both in adults and in children. Further support for the efficacy of this procedure is provided by its use in the treatment of other disorders. FMT therapy is also a more readily accessible and less costly procedure than some standard or other alternative therapies. The application of faecal for such transplants is the first step in a long journey. The next step in this journey has already been taken with the development of a standardised filtered, frozen preparation of stool for FMT and use of a ‘universal donor’.[30, 75] The first randomised controlled trial (RCT) testing the efficacy of FMT in the treatment of CDI in adults gave us an approving consequence. More well-designed RCT studies are necessary to confirm the efficacy of this therapy and define best practices for its use, including standards for pre-FMT treatment.

Conclusions

Although further controlled evidence is needed, FMT therapy appears to be successful in both adults and children with gastrointestinal diseases such as CDI and possibly IBD. With the development of convenient equipment for faecal preparation and oral administrable capsules filled with faecal suspension or freeze-dried preparations, more patients may benefit from a standardised FMT therapy with less aesthetic concerns, greater convenience and perhaps higher efficacy.

Authorship

Guarantor of the article: Dr Kaichun Wu.

Author contributions: Kaichun Wu performed the systematic review and authored the work. Sumei Sha, Jie Liang and Min Chen assisted with data interpretation, wrote the manuscript and approved the final version. Bin Xu, Cong Liang and Ni Wei performed the research, collected and analysed the data. All authors approved the final version of the manuscript.

Acknowledgement

Declaration of personal interests: None.

Declaration of funding interests: This study was supported by grants from the Ministry of Health National Commonweal Research Project (no. 200802156), National Healthcare Research Project (no. 201002020), National Scientific Support Project (nos. 2012BAI06B03 and BSW11J013), National Natural Science Foundation of China (Nos. 81322037, 81170360 and 81370504) and National Excellent Doctoral Dissertation of PR China (No. 201182). The funding bodies had no role in the study design, collection, analysis or interpretation of the data.

Ancillary