Article Text

The use of faecal microbiota transplant as treatment for recurrent or refractory Clostridioides difficile infection and other potential indications: second edition of joint British Society of Gastroenterology (BSG) and Healthcare Infection Society (HIS) guidelines
Free
  1. Benjamin H Mullish1,2,
  2. Blair Merrick3,
  3. Mohammed Nabil Quraishi4,5,6,
  4. Aggie Bak7,
  5. Christopher A Green8,9,
  6. David J Moore10,
  7. Robert J Porter11,
  8. Ngozi T Elumogo12,13,
  9. Jonathan P Segal14,15,
  10. Naveen Sharma4,5,6,
  11. Belinda Marsh16,
  12. Graziella Kontkowski16,17,
  13. Susan E Manzoor5,
  14. Ailsa L Hart1,18,
  15. Christopher Settle19,
  16. Josbert J Keller20,21,
  17. Peter Hawkey5,22,
  18. Tariq H Iqbal4,5,6,
  19. Simon D Goldenberg3,
  20. Horace R T Williams1,2
  1. 1Division of Digestive Diseases, Department of Metabolism, Digestion and Reproduction, Faculty of Medicine, Imperial College London, London, UK
  2. 2Departments of Gastroenterology and Hepatology, St Mary's Hospital, Imperial College Healthcare NHS Trust, London, UK
  3. 3Centre for Clinical Infection and Diagnostics Research, Guy’s and St Thomas’ NHS Foundation Trust, King’s College London, London, UK
  4. 4Department of Gastroenterology, University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK
  5. 5Microbiome Treatment Centre, University of Birmingham, Edgbaston, UK
  6. 6Institute of Cancer and Genomic Sciences, University of Birmingham, London, UK
  7. 7Healthcare Infection Society, London, UK
  8. 8Department of Infectious Diseases & Tropical Medicine, University Hospitals NHS Foundation Trust, Birmingham Heartlands Hospital, Birmingham, UK
  9. 9School of Chemical Engineering, University of Birmingham, Birmingham, UK
  10. 10Institute of Applied Health Research, College of Medical and Dental Sciences, University of Birmingham, Birmingham, UK
  11. 11Department of Microbiology, Royal Devon and Exeter Hospitals, Barrack Road, UK
  12. 12Quadram Institute Bioscience, Norwich Research Park, Norwich, UK
  13. 13Norfolk and Norwich University Hospital, Norwich, UK
  14. 14Department of Gastroenterology, Royal Melbourne Hospital, Melbourne, Victoria, Australia
  15. 15Department of Medicine, University of Melbourne, Melbourne, Victoria, Australia
  16. 16Lay representative for FMT Working Party, Healthcare Infection Society, London, UK
  17. 17C.diff support, London, UK
  18. 18Department of Gastroenterology and Inflammatory Bowel Disease Unit, St Mark's Hospital and Academic Institute, Middlesex, UK
  19. 19South Tyneside and Sunderland NHS Foundation Trust, South Shields, UK
  20. 20Department of Gastroenterology, Haaglanden Medisch Centrum, The Hague, The Netherlands
  21. 21Department of Gastroenterology and Hepatology, Leiden University Medical Center, Leiden, The Netherlands
  22. 22Public Health Laboratory, Faculty of Medicine, University of Birmingham, Birmingham, UK
  1. Correspondence to Dr Horace R T Williams, Division of Digestive Diseases, Department of Metabolism, Digestion and Reproduction, Imperial College London, London, UK; h.williams{at}imperial.ac.uk; Dr Simon D Goldenberg, Centre for Clinical Infection and Diagnostics Research, Guy’s and St Thomas’ NHS Foundation Trust, King’s College London, London W2 1NY, UK; Simon.Goldenberg{at}gstt.nhs.uk

Abstract

The first British Society of Gastroenterology (BSG) and Healthcare Infection Society (HIS)-endorsed faecal microbiota transplant (FMT) guidelines were published in 2018. Over the past 5 years, there has been considerable growth in the evidence base (including publication of outcomes from large national FMT registries), necessitating an updated critical review of the literature and a second edition of the BSG/HIS FMT guidelines. These have been produced in accordance with National Institute for Health and Care Excellence-accredited methodology, thus have particular relevance for UK-based clinicians, but are intended to be of pertinence internationally. This second edition of the guidelines have been divided into recommendations, good practice points and recommendations against certain practices. With respect to FMT for Clostridioides difficile infection (CDI), key focus areas centred around timing of administration, increasing clinical experience of encapsulated FMT preparations and optimising donor screening. The latter topic is of particular relevance given the COVID-19 pandemic, and cases of patient morbidity and mortality resulting from FMT-related pathogen transmission. The guidelines also considered emergent literature on the use of FMT in non-CDI settings (including both gastrointestinal and non-gastrointestinal indications), reviewing relevant randomised controlled trials. Recommendations are provided regarding special areas (including compassionate FMT use), and considerations regarding the evolving landscape of FMT and microbiome therapeutics.

  • BACTERIAL INFECTION
  • COLONIC BACTERIA
  • COLONIC MICROFLORA
  • DIARRHOEAL DISEASE
  • ENTERIC BACTERIAL MICROFLORA

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Executive summary of recommendations

Patient summary

Faecal microbiota transplant (FMT), sometimes also known as stool or poo transplantation, can be an effective treatment for patients with C. difficile (commonly known as C. diff) infection (CDI). It is usually given when the infection comes back after antibiotic treatment (relapse), or occasionally if antibiotics do not work (refractory). It is not fully understood how FMT helps patients with CDI, but it is thought it is partly to do with restoring beneficial gut microorganisms (eg, bacteria) and the chemicals (eg, metabolites) they produce.

The first British Society of Gastroenterology (BSG)/Healthcare Infection Society (HIS) guidelines on the use of FMT for C. diff were published in 2018, and since this time, new evidence has become available. This has prompted this second edition of the guidelines. Key recommendations focus on which patients should be offered FMT, when it should be offered and the best ways to administer it. The guidelines also describe important considerations for screening of stool donors to ensure the safety and success of FMT. Two further topics are focused on in this second edition. One is the evidence for the use of FMT for conditions other than CDI, including irritable bowel syndrome (IBS), ulcerative colitis and Crohn’s disease, as well as conditions outside of the gut, such as obesity and metabolic syndrome. The second topic considers patients with conditions in which there are no other treatment options available to them, and if they can be offered FMT: this is called compassionate use.

Introduction

FMT (sometimes referred to by other names, including ‘intestinal microbiota transplant/transfer1) describes the transfer of minimally manipulated faeces from a healthy screened donor to a patient for the treatment of disease. FMT is now entering its second decade of use in modern mainstream medicine, with the first randomised trial reporting its utility following antibiotic treatment in recurrent CDI (rCDI) in 2013.2 The first BSG/HIS-endorsed FMT guidelines were published in 2018,2 and the interest continues to grow in the use of FMT, both for CDI and for its potential in the management of non-CDI conditions.3

Since the first BSG/HIS FMT guidelines in 2018, there has been publication of European and North American CDI-related guidelines4 that have also addressed FMT, consensus reports relating to aspects of FMT service design and delivery,5 and other BSG guidelines that have made consideration of a role for FMT in a non-CDI setting, for example, for IBD.6 More recently, the National Institute for Health and Care Excellence (NICE) medical technologies guidance summarised the clinical and cost effectiveness of FMT, from a UK National Health Service (NHS) perspective.7 Despite these publications, the BSG and HIS advocated for a second edition of the UK FMT guidelines (with the focused version presented here and full version available in online supplemental file A) for a number of reasons. Firstly, the high levels of clinical interest within this field mean that this has been a fast-moving area with a rapidly growing literature base. Particular areas of evolution since the last guideline iteration have included randomised trials in both CDI and non-CDI settings, the reporting of data from regional and national FMT registries (with longer periods of follow-up and larger numbers of patients than were previously described), and concerns related to donor screening (relating both to the COVID-19 pandemic and high profile reports of FMT-related pathogen transmission with adverse patient outcomes). Secondly, while the NICE medical technologies guidance presented a general evaluation of the clinical use of FMT, its remit did not include guidance as to many of the more specific areas related to FMT provision and administration that are of greatest relevance to practising clinicians in this field (including donor selection and screening and material preparation), or consideration of non-CDI indications. As such, there was a compelling case to apply NICE-accredited methodology to the current evidence base and provide clinicians with the highest-quality recommendations and guidance on which to base their practice of FMT use in adults.

Supplemental material

The focus of these guidelines was on the use of ‘conventional’ FMT, to inform use in healthcare settings (primarily the NHS) and in academia. As such, as per the prior guidelines, studies were considered only if they explored the administration of whole stool, and not modified products, such as cultured microorganisms (or their proteins, metabolites or other components) or microbiota suspensions. The guideline development team (referred to as Working Party) are aware of developments in the USA in this space, particularly the recent Food and Drug Administration (FDA) approval of ‘next-generation’ FMT products, including RBX2660/Rebyota (Ferring; a rectally administered FMT-type product8) and SER-109/Vowst (Seres/Nestle; a purified spore-based product9) for preventing CDI relapses. Clinical trials that contributed to the licensing of these products investigated the performance of these agents compared with standard-of-care anti-CDI antibiotics. None explored efficacy compared with ‘conventional’ FMT. At the time of writing, no such products were licensed for use within the UK or European Union, and none have been licensed in any region as part of management of a non-CDI indication.

Glossary of terms used is provided in online supplemental file B.

Aims and scope

The main purpose of this second edition of the guidelines was to set recommendations and best practice for the optimal provision of effective and safe FMT for recurrent or refractory CDI (defined in box 1) in adult (≥18 years) patients. The secondary purpose was to provide guidance for using FMT in conditions other than CDI in the adult population. These recommendations focused on the provision of FMT in the UK, although many aspects are also relevant internationally. The focus was on ‘minimally manipulated’ stool, and not the ‘next-generation’ FMT products (ie, defined microbial communities as ‘microbiome therapeutics’). The diagnosis and management of CDI in general were considered outside the scope of these guidelines.

Box 1

Commonly accepted CDI definitions*

  • Recurrent CDI: infection symptoms resolved after treatment but recurred within 8 weeks. It is currently difficult to establish a difference between a relapse of the disease or the occurrence of a new infection.

  • Refractory CDI: CDI which is not responding to antibiotic treatment. This type of CDI may or may not be considered fulminant CDI.

  • Severe CDI: when fever, leucocytosis and rise in serum creatinine are present, which may also be supported by further diagnostic abnormalities, for example, distension of the large intestine seen at imaging.

  • Fulminant CDI: also known as severe complicated, occurs when one of the following CDI-related factors are present: hypotension, septic shock, elevated serum lactate, ileus, toxic megacolon, bowel perforation or a fulminant course of disease.

Please note that clinically, many of these definitions overlap and it is not always possible to clearly group patients into these categories. Additionally, over the disease course, this may change, for example, refractory CDI may become fulminant.

*Taken from ESCMID guidelines (https://doi.org/10.1016/j.cmi.2021.09.038).

CDI, Clostridioides difficile infection; ESCMID, European Society of Clinical Microbiology and Infectious Diseases.

Box 2

Recommended Donor History Questionnaire

Positive response to any of these questions may exclude further consideration regarding donation at that time; it may be appropriate to rescreen and consider for donation at a later time point based on the particular scenario.

  • Receipt of antibiotics and/or other medications potentially associated with gut microbiome perturbation, to include (but not limited to) proton pump inhibitor, statin, immunosuppression, and/or chemotherapy, within the past 3 months.

  • Known prior exposure to HIV and/or viral hepatitis, within the past 3 months.

  • Known previous or latent tuberculosis.

  • Use of illicit drugs, any tattoo, body piercing, needlestick injury, blood transfusion, acupuncture (outside of licensed or approved UK facilities), all within the previous 4 months.

  • New or multiple (more than one) sexual partners within the past 3 months.

  • Sex with somebody diagnosed with HTLV-1 and HTLV-2*.

  • Previously living in areas with high prevalence of HTLV-1 and HTLV-2*.

  • Receipt of a live-attenuated vaccine within the past 6 months.

  • Cold sores, anal ulcers, anal sores, pruritus ani within the past 3 months.

  • Underlying gastrointestinal conditions/symptoms (eg, history of IBD, IBS, chronic diarrhoea, chronic constipation, coeliac disease, bowel resection or bariatric surgery).

  • Acute diarrhoea/gastrointestinal symptoms within the past 2 weeks.

  • Family history of any significant gastrointestinal conditions (eg, family history of IBD or colorectal cancer).

  • History of atopy (eg, asthma, eosinophilic disorders).

  • Any systemic autoimmune conditions.

  • Any metabolic conditions, including diabetes and obesity.

  • Any neurological or psychiatric conditions.

  • History of chronic pain syndromes, including chronic fatigue syndrome and fibromyalgia.

  • History of any malignancy.

  • History of receiving growth hormone, insulin from cows or clotting factor concentrates, or known risk of prion disease.

  • History of receiving an experimental medicine (including vaccines) within the past 6 months.

  • History of travel to tropical countries within the past 6 months.

  • *This question to be asked in centres where laboratory screening for HTLV-1 and HTLV-2 may be difficult; areas to focus on, but not limited to: Japan, the Caribbean, and South America.

  • HTLV, human T-cell lymphotropic virus; IBD, inflammatory bowel disease; IBS, irritable bowel syndrome.

Box 3

Recommended blood screening for donors

Pathogen screening:

  • Hepatitis A IgM

  • Hepatitis B (HBsAg and HBcAb)

  • Hepatitis C antibody

  • Hepatitis E IgM

  • HIV-1 and HIV-2 antibodies

  • HTLV-1 and HTLV-2 antibodies

  • Treponema pallidum antibodies (TPHA, VDRL)

  • Epstein-Barr virus (EBV) IgM and IgG*

  • Cytomegalovirus (CMV) IgM and IgG*

  • Strongyloides stercoralis IgG

  • Entamoeba histolytica serology

  • Cysticercal serology

General/metabolic screening:

  • Full blood count with differential

  • Creatinine and electrolytes

  • Liver enzymes and liver function tests

  • C reactive protein

  • *EBV and CMV testing is recommended where there is the potential that the FMT prepared from that donor will be administered to immunosuppressed patients at risk of severe infection if exposed to CMV and EBV.

  • FMT, faecal microbiota transplant; HBcAb, hepatitis B core antibody; HBsAg, hepatitis B surface antigen; HTLV, human T-cell lymphotropic virus; TPHA, T. pallidum haemagglutination assay; VDRL, Venereal Disease Research Laboratory.

Box 4

Recommended stool screening for donors

  • Clostridioides difficile tcdB (toxin B) by PCR*

  • Campylobacter, Salmonella and Shigella, preferably by PCR

  • Shiga toxin-producing Escherichia coli by PCR

  • Other enteropathogenic E. coli, including, but not limited to, enteropathogenic E. coli by PCR

  • Multidrug-resistant bacteria, including, but not limited to, carbapenemase-producing Enterobacterales, extended-spectrum beta-lactamases and vancomycin-resistant enterococci**

  • Stool ova, cysts and parasite analysis, including:

    • Cryptosporidium and Giardia antigen or PCR

    • Acid fast staining for Cyclospora, Isospora and Microsporidia

  • Norovirus and rotavirus PCR

  • SARS-CoV-2***

  • Helicobacter pylori stool antigen****

  • * Glutamate dehydrogenase (GDH) screening for possible C. difficile is not required or recommended; where performed, a positive GDH would not be sufficient to exclude a donor on the grounds of ‘positive C. difficile status’.

  • **Methicillin-resistant Staphylococcus aureus is primarily recognised as a skin rather than a gastrointestinal organism; therefore, screening is not universally recommended.

  • ***Based on current prevalence and laboratory expertise, a broader viral screen may be appropriate, ideally via multiplex panel, which may include, for example, sapovirus and poliovirus.

  • ****Consider testing but not necessarily to exclude as a donor if positive; may potentially wish to consider informing any recipients of H. pylori stool antigen-positive material, especially if recipients do not have a background of/are not currently H. pylori stool antigen positive.

Box 5

Post-baseline bookend screening stool microbiology

  • Clostridioides difficile tcdB (toxin B)

  • Campylobacter, Salmonella and Shigella

  • Shiga toxin-producing Escherichia coli

  • Other enteropathogenic E. coli, including, but not limited, to enteropathogenic E. coli

  • Microsporidia

  • Norovirus and rotavirus PCR

  • Cryptosporidium

  • SARS-CoV-2

  • Cyclospora

Methodology

Topics for these guidelines were derived from the initial discussions of the Working Party during the stakeholder meeting. The included questions (online supplemental appendix 1) were adapted from those in the previous version of the guidelines published in 2018.1 Methods were followed in accordance with the NICE manual for conducting evidence syntheses (online supplemental file C).

Supplemental material

Data sources and search strategy

Three electronic databases (MEDLINE, Embase, Cochrane Central Register of Controlled Trials) were searched with the last search date in July 2023. Search terms were constructed using relevant index and free-text terms (online supplemental appendix 1). Reference lists of identified relevant articles were scanned for additional studies and forward reference searching (identifying articles which cite relevant articles) was performed. The searches were restricted to primary articles published in the English language.

Study eligibility and selection criteria

Search results were downloaded to Covidence software and screened for relevance. Two reviewers discussed their disagreements first, and the third reviewer was available to arbitrate but was not needed. The results of study selection and the list of excluded studies for all questions are available in online supplemental appendix 2.

Data extraction and quality assessment

Included epidemiological studies were appraised for quality using checklists (links available in online supplemental appendix 3A). The results of quality appraisal are available in online supplemental appendix 3B.

Data were extracted by one reviewer and checked by other reviewers. For each question, the data from the included studies were extracted to create the tables of study description and summary of findings tables (online supplemental appendix 4).

Rating of evidence and recommendations

The strength of the evidence was defined by GRADE (Grading of Recommendations Assessment, Development and Evaluation) tables (online supplemental appendix 5) and using the ratings ‘high’, ‘moderate’, ‘low’ and ‘very low’ to construct the evidence statements, which reflected the Working Party’s confidence in the evidence. The strength of recommendation was adopted from GRADE and reflects the strength of each evidence statement.

Consultation process

Feedback on draft guidelines was received from the participating organisations and through consultation with relevant stakeholders. The Working Party reviewed stakeholder comments and collectively agreed revisions (online supplemental file D).

Guideline development team and conflicts of interest

Members of the Working Party represent professional societies, that is, BSG and HIS, as well as clinical microbiologists, gastroenterologists, infection prevention and control doctors, clinical and academic researchers, FMT production manager, methodologists and two lay members. Individual members were mostly UK based, but some international experts were also chosen to ensure that the guidelines are also relevant to an international audience. BSG and HIS commissioned the authors to undertake this Working Party report. The authors received no specific funding for this work. Financial support for the time required to obtain the evidence and write the manuscript was provided by the authors’ respective employing institutions. BHM was the recipient of a National Institute for Health and Care Research (NIHR) Academic Clinical Lectureship (CL-2019-21-002). The Division of Digestive Diseases at Imperial College London receives financial and infrastructure support from the NIHR Imperial Biomedical Research Centre based at Imperial College Healthcare NHS Trust and Imperial College London. The authors would like to thank Dr Rohma Ghani for her assistance on the topic of donor screening, Dr Bin Gao for reviewing the studies related to FMT given to patients with functional constipation, Dr Andrew Flatt for advice on donor screening, Professor Mark Gilchrist for advice on medical product regulation, and Professor Jessica Allegretti, Professor Christian Lodberg Hvas and Dr Simon Baunwall for providing additional data from the included studies. The views expressed in this publication are those of the authors and have been endorsed by BSG and HIS and approved following a consultation with external stakeholders. Authors declared no substantial conflicts of interest which would prevent them from being the members of the guidelines panel. All conflicts of interest are disclosed in online supplemental file C.

Scheduled review

The guidelines will be reviewed at least every 4 years and updated if change(s) are necessary or if evidence emerges that requires a change in practice.

Implementation

The Working Party agreed that there is no anticipated additional cost associated with implementation of these guidelines unless existing practice falls well below currently accepted standards. Assessing the cost-effectiveness of different treatments is not within the scope of this guidance. The practices recommended by these guidelines are currently used in most centres offering FMT in the UK. There is a potential cost-saving and other benefits (eg, reducing the carbon footprint) when certain recommendations are followed (eg, donor screening or using aerobic processes for FMT preparation). Lay materials and continuing professional development questions are available in the online supplemental files E and F.

Rationale for recommendations

Effectiveness and safety of FMT in treating CDI

There is clear evidence of the growing use of FMT globally. With the availability of randomised trial outcome data, FMT has become an accepted treatment for recurrent and refractory CDI. A recent pan-European survey suggested a disparity in access to FMT between countries (or even between regions within countries), suggesting ongoing significant underutilisation in patients who may stand to benefit from FMT.10 Previous BSG/HIS guidelines3 recommended that FMT should be offered to patients with refractory CDI, or those with risk factors for recurrence, but not as first-line treatment. At the time of their publication, there were fewer randomised trials and comparison treatment was limited to vancomycin. Due to a small number of studies conducted before the first edition of the guidelines was published, meta-analyses were not possible and the evidence for effectiveness was not well established. Additionally, effectiveness and, more importantly, safety of FMT for some patient populations - including those who were immunocompromised or immunosuppressed, frail and older patients, and patients with certain comorbidities - was unknown.

Of note, FMT use in the context of CDI is predominantly described as being administered after a course of anti-CDI antibiotics. Depending on the study reviewed, FMT may be either viewed as a direct part of the treatment of an episode of CDI (ie, consolidation of therapy after anti-CDI antibiotics), or that the anti-CDI antibiotics are the central therapy and that the role of FMT is primarily prevention of further recurrence. Growing understanding about mechanisms of efficacy of FMT in CDI—including FMT’s roles in both direct inhibition of the growth of C. difficile, as well as prevention of spore germination,11 mean that both interpretations merit consideration. Reflecting this view, FMT in CDI will interchangeably be referred to as a modality of treatment and intervention to prevention of recurrence within this guideline, with the assumption that FMT has been administered only after a preceding course of anti-CDI antibiotics unless otherwise stated.

General population with CDI

Effectiveness of FMT versus standard care or placebo: there was strong evidence which suggested that FMT is more effective than standard care or placebo for preventing CDI recurrence in the general population.2 12–16

Adverse events following FMT versus standard care or placebo: there was strong evidence which suggested no negative effect of FMT.2 12–16

Patients with severe, complicated or fulminant CDI

Effectiveness of FMT in patients with severe CDI: there was weak evidence which suggested that FMT is beneficial in this patient group.17

Effectiveness of FMT in patients with severe CDI compared with patients with mild/moderate CDI: there was moderate evidence which suggested there was no difference between these two patient groups.18–24

Effectiveness of FMT in patients with refractory or fulminant CDI versus recurrent CDI: there was inconsistent evidence which suggested no difference in effect for these patient groups.25–29

Effectiveness of FMT in patients with pseudomembranous colitis compared with other patients: there was weak evidence, and it is not clear whether in these patients FMT may be less successful.19 22

Adverse events in patients with severe, refractory or fulminant CDI: there was weak evidence which suggested there was no increased risk associated with FMT for these types of patients.17 18 25

Adverse events in patients with pseudomembranous colitis: there were no studies.

First episode of CDI

Effectiveness of FMT: there was moderate evidence which suggested that FMT is effective in these patients.13 30

Adverse events: there was moderate evidence which suggested no negative effect.13

Patients with coexisting IBD and CDI

Effectiveness of FMT: there was weak evidence that suggested FMT was effective in treating CDI in patients with IBD.31–35

Effectiveness of FMT in patients with IBD with CDI compared with patients without IBD: there was moderate evidence which suggested that FMT for CDI is equally successful in patients who have IBD and those who do not.18 22 23 25 27 36–41

Effect on adverse events: there was weak evidence, but it suggested that FMT is safe in patients with IBD treated for CDI.28 31 33 34 36 However, two studies also highlighted that some patients with IBD may experience a flare following FMT.31 36

Immunocompromised or immunosuppressed patients with CDI

Effectiveness of FMT: there was weak evidence which suggested that FMT is effective in treating CDI in patients who are immunocompromised or immunosuppressed.42 43

Effectiveness in immunocompromised/immunosuppressed patients compared with immunocompetent patients: there was moderate evidence which suggested that there was no difference in effectiveness between these two patient groups.19 21–23 26 28 37–41 44

Adverse events: there was weak evidence which suggested that FMT is safe in this patient group.42 43

Patients with cancer with CDI

Effectiveness of FMT: there was weak evidence which suggested that FMT is effective in this patient group.45 46

Effectiveness in patients with cancer compared with patients with no cancer: there was weak evidence, but it suggested that there was no difference in the effectiveness between these two patient groups.19 21 40

Adverse events: there was weak evidence which suggested that FMT was safe in this patient group.45 46

Post-solid organ transplant patients with CDI

Effectiveness of FMT: there was weak evidence which suggested that FMT is effective in this patient group.47

Effectiveness in solid organ transplant patients compared with patients with no solid organ transplant: there were no studies.

Adverse events: there was weak evidence which suggested that FMT is safe in this patient group.47

Patients with liver disease and CDI

Effectiveness of FMT: there was weak evidence which suggested FMT is effective in this patient group.48

Effectiveness in patients with liver disease compared with patients without liver disease: there was weak evidence which suggested no difference in the effectiveness of FMT between these two groups of patients.38 40 49

Adverse events: there was weak evidence which suggested that FMT was safe in this patient group.48

Patients with kidney disease and CDI

Effectiveness of FMT: there were no studies.

Effectiveness in patients with kidney disease compared with patients without kidney disease: there was weak evidence which suggested that there is no difference in the effectiveness of FMT between these patient groups.19 23 38 40

Adverse events: there were no studies.

Patients with diabetes mellitus and CDI

Effectiveness of FMT: there were no studies.

Effectiveness in patients with diabetes mellitus compared with patients without diabetes mellitus: there was weak evidence which suggested that there is no difference in the effectiveness of FMT between these patient groups.19 39 40

Adverse events: there were no studies.

Patients with cardiovascular disease and CDI

Effectiveness of FMT: there were no studies.

Effectiveness in patients with cardiovascular disease (CVD) compared with patients without CVD: there was weak evidence, which suggested that there is no difference in the effectiveness of FMT between these patient groups.39

Adverse events: there were no studies.

Patients with recurrent urinary tract infections and CDI

Effectiveness of FMT: there were no studies.

Effectiveness in patients with urinary tract infection (UTI) compared with patients without UTI: there was weak evidence, which suggested that there is no difference in the effectiveness of FMT between these patient groups.23

Adverse events: there were no studies.

Patients with COVID-19 infection and CDI

Effectiveness of FMT: there was weak evidence which suggested that FMT is effective in this patient group.50

Effectiveness in patients with COVID-19 compared with patients without COVID-19: there were no studies.

Adverse events: there was weak evidence which suggested FMT is safe in this patient group.50

Patients with CDI and other conditions

Effectiveness of FMT: there were no studies.

Effectiveness in patients with other conditions compared with patients without these conditions: there was weak evidence, which suggested that there is no difference in the effectiveness of FMT between these patient groups.19 22 38 39

Adverse events: there were no studies.

Patients with CDI and multiple comorbidities

Effectiveness of FMT: there were no studies.

Effectiveness in patients with multiple comorbidities compared with patients without comorbidities: there was weak evidence which suggested that FMT may be less successful in patients with multiple comorbidities.20 27 37 44 51 52

Adverse events: there were no studies.

Additional data from excluded studies

Quality of life

One study53 reported improved quality of life after the patients underwent FMT for CDI.

Mortality

Two studies54 55 reported no difference in mortality rates, one56 reported that the incidence of CDI-related mortality decreased when an FMT programme was introduced, one23 reported that early FMT reduced mortality in severe cases, and one study57 reported that patients who received FMT had a 77% decrease in odds of mortality.

Long-term effectiveness

Six studies23 58–62 reported that at long-term follow-up (up to 1 year), FMT was still effective.

Asymptomatic carriage after FMT

One study63 reported that asymptomatic carriage of C. difficile after FMT is rare.

New or worsening symptoms following FMT

One study23 reported that 1 year after follow-up, nausea was present in 18% of the patients, abdominal pain in 21% and diarrhoea in 33%, but that no serious events related to FMT occurred. One study59 reported that within a year after FMT, the prevalence of constipation increased, but that most of the cases did not need treatment. Other symptoms included urgency, cramping and an increased incidence of IBS. Two years after FMT, new conditions included weight gain, diabetes mellitus, dyslipidaemia, thyroid problems, gastrointestinal problems and serious infections. These conditions were not considered directly linked to FMT. Other studies reported the onset of the following new issues,36 54 60 62 but none of these conditions were assessed for causality. One study reported worsening pre-existing chronic IBD and rheumatoid arthritis.60 One study64 reported that there was a slightly higher incidence of myocardial infarction in FMT group compared with non-FMT at 1 year follow-up, but that the incidence of other conditions was similar. At 10-year follow-up, one study65 reported that there were no new diagnoses of autoimmune diseases, gastrointestinal disorders or malignancies and that there were no deaths which were attributed to FMT.

Resolution or improvement of conditions following FMT

Three studies reported resolution or improvement of existing conditions following FMT,54 60 62 including eradication of multidrug-resistant microorganisms,54 improvement of undifferentiated colitis, Crohn’s disease, ulcerative colitis, diabetes mellitus and Parkinson’s disease,62 and improvement of IBS, IBD and alopecia areata.60 None of these studies investigated whether these improvements were directly associated with FMT.

The Working Party discussed the above evidence and concluded that FMT administered after CDI treatment with appropriate antibiotics appears to be more effective than placebo, or additional doses of vancomycin or fidaxomicin, in prevention of CDI recurrence. However, the sensitivity analyses performed due to high heterogeneity suggest that its effectiveness depends on many factors, including the route of FMT administration, the number of FMTs given, the type of patient, and the length of follow-up. It is also important to highlight that the high heterogeneity was also a result of different types of comparisons, which are typically used in clinical practice and constitute standard care; for example, in some studies, participants were given initial antibiotics to treat CDI and received placebo as a part of standard care while in other studies, participants received the initial antibiotics for treatment as well as additional doses of vancomycin or fidaxomicin as a comparison with FMT. In either case, FMT was more effective than any of these standard regimens. The results of one randomised controlled trial (RCT)5 support previous observational reports that retention enema is not an efficient route of administration.

Additionally, FMT seems to be beneficial for patients with different types of comorbidity regardless of the severity or phenotype of CDI and the number of CDI episodes preceding FMT. The Working Party acknowledged that some types of comorbidities and multiple comorbidities may make the FMT less effective, and that for these patients, more than one FMT may be required. Clinically, this would be similar for all patients because subsequent FMT, preferably from a different donor, should be offered if the first FMT fails. One dose of FMT may be less effective in patients with severe or pseudomembranous colitis and to achieve a desired effect, these patients could benefit from additional doses. However, clinically, this issue may not be relevant because in practice, patients with CDI are not routinely assessed for the presence of pseudomembranous colitis. Therefore, the clinical pathway for these patients would remain similar to patients with other CDI types. Nevertheless, FMT in these patients still appears to be better than placebo or antibiotics alone. Thus, FMT should be given for different types of patients, regardless of their comorbidities or the type of CDI. As per the previous iteration of the guidelines, the Working Party discussed that the only absolute contraindication for FMT is the presence of anaphylactic food allergy.

In previous guidelines, there was a concern that FMT may cause harm in some types of patients, including those who are immunocompromised or immunosuppressed, those with liver or kidney disease or those with IBD. However, the evidence now suggests that the incidence of adverse events, regardless of their severity, appears to be similar in different types of patients. Thus, the Working Party agreed that FMT should still be considered as a treatment option for patients with comorbidities based on its safety. Moreover, in the general population, the incidence of adverse events in patients who receive FMT does not appear to be different when compared with patients who receive placebo or anti-CDI antibiotics. The Working Party would also like to stress that, due to the similar incidence of occurrence in different treatment groups, gastrointestinal events such as diarrhoea, nausea or bloating are probably more likely to be associated with CDI itself and possibly some co-interventions (eg, bowel preparation) rather than with FMT treatment. Based on clinical experience of the Working Party members, adverse events, none of which were captured by the included studies, may occasionally occur, but their incidence is very rare. A recent systematic review,66 which investigated the occurrence of adverse events after FMT, reported that the overall rate of severe adverse events was 0.65% (95% CI 0.45% to 0.89%). The population in this study included patients with IBD (4.8%) as well as immunosuppressed/immunocompromised patients (8%). For specific adverse events, the incidence was 0.19% (95% CI 0.09% to 0.31%) for sepsis or sepsis-like conditions, 0.27% (95% CI 0.15% to 0.43%) for aspiration pneumonia and 0.20% (95% CI 0.09% to 0.34%) for bowel perforation. Mild adverse events were also relatively rare, with constipation reported in 1.03% (95% CI 0.77% to 1.33%) of the patients, abdominal pain in 1.66% (95% CI 1.33% to 2.03%), nausea in 0.92% (95% CI 0.67% to 1.20%), vomiting in 0.34% (95% CI 0.20% to 0.52%), flatulence in 0.70% (95% CI 0.49% to 0.94%) and febrile episodes in 0.33% (95% CI 0.19% to 0.50%) of patients following FMT. In general, the majority of adverse events seem to occur either due to unsafe FMT products or unsafe practice of administration, both of which are avoidable when careful donor screening is in place and appropriate care is given to FMT recipients. Other events may be unpreventable, for example, diarrhoea due to glycerol being used as cryoprotectant, but these are relatively minor and self-limiting.

The data from the excluded studies point out that the desired effects of FMT are generally long-lasting with many patients experiencing no recurrence of CDI and no evidence of adverse events occurring months to years after FMT. There are some patients who experience recurrence or relapse and the Working Party discussed how these patients should be managed. It was concluded that current evidence23 and clinical practice support the treatment of these patients with either further FMT or anti-CDI antibiotic therapy.

The Working Party discussed whether, due to an apparent benefit, FMT should be offered as a treatment for patients with the first episode of CDI. The effectiveness for patients experiencing the first or second CDI has recently been established in one RCT.13 However, due to the fact that FMT may be more invasive and expensive compared to antibiotics, that a relatively high success rate may be achieved with anti-CDI antibiotics alone, together with the challenges in donor recruitment and adequate FMT provision, then FMT is not currently recommended for a primary CDI episode. Instead, this issue can be investigated in future studies.

Recipient factors influencing the outcome of FMT for patients with CDI

The evidence above demonstrates that FMT is generally effective in the majority of individuals regardless of their health status. Despite this, there are still patients in whom FMT fails. Risk factors for CDI recurrence after FMT are poorly understood, but certain patient characteristics such as advanced age, female sex and some medications have been proposed as potential predictors for failure.67 There may also be some additional modifiable factors which could be optimised before FMT is given and these have not yet been explored. Despite some studies reporting some patient characteristics as risk factors, the results have been mostly inconsistent. Additionally, there remain concerns about the safety of FMT for some patients. Underlying vulnerabilities such as older age and the effect of some medications could potentially increase an individual’s risk of severe adverse events associated with FMT. Previous BSG/HIS guidelines3 did not identify any risk factors for CDI recurrence other than post-FMT antibiotics. The guidelines also found very little evidence that would demonstrate the safety of FMT in more vulnerable populations. As a result, the guidelines recommended caution when administering FMT to people with certain conditions such as immunosuppression or liver disease and suggested that antibiotic therapy should be avoided or delayed when possible.

Demographic factors

Age

Effect on success rates: there was moderate evidence which suggested that this does not influence the effectiveness of FMT.19–23 26–28 37–40 44 68 69

Effect on adverse events: there was weak evidence which suggested that adverse events are similar across all age groups.68

Sex

Effect on success rates: there was moderate evidence which suggested that this does not influence the effectiveness of FMT.19–21 23 26–28 37–40 44

Effect on adverse events: there were no studies.

Body mass index

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.19 39

Effect on adverse events: there were no studies.

Factors associated with CDI

Number of CDI episodes before FMT

Effect on success rates: there was moderate evidence which suggested that this does not influence the effectiveness of FMT.19–21 23 28 38 44 69

Effect on adverse events: there were no studies.

Hospitalisation due to CDI

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.19 38

Effect on adverse events: there were no studies.

Antibiotics used for treatment of CDI before FMT

Effect on success rates: there was weak evidence which suggested that these do not influence the effectiveness of FMT.19 22 39 40 69

Effect on adverse events: there were no studies.

C. difficile strain

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.21 23 41

Effect on adverse events: there were no studies.

Healthcare-acquired CDI

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.20

Effect on adverse events: there were no studies.

Other risk factors

Use of PPIs and other antisecretory medications

Effect on success rates: there was moderate evidence which suggested that these do not influence the effectiveness of FMT.19 20 22 23 26 28 37 38 40 41

Effect on adverse events: there were no studies.

Use of corticosteroids preceding the administration of FMT

Effect on success rates: there was weak evidence which suggested that these do not influence the effectiveness of FMT.40

Effect on adverse events: there were no studies.

Use of lactulose preceding the administration of FMT

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.40

Effect on adverse events: there were no studies.

Probiotic use preceding the administration of FMT

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.19 22

Effect on adverse events: there were no studies.

Non-CDI antibiotic use preceding the administration of FMT

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.23 26 40

Effect on adverse events: there were no studies.

Use of narcotics preceding the administration of FMT

Effect on success rates: there was weak evidence which suggested that these do not influence the effectiveness of FMT.39

Effect on adverse events: there were no studies.

Hospitalised at or before FMT

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.22 26 28 39

Effect on adverse events: there were no studies.

Blood biomarkers

Effect on success rates: there was weak evidence which suggested that these do not influence the effectiveness of FMT.20 28 51 However, one study51 reported a higher risk of recurrence of CDI in patients with zinc deficiency as well as a beneficial effect for zinc-deficient patients who were given zinc supplements.

Effect on adverse events: there were no studies.

Other risk factors

Effect on success rates: there was weak evidence which suggested that these do not influence the effectiveness of FMT.27 38 41 69

Effect on adverse events: there were no studies.

Upon reviewing the above evidence, the Working Party agreed that there are currently no identified factors which affect the effectiveness of FMT. There may be some characteristics of CDI infection that may result in FMT being less effective; however, as was highlighted in a previous section, FMT is still more effective than standard antibiotics and placebo. Adverse events were assessed only for patients’ age and the evidence suggested that age had no effect. The Working Party agreed that the paucity of studies reporting adverse events for patients with different characteristics likely represents the lack of effect of these characteristics on the incidence and severity of adverse events. Based on these conclusions, the Working Party agreed that FMT should not be declined or delayed based on any patient-related or CDI-related characteristic.

Additionally, the Working Party agreed that further studies investigating the effect of simple non-modifiable risk factors (eg, age, sex, etc) are not necessary because the existing studies suggest that these factors are not likely to influence the effectiveness or adverse events of FMT to the point where antibiotics and/or other therapies should be considered as an alternative. As such, future studies should focus on investigating modifiable risk factors which can be corrected before FMT is given so that its outcomes are optimised. A recent review70 identified possible recipient factors which facilitated donor microbiota engraftment, including genetics, inflammation status and environmental factors (eg, diet). Further studies are needed to identify if these factors can influence clinical outcomes of FMT.

Donor factors influencing the outcome of FMT for patients with CDI

A robust donor screening programme is an essential part of FMT services to ensure safety for FMT recipients. Donor recruitment is challenging; using standard criteria applied in many FMT services to ensure safety and efficacy, one recent study reported that only 1.7% of prospective candidates qualified as suitable donors.71 Moreover, the study reported that due to a lengthy screening process, as many as 39% of the candidates were lost to follow-up even before their suitability was established. The reluctance of the public to donate their stool is also well documented and seems to stem from the social perception of stool, the lack of awareness of the importance of donation, and the logistic difficulties in collection and transport of the stool.72 Evidently, there is a need for a pragmatic approach for the recruitment and screening of potential donors.

The primary aim of donor screening is mitigating risk of pathogen transmission via FMT. A secondary aim of donor screening is to exclude potential donors who may have an ‘aberrant/adverse’ gut microbiome. While the complexity and relative novelty of exploration of the gut microbiome mean that there is no clear agreed definition of what a ‘healthy’ or ‘unhealthy’ gut microbiome is,73 either compositionally or functionally, there is the theoretical potential for transmission of gut microbiome traits (and therefore potential for transmission of risk of diseases with a link to the gut microbiome) via FMT. There are also some studies that include microbiome sequencing and other approaches to try and find which bacteria transplanted from donor to recipient are associated with success.74 75 So far, it has been difficult to define a core set of bacteria or functions underlying a good donor or successful FMT. At the moment, there is little evidence which allows FMT services to define a healthy microbiome which is most optimal for donation. Previous BSG/HIS guidelines3 acknowledged that research into donor factors is lacking. Therefore, the guidelines recommended a general approach that all healthy adults under 60 years of age with body mass index (BMI) under 30 kg/m2 could be potential candidates for donor screening. The recommendations then focused on an initial screening using a health and travel questionnaire, followed up by a battery of laboratory testing of blood and stools to further ensure the safety of FMT material. The guidelines also recommended regular reassessment of donors to ensure continuing safety. Since the guidelines were published, more evidence has become available, especially around the experience of donor screening and the retention of possible donors. The emergence of the COVID-19 pandemic also raised questions whether prospective donors should be tested for other, non-gastrointestinal pathogens, to ensure the safety of recipients.

Related versus not related donor

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.22 24 52

Effect on adverse events: there were no studies.

Age of the donor

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.23 27

Effect on adverse events: there were no studies.

Sex of the donor

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.23

Effect on adverse events: there were no studies.

Amount of stool produced

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.27

Effect on adverse events: there were no studies.

Microbiome composition of the donor

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.27

Effect on adverse events: there were no studies.

The Working Party reviewed the above evidence and concluded that it is likely that routinely measured donor factors do not influence the effectiveness of FMT for treatment of CDI. The Working Party agreed that the use of universal donors is the most practical and cost-effective way to obtain donor stools. The previous practice of using related donors, which in early days before stool banks existed were the most reliable source of donor stools, is now outdated and should be avoided. There is no established evidence that stools from a related donor influence the effectiveness of the FMT, but there may be logistical difficulties and potentially additional costs related to donor screening. There is also a concern that stool microbiota may be less diverse in these donors. As a related donor may cohabit with a recipient, the overlap of environmental factors with the patient (eg, diet) may affect their gut microbiome and the success of FMT.

There were no studies which investigated whether the donor factors affected the incidence or severity of adverse events, but the members agreed that, apart from the composition of the microbiota, they are not likely to influence the effectiveness of FMT. As mentioned above, some studies demonstrate that the composition of microbiota of the donor stool may predict the success or failure of FMT,74 75 but none of these studies met the inclusion criteria for these guidelines. The Working Party stressed that wherever donor factors have been investigated, this was done in situations in which all donors were screened for possible transmissible diseases and where safety of FMT material was established. Therefore, they stated that screening of all donors must remain in place to ensure the safety of FMT recipients. All donors should also be rescreened regularly to ensure ongoing safety.

Rationale for recommendations on overall approach to donor screening

The Working Party agreed a robust donor screening procedure remains mandatory. As per the original version of these guidelines, the screening should continue to comprise a questionnaire, to identify risk factors for an aberrant microbiome and pathogen carriage, and laboratory-based testing for pathogen detection. This should be an ongoing process that is repeated at appropriate intervals.

The Working Party discussed the reported FMT complications since the last guidelines which might influence updates in the recommended donor screening protocols. From one perspective, there have been a number of reported cases of infection post-FMT apparently related to pathogen transmission which may have been mitigated by additional donor screening processes, including C. perfringens,76 atypical enteropathogenic Escherichia coli77 and Shiga toxin-producing E. coli.78 It is also important to highlight the well-publicised case of FMT-related infection transmission in two immunosuppressed patients who developed bloodstream infection after transmission of E. coli carrying an extended-spectrum beta-lactamase via FMT, leading to one death.79 80 There had been considerable concern since the emergence of SARS-CoV-2 regarding its potential for transmission via FMT (particularly related to its potential route of entry via the luminal tract and well-described gastrointestinal symptoms related to infection), and rapid consensus updates to donor screening were introduced to mitigate risk.81 However, despite this theoretical risk, there are no reported cases of FMT-related SARS-CoV-2 transmission described, to the knowledge of the Working Party. Since the last guideline, there has been an increased period of time for reporting of registry data and of prospective case series. Overall, FMT for rCDI appears safe with several years of follow-up post-treatment; there have been very few cases of infection potentially attributable to FMT and very low rates of new diseases which might feasibly be attributable to FMT.23 36 54 58–62 64–66 There is a need to strike an appropriate balance between screening practices that are robust enough to mitigate the potential risks of providing FMT while allowing sufficient pragmatism. Overly stringent screening focused on theoretical risk of every possible pathogen risks making the process impossible to comply with.

Regarding the recommended Donor History Questionnaire, the Working Party provided some updates to this compared with the original version of this guideline (box 2). For instance, the assessment for risk factors for bloodborne viruses has been updated to be consistent with those from UK Blood and Transplant. The Working Party noted that FMT services in certain settings aimed to recruit donors from within blood donation services, given the degree of overlap in assessment between blood and stool donation, although no such approach was currently being undertaken within the UK. Additional assessments have now been recommended, for example, enquiring about recent cold sores, anal ulcers and/or persistent pruritus ani, to screen for organisms that colonise the oral, rectal or perineal mucosa, including herpes simplex virus, pinworm and mpox (previously monkeypox) virus. Of note, the Working Party discussed that while a health questionnaire assessment is mandatory, it is beyond the scope of the committee to mandate specific content or specific exclusion criteria, and box 2 represents recommendations based on suggested best practice rather than compulsory questions. Questionnaire content and clinical interpretation of responses should be discussed and agreed at a local level following a robust risk assessment.

Laboratory-based blood screening of potential donors remains mandatory (box 3). The Working Party discussed that while a number of the pathogens listed in box 2 are not recognised to transmit via the faecal-oral route (being predominantly bloodborne pathogens), and the theoretical risk of them being transmitted via FMT being therefore low, there was still justification to screen for them out of a principle of caution. The Working Party again discussed and upheld their recommendation regarding Epstein-Barr virus and cytomegalovirus (CMV) testing being only recommended where there is the potential that the FMT prepared from that donor will be administered to immunosuppressed patients at risk of severe infection. Of interest, recent evidence suggests that only a very small proportion (approximately 1%) of CMV IgG or IgM-positive donors have detectable stool CMV DNA on PCR, and no CMV IgM-positive donors or those with stool CMV DNA have infectious virus on cell culture.82 Nevertheless, this recommendation has also been upheld on the principle of an abundance of caution. While the Working Party recommended consideration of a set of general/metabolic blood tests for donors, they did not set specific limits/thresholds for values. The examples were discussed of a donor with, for instance, incidental marked anaemia or raised C reactive protein as being at high risk of having significant undiagnosed disease which may impact the gut microbiome, and therefore being unsuitable for material donation.

The Working Party discussed the need to update stool pathogen screening compared with the last version of the guideline (box 4). In one respect, they acknowledged the need to recommend additional screening, with faecal SARS-CoV-2 being of relevance given its potential for faecal-oral transmission, as discussed above. The Working Party recognised that a global consensus document designed for European practice developed at the height of the COVID-19 pandemic had recommended SARS-CoV-2 screening of each donated stool sample.81 The Working Party concluded that while an argument could be made for continuing with this approach based on risk assessment at present, the currently evolving risk landscape related to SARS-CoV-2 (related to a number of factors, including national COVID-19 vaccination roll-out) may mean that a modified protocol for SARS-CoV-2 screening may become appropriate over the lifetime of this guideline. Similarly, the Working Party noted a report of atypical enteropathogenic E. coli transmission related to FMT, and as such felt that more considered screening for this in donors was justified.77 The Working Party also discussed that new evidence had emerged since the last version of the guidelines that suggested against certain gastrointestinal pathobionts being transmitted via FMT. In particular, a Danish FMT service recently described 13 out of 40 donors as being Helicobacter pylori stool antigen positive, but that 26 FMTs administered from five positive donors had not resulted in any recipients becoming H. pylori stool antigen positive at a median of 59 days.83 While these data do not support the need for H. pylori stool antigen being part of screening, the Working Party also discussed the different risk burden that theoretical H. pylori transmission might have in the UK versus in the Far East, given its association with gastric cancer. It was noted that there are recent data demonstrating transmission of Blastocystis via FMT, but that this did not influence success of FMT as treatment for rCDI, and it was not associated with any gastrointestinal symptomatology over months of follow-up, suggesting no need to intensify donor screening for this organism.84

The Working Party noted recent literature exploring the impact of FMT upon the gut microbiota dynamics of potentially procarcinogenic bacteria. This topic first came to light from a study of 11 paediatric patients with rCDI (of whom 6 had underlying IBD), in whom 4 patients were found to have sustained acquisition of procarcinogenic bacteria post-FMT, after transmission from colonised donors. It was also noted that two patients experienced clearance of such bacteria after FMT from a negative donor.85 Using full genome sequencing, one of these patients acquiring procarcinogenic bacteria was shown to have durable donor-to-recipient transmission of E. coli with the colibactin gene (clbB), which has been associated with colonic tumours.86 A further retrospective study87 analysed stool metagenomes of matched pre-FMT versus post-FMT samples from 49 patients with rCDI, together with their matched donors. This showed higher prevalence and abundance of potentially procarcinogenic polyketide synthase-positive (pks+) E. coli in the gut microbiome of patients with rCDI compared with their healthy donors, and that the pks status of the post-FMT gut microbiome related to the pks status of the donor being used (with pks being negative in five out of eight of their donors at all time points sampled and detected in overall low levels otherwise). More specifically, persistence (8 out of 9 patients) or clearance (13 out of 18 patients) of pks+ E. coli in pks+ patients correlated with pks in the donor (p=0.004). While these data are of interest, the Working Party concluded that the small number of publications on this topic, unclear understanding of the true potential causative procarcinogenic nature of the bacteria being studied, and overall reassuring safety profile of FMT meant that there was no current clinical indication for routine metagenome screening for such bacteria or their genes as part of donor screening. Additionally, since the durability of engraftment of donor strains after a single FMT is variable but may be only several months in the case of a reasonable proportion of taxa,73 the real procarcinogenic risk could be even lower than previously suggested, should bacteria with these gene cassettes be those with limited colonisation duration. Further studies within this field should be undertaken and results monitored. The Working Party noted that FMT for rCDI is often being used in an older and frail population for whom the risk-to-benefit ratio of FMT is being considered over a fairly short period, that is, patients with limited alternate therapeutic options, with the aim of minimising further hospital admissions. This ratio would be different in the context of younger patients, particularly where FMT was used on a more exploratory basis, and this may influence the importance of considering the potential future role for screening for such bacteria.

The Working Party also noted that a number of studies had proposed using stool metagenomics as a tool to assess stool donors, and proposed a variety of ecological or taxonomy-based metrics to select out and stratify potentially ‘ideal’ donors.88 Discussions within the Working Party concluded that while this was of research interest, there was no justification for use of any assessment of this nature as part of the donor screening/selection process at present. It was also observed that a small number of studies had suggested a potential role for additional modalities of laboratory assessment as part of donor screening; for instance, one study observed a trend towards increased gastrointestinal symptoms post-FMT for rCDI after receipt of FMT from a donor with positive small intestinal bacterial overgrowth, as assessed by positive lactulose breath test.89 Again, the Working Party felt that while this was of interest and supported future research, there was no current justification for this to be incorporated into the donor screening process.

As per their discussions regarding the health questionnaire, the Working Party felt that it was beyond the scope to mandate or exclude specific laboratory tests. Thus, the lists given in boxes 3 and 4 reflect suggested best practice but not compulsory testing. Laboratory-based testing and clinical interpretation of results should be performed and agreed at a local level following a robust risk assessment. Consistent with this, the Working Party noted the differences in laboratory donor screening approaches that are reported in different regions globally. These are consistent with the different prevalence and risk profile of different pathogens within each region.90 As highlighted by the case of COVID-19, the list of pathogens for which testing is undertaken needs to be constantly reviewed, revised and updated, based on local epidemiology and the latest evidence base. One area that may require particular focus in this regard is the potential for emergence of new viral pathogens or rise in population prevalence of known viral pathogens with established faecal-oral transmission, for example, poliovirus; the pertinence of this is highlighted by its detection within sewage water in London in 2022.91 92

The Working Party no longer supports the use of fresh FMT, because this approach does not allow for direct testing of the donor stool used to manufacture FMT prior to administration and does not allow for a period of quarantine in the case where additional donor testing may be required. Stool may be processed into FMT immediately from donors who have passed baseline screening, but the Working Party agreed that it should initially be quarantined. The Working Party also agreed that post-baseline screening is required prior to release of FMT from quarantine to further mitigate the risk of pathogen transmission. This post-baseline donor screening needs to take a safe but pragmatic approach, and should cover two aspects:

  • Bookend testing (box 5) on donated stool to pick up acquisition of asymptomatic, transmissible enteric pathogens during the donation period. Again, the exact framework should be defined by local policies and donation schedules, following a robust risk assessment. However, the Working Party recognised that there is a need to define the longest period the donor can donate without testing to ensure that safety of the recipient is not compromised. The Working Party agreed that this period should be no longer than 4 months. Bookend testing could include testing of pooled aliquots of donor stool used for manufacturing FMT. FMT could only be considered for release from quarantine once results have been demonstrated to be clear.

  • Bookend assessment and/or testing of donor to identify risk factors for pathogen acquisition since baseline screening. The exact framework should be defined by local policies and donation schedules, following a robust risk assessment. This could involve a donor questionnaire at each donation. FMT could only be considered for release from quarantine if no specific risks were identified. FMT manufactured from donors identified as having acquired risk factors during the donation period (such as unprotected sex with a new partner) would need to undergo continued quarantine, and only be considered from release once the appropriate repeat blood testing had been performed and results were demonstrated to be clear, ensuring that there had been a sufficient time period to allow for seroconversion.

Preparation-related factors influencing the outcome of FMT for patients with CDI

The effectiveness of FMT is presumed to depend on transferred commensal microbiota being able to engraft and proliferate in the recipient’s colon. Thus, preservation of viability of relevant bacteria during processing and storage is considered an important factor for FMT effectiveness. At the moment, there is no standard approach to how donated stools are processed and stored, although it has been suggested that variations in processing seem to have little influence on FMT effectiveness for rCDI.93 Due to the difficulties with donor recruitment, as well as an additional benefit of quarantine of the donor stools, the desire is to keep FMT product for as long as possible. Longer storage is also helpful if an interruption of donor supply or manufacturing process occurs, an example of which was observed during the recent pandemic. There is a need for studies to determine the time thresholds and optimal conditions in which FMT products need to be processed and used. The determination of appropriate storage temperatures is also important for cost-effectiveness and environmental considerations. Previous BSG/HIS guidelines3 found mostly low-quality evidence in relation to stool processing and storage. Based on standard practice, they recommended that stools should be processed within 6 hours of defecation, stored at −80°C and used within 6 months of processing.

Fresh versus frozen stool

Effect on success rates: there was moderate evidence which suggested that fresh and frozen stools are equally effective.19 21 27 28 69

Effect on adverse events: there was weak evidence which suggested that this does not influence the effectiveness of FMT.29

Stool frozen at −20°C vs −80°C

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.94

Effect on adverse events: there were no studies.

Lyophilised stool

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.95–97

Effect on adverse events: there was weak evidence which suggested FMT from lyophilised stools is safe.96

Type of capsule

Effect on success rates: there was weak evidence which suggested that this does not influence the effectiveness of FMT.98

Effect on adverse events: there were no studies.

Processing time

Effect on success rates: there was weak evidence which suggested that processing time for 150 min or longer does not influence the effectiveness of FMT.23 99

Effect on adverse events: there were no studies.

Storage time

Effect on success rates: there was weak evidence which suggested that storing frozen products for more than a year may not influence the effectiveness of FMT.23 94 99

Effect on adverse events: there were no studies.

Additional data from excluded studies

Anaerobic versus aerobic processing

Two studies93 100 reported that processing the stool samples under anaerobic conditions helps to preserve microbial diversity93 and viability.100 On the other hand, one study101 reported that oxygen-free atmosphere was not necessary as long as the air above collected samples was removed.

Effect of freezing

Two studies93 102 reported that freezing resulted in the loss of microbial diversity of the processed stool samples. One study102 reported that preparation in maltodextrin-trehalose solutions, storage at −80°C standard freezer and rapid thawing at 37°C provided the best results for the samples to retain their revivification potential. The same solution was also reported to be effective in preserving lyophilised samples.101

Emulsion process

One study103 showed that magnet plate emulsion and Seward Stomacher Emulsion were similar in terms of maintaining microbial load.

The Working Party concluded that there is currently no evidence to suggest that any preparation factors in particular have an effect on the effectiveness or the incidence and severity of adverse events of FMT for CDI. The literature from the excluded studies suggests that anaerobic process and freezing the products have an effect on the viability of the microbiota, but there still seems to be an adequate clinical effect regardless of these findings. In terms of efficacy, it is currently not known how long fresh stools can be kept before they are processed and how long the FMT products can be stored frozen. However, the literature suggests that up to 180 min before processing starts and up to 12 months of storage time are acceptable. Due to a relatively low impact on effectiveness, the Working Party suggested that other factors such as overall safety, cost-effectiveness, convenience and environmental concerns should be considered when preparing and storing FMT products. It is preferred that the products are stored frozen because this provides convenience and additional safety, as the delay in administration allows more time to withdraw faeces if a donor becomes ill or tests positive for a transmissible pathogen. Current practice in much of the UK is to start the processing of the stools as soon as possible and no longer than within 150 min from the time of defecation to freezing. The Working Party stated that there is no reason to challenge this practice. A threshold of 150 min is used within a number of studies, reflecting a balance between enough time to pragmatically transfer stool from production by donor to an FMT laboratory, and yet a short enough time to mitigate alterations to microbiome composition and functionality.104 Either aerobic or anaerobic process is acceptable, and in line with standard practice, cryoprotectant needs to be added. Additionally, the Working Party reported that many centres in the UK and in mainland Europe have successfully used older products and they concluded that the storage time of the frozen FMT products can be extended from 6 to 12 months and that the temperature of the freezer can be increased to −70°C to minimise the environmental impact. It is currently not known whether the products could be stored at −20°C for up to 12 months. The Working Party expressed concerns that storage at this temperature could result in the loss of bacterial counts, and therefore recommended that this practice should be avoided until there is more evidence to support it. Since the FMT needs time to be thawed, and there is a concern that the microbial cultures will start to deteriorate, the Working Party recommended that this is done in an ambient temperature and used within 6 hours of thawing. The Working Party also agreed that thawing in water baths should be avoided because this may lead to contamination of FMT with waterborne pathogens. The decision whether and how stools should be encapsulated or lyophilised can be left to individual laboratories and will depend on the availability of the equipment.

The Working Party agreed to provide the advice in line of the recommendations from the previous edition of the guidelines,3 which suggested, based on data from two systematic reviews, that 50 g of stool should be used for FMT. Previous edition of the guidelines also recommended that stools should be mixed with 1:5 proportion to a diluent. However, the Working Party also agreed that these should be considered as arbitrary figures, not currently supported by the evidence. Thus, FMT processing facilities may choose to adjust this volume and proportion depending on a clinical need and the availability of the donor stools. While the bottom limit for the volume of the stool to be used has not yet been established, it has been acknowledged that some FMT centres use 30 g of stools diluted to 1:6 ratio, and this is still clinically effective.

Route of delivery and other administration factors influencing the outcome of FMT for patients with CDI

FMT can be delivered via the upper and lower gastrointestinal tract, allowing it to reach different parts of the digestive tract. Different delivery routes may have different rates of success but are also associated with different risks and adverse events and may therefore not be suitable for all patients. There are also other factors to consider during FMT administration. It is still not clear whether taking certain medications or undergoing bowel preparation shortly before FMT could influence its outcome. Previous BSG/HIS guidelines3 acknowledged that lower and upper gastrointestinal tract administration have similar success rates and adverse events and that both could be used if clinically appropriate. However, due to the evidence suggesting lower efficacy associated with enema administration, this route of delivery was only recommended when neither upper gastrointestinal routes, nor colonoscopy, would be considered appropriate. Additionally, at the time of publication, there was a paucity of evidence regarding encapsulated FMT; thus, no recommendations were made regarding its use. Regarding other factors, the evidence was low, but the guidelines suggested the use of bowel lavage and a single dose of antimotility agent if FMT was to be delivered via lower gastrointestinal route, and the use of PPIs and prokinetics when FMT was administered via the upper gastrointestinal tract.

Route of delivery

Colonoscopy versus other methods

Effect on success rates: there was moderate evidence which suggested that a colonoscopic route may be slightly more effective when compared with other administration routes combined.19 21 25 26 38 39 95 105 106

Effect on adverse events: there was weak evidence which suggested colonoscopic delivery has no effect on adverse events.25 38 106

Enema versus other methods

Effect on success rates: there was inconsistent evidence but it suggested that enema may be less effective than other methods.26 107 108

Effect on adverse events: there was very weak evidence which suggested that delivery via enema had no effect on adverse events when compared with other administration routes.42 108

Lower gastrointestinal (unspecified) versus other methods

Effect on success rates: there was very weak evidence which suggested no difference in effect when comparing lower gastrointestinal tract administration with other methods when combined.23 27 109

Effect on adverse events: there was very weak evidence which suggested that delivery via lower gastrointestinal route had no effect on adverse events when compared with other administration routes.109

Upper gastrointestinal versus other methods

Effect on success rates: there was weak evidence which suggested no difference in effect when comparing upper gastrointestinal tract administration with other methods when combined.19 21 23 25–27 105 106 109 110

Effect on adverse events: there was weak evidence which suggested that upper gastrointestinal tract administration had no effect on adverse events when compared with other administration routes.25 105 106 109

Oral capsules versus other methods

Effect on success rates: there was weak evidence which suggested no difference in effect when comparing oral capsules with other delivery methods when combined.21 26 38 95 105–109

Effect on adverse events: there was weak evidence which suggested that oral capsules had no effect on adverse events when compared with other administration routes.38 43 44 105 106 109

Bidirectional (upper and lower gastrointestinal simultaneously) versus other methods

Effect on success rates: there was very weak evidence which suggested a potential benefit when comparing bidirectional method of FMT administration with other routes when combined.105

Effect on adverse events: there was very weak evidence which suggested that bidirectional method had no effect on adverse events when compared with other administration routes.105

Other factors

Location of delivery

Effect on success rates: there was very weak evidence which suggested this did not influence the effectiveness of FMT.39

Effect on adverse events: there were no studies.

Volume of FMT infused

Effect on success rates: there was very weak evidence which suggested this did not influence the effectiveness of FMT.26 39

Effect on adverse events: there were no studies.

PPI use

Effect on success rates: there was very weak evidence which suggested this did not influence the effectiveness of FMT.21

Effect on adverse events: there were no studies.

Antimotility agents used

Effect on success rates: there was very weak evidence which suggested these did not influence the effectiveness of FMT.21 39

Effect on adverse events: there were no studies.

Bowel lavage/preparation used

Effect on success rates: there was very weak evidence which suggested that this increases the effectiveness of FMT.21 22 39

Effect on adverse events: there were no studies.

The Working Party discussed the above evidence and concluded that most routes of administration are effective and where differences in effectiveness exist, they are subtle and not significant clinically. Thus, any of these methods can be considered for FMT delivery. Based on the current evidence presented here and in the Effectiveness and safety of FMT in treating CDI section, there is some concern that enema may be the least effective route and, as such, it is preferred that whenever possible, this should be avoided. Enema could still be considered as a method of delivery when other options are not feasible. The Working Party observed that there was no additional data regarding flexible sigmoidoscopy specifically; it was felt that given the nature of this procedure, the efficacy of FMT via this route (and therefore recommendations pertaining to it) would broadly be similar to colonoscopy, while recognising that colonoscopy allows more proximal access to the colon and therefore a higher chance of material retention (and therefore potentially success). For all routes of delivery, FMT appears to be equally safe, although there may be some general risks associated with some delivery methods (eg, endoscopy). Therefore, the Working Party recommends that other factors, such as cost, patient preference, patient safety and environmental concerns, should be considered when choosing the route of FMT delivery. As an example, when available, oral capsules could be offered to avoid unnecessary endoscopy to reduce potential unnecessary harm, cost and environmental impact.111 However, the Working Party also noted that the methods of encapsulation and the administration of encapsulated FMT to patients differ considerably between the centres and more research is currently needed to determine the most optimal regimen for this route of FMT delivery.

There is currently very little evidence that the site of delivery (within the gastrointestinal tract) is important for FMT effectiveness, and the Working Party agreed that the only important factor to consider is that FMT must be delivered to a part of the colon where it can be retained. The members agreed that bowel lavage/preparation, which is currently recommended for lower and upper gastrointestinal tract delivery, should continue in the light of the evidence suggesting a potential benefit. While the quality of the evidence is low, the Working Party concluded that there is no benefit associated with either the administration of PPI or other antisecretory medications, or antimotility medication. Therefore, PPI and other antisecretory medications are not necessary, and the Working Party advises against the use of antimotility agents in line with general consensus that these may promote C. difficile toxin retention. Additionally, there seems to be no effect associated with the volume of FMT used, although the Working Party acknowledged that it is not the volume of the infusion but the amount and concentration of the stool microbiota which is a determining factor and that the volume of faeces that needs to be infused will also depend on other factors such as water and undigested food content, and the overall mass of the stool. Future studies need to address the issue of a minimum effective dose that needs to be administered for a successful FMT.

The Working Party also discussed the effect of anti-CDI antibiotics administered before FMT. Overall, the Working Party noted that this was not supported by evidence, but, intuitively, recognised that there is a need for a balance between sufficient anti-CDI antibiotics to minimise the burden of C. difficile prior to administration of FMT and enough of a gap from the time antibiotics were given so that the risk of damaging the new microbiome is minimised. The opinion of the Working Party was that 24 hours met an appropriate balance.

Post-FMT factors influencing the outcome of FMT for patients with CDI

The risk factors for failure after administration of FMT, especially associated with the use of antibiotic therapy, started to emerge at the time the first BSG/HIS guidelines3 were about to be published. The guidelines identified two studies which mentioned a potential link between the administration of non-CDI antibiotics in a short time after the FMT was given, and subsequently suggested that antibiotic therapy should ideally not be administered within the first 8 weeks, and that an infectious disease specialist or a medical microbiologist should be consulted before the therapy is given. Other potential factors (eg, diet or the use of probiotics) have also been discussed but their influence on FMT outcome remains unclear.

Use of non-CDI antibiotics

Effect on success rates: there was weak evidence which suggested a potential negative effect on the effectiveness of FMT.19 22 23

Effect on adverse events: there were no studies.

Other post-FMT factors

Effect on success rates: there was very weak evidence which suggested these do not influence the effectiveness of FMT.15 22 23

Effect on adverse events: there were no studies.

The Working Party agreed that there is a concern, although evidence is weak, that post-FMT, non-CDI antibiotics are a potential risk factor for FMT failure. As such, the Working Party recommended that for patients who require antibiotics, either long-term or within 8 weeks of FMT, this decision needs a formal assessment and a discussion with infection specialists or other appropriate specialist healthcare professionals. Currently, there is no reason to suspect that factors other than post-FMT antibiotics are risk factors for FMT failure.

Prophylactic FMT treatment to prevent CDI

Prophylaxis has become one area of interest in CDI more broadly and FMT is proposed as a potential therapy among other more traditional agents such as vancomycin and probiotics.112 Although no studies were identified, the recognition has grown that CDI pathogenesis relates to gut microbiome disruption113; therefore, there is a biological rationale that restoration of gut microbiome in vulnerable patients (eg, patients with extensive exposure to antibiotics) via FMT could be a reasonable strategy to prevent CDI. Current debate also focuses on the definition of prophylaxis, specifically whether it should describe the prevention of recurrence or the prevention of new CDI in patients at risk. Previous BSG/HIS guidelines did not address this topic and thus, no recommendations were made.

No studies were found in the existing literature which assessed the effect of prophylactic treatment on any of the included outcomes.

Additional data from excluded studies

The Working Party is aware of one ongoing trial which aims to evaluate the effectiveness of FMT (oral capsules) for the prevention of CDI in patients with a history of CDI currently taking antibiotics.114

Due to the lack of existing evidence, the Working Party agreed that no recommendation can be made in favour or against prophylactic FMT. Instead, the Working Party suggests that studies addressing this issue should be undertaken in the future to establish its feasibility and cost-effectiveness.

FMT for non-CDI indications

In current clinical practice, FMT is only recommended for the treatment of recurrent CDI. Due to its success with CDI, FMT has been investigated for other diseases in which the gut microbiota has been implicated as a pathogenic agent. Previous BSG/HIS guidelines3 reported that the majority of the studies investigating the effectiveness of FMT for non-CDI indications were of poor design and quality, and that only a small number of RCTs existed. The conditions which were reported in the previous guidelines included ulcerative colitis, IBS, hepatic encephalopathy and metabolic syndrome, all of which showed a potential benefit. However, the lack of evidence regarding the choice of suitable patients and the most appropriate methods for FMT preparation and administration led the Working Party to a decision not to recommend FMT in the context other than research. At the time the guidelines were published, it was also noted that there were ongoing trials for other conditions. Since then, more diseases have now been linked with the gut microbiome, and a large number of systematic reviews and meta-analyses investigating the effectiveness of FMT for these conditions have become available.

Ulcerative colitis

Effect on inducing remission: there was moderate evidence which suggested FMT is effective in inducing remission in patients with ulcerative colitis.115–125

Effect on adverse events: there was strong evidence which suggested that FMT does not have an effect on the adverse events in this group of patients.115–117

Additional data from excluded studies: one study126 reported that patients who received FMT and also followed an anti-inflammatory diet were more likely to achieve remission at 8 weeks when compared with patients who received standard care.

Another study127, which assessed the effectiveness of FMT as a maintenance therapy for patients with ulcerative colitis in remission, reported that 12 months after the intervention, the incidence of remission was similar in a group of patients who received FMT from a healthy donor and those who received autologous FMT.

Crohn’s disease

Effect on success rates: there was weak evidence which suggested FMT is effective in maintaining remission in patients with Crohn’s disease.128

Effect on adverse events: there were no studies.

Pouchitis

Effect on success rates: there was weak evidence which suggested that FMT has no effect on treatment of pouchitis.129 130

Effect on adverse events: there was weak evidence which suggested that FMT does not have an effect on the adverse events in this group of patients.129 130

Irritable bowel syndrome

Effect on success rates: there was inconsistent evidence, and it was not possible to determine the effectiveness of FMT on achieving IBS remission. 120 125 127 131–143

Effect on adverse events: there was strong evidence which suggested that FMT does not have an effect on the adverse events in this group of patients.131–133

Effect on quality of life: there was moderate evidence which suggested that IBS may improve quality of life for patients with IBS.131–133

Additional data from excluded studies: one review139 suggested that while FMT may not show an overall advantage, the delivery via upper gastrointestinal (via duodenoscopy or nasojejunal tube) may be more effective than the delivery via other methods.

Constipation

Effect on success rates: there was weak evidence which suggested FMT is effective in improving symptoms in patients with functional constipation.144

Effect on adverse events: there were no studies.

Effect on quality of life: there was weak evidence which suggested FMT may improve the quality of life in patients with constipation.144

Preventing hepatic encephalopathy in patients with decompensated cirrhosis

Effect on success rates: there was weak evidence which suggested FMT is effective in preventing hepatic encephalopathy.145 146

Effect on adverse events: there was weak evidence which suggested a possible negative effect of FMT on adverse events in this patient group.145

Metabolic syndrome

Effect on success rates: there was weak evidence which suggested that FMT had no effect on improving biomarkers of metabolic syndrome.147 148

Effect on adverse events: there were no studies.

Additional data from excluded studies: four RCTs149–152 reported no improvements in most of the markers associated with metabolic syndrome.

Obesity

Effect on success rates: there was moderate evidence which suggested no effect on reducing BMI in obese patients.153

Effect on adverse events: there were no studies.

Other conditions

Literature searches were conducted for other conditions for which it was known that FMT was investigated as a potential treatment options. No studies which fit the inclusion criteria were identified for the following conditions: autism spectrum disorder, multidrug resistance, immune checkpoint inhibitor (ICI) colitis and graft versus host disease.

The searches identified other conditions which were not searched for systematically but for which RCTs now exist. These included one study which reported that FMT may halt progression of new-onset type 1 diabetes mellitus,154 one study which reported an increase in gut motility and some self-reported improvement in symptoms of Parkinson’s disease,155 one study which reported no effect on controlling peripheral psoriatic arthritis,156 and one study which reported a reduced intestinal inflammation and an improvement in symptoms of progressive supranuclear palsy-Richardson’s syndrome.157

Data from excluded studies

Infection/colonisation of gastrointestinal tract with multidrug-resistant organisms

One RCT158 reported no difference in decolonisation success when comparing patients who received FMT with antibiotics with patients who did not receive any treatment. A follow-up to this RCT159 reported that the treatment with oral antibiotics temporary decreased the richness and diversity of gut microbiota but that after the administration of FMT, the proportion of Enterobacteriaceae decreased. One review160 reported that decolonisation rates after FMT ranged from 20% to 90% for different types of microorganisms, but it reported that the spontaneous clearance was not considered in the studies.

Alcoholic hepatitis

One RCT161 reported that at 28 days and 90 days of follow-up, patients who received FMT and antibiotics had higher rates of survival and that hepatic encephalopathy and ascites resolved in more patients in this group. Another RCT162 reported that there was a lower rate of 90-day survival in patients who received prednisolone (34 of 60, 57%) when compared with those who received FMT (45 of 60, 75%, p=0.044).

The Working Party reviewed the above evidence and concluded that FMT cannot currently be recommended as a treatment of conditions other than CDI. The evidence indicates that patients with ulcerative colitis may benefit from FMT; however, at the moment, there is little information about the most effective protocols for the use of FMT in this condition and how its effectiveness and cost compare with other well-established treatment options. Most of the studies focused on the induction of remission in these patients but there is also a need for future studies to determine the role of FMT in maintaining remission. Some studies already identified that further FMT may be needed for achieving long-lasting effects.116 123 163–165 The Working Party agrees with the recent consensus166 of the experts who concluded that, at the moment, the studies are too small and methodologically heterogeneous to determine the effectiveness of FMT for IBD, including ulcerative colitis, and that the risk of serious side effects, including exacerbation of IBD, cannot be ignored. As such, the Working Party agreed that FMT may be offered to patients with ulcerative colitis who are not suitable for the licensed treatment options or in whom these options have failed. There is also weak evidence which suggests that patients with other conditions, namely Crohn’s disease, IBS and constipation, may benefit from FMT, but more research is required before any clinical decisions are made. For other conditions, including metabolic syndrome, autism spectrum, pouchitis, preventing hepatic encephalopathy, obesity and the treatment of multidrug-resistant microorganisms, further research is required to establish whether or not FMT is safe and effective. In the meantime, the Working Party agreed that FMT may be considered when conventional treatment fails, and when the patients meet the eligibility criteria for compassionate use of FMT (described in the next section).

Compassionate use of FMT

While clinical trials are a preferred option for accessing unlicensed medicinal products, this is not always possible. This may be because a patient may be too ill to enter a clinical trial, fails to meet some aspects of inclusion criteria or that no trial is ongoing at the time the treatment is needed. For this reason, compassionate use programmes (also known as early access or special access) were developed to provide access to unlicensed treatments.167 These treatments may include products which are still in clinical development or those that are already licensed in other countries. Examples of compassionate use programmes include Expanded Access Program in the USA, Compassionate Use Program (for a group of patients with a specified condition) and Named Patient Program (NPP, for named patients) in the EU and Early Access to Medicines Scheme (EAMS) in the UK.168 The EAMS is available for manufacturers to apply for the early access to their products; however, another scheme, The Supply of Unlicenced Medicinal Products (‘specials’), allows the clinicians to request the unlicensed products in a manner similar to the EU’s NPP.169 This scheme allows a supply of the medicinal products to the individual patients with special needs which a licensed product cannot meet’169 and also includes the off-label use of these products. In the section below, the term compassionate use programme was used to refer to the ‘specials’ scheme as well as other similar programmes in other countries.

Since publication of the last iteration of the guidelines, the range of medical conditions with a potential pathogenic link to a perturbed gut microbiome has continued to expand. A number of these conditions have no or limited treatment options. In many cases, the Working Party recognised that these remained associations, often without clear supporting mechanistic links that might deconvolute whether gut microbiome perturbation was a cause of the condition, consequence or an epiphenomenon. A body of research has also explored whether FMT, alongside a conventional drug treatment, might augment the efficacy of that therapy, help to recover efficacy where this has been lost, or mitigate side effects of that medication. One prominent example of this scenario is cancer immunotherapy with ICIs, where early-phase trial evidence suggests healthy donor FMT prior to anti-PD1 treatment for melanoma may boost efficacy in a subset of patients.170 Further clinical trials demonstrated that FMT derived from anti-PD1 responders may be used to regain treatment response in certain patients with melanoma who had become refractory to treatment,171 172 and also shows promise as an approach to potentially mitigate ICI-induced colitis in patients refractory to conventional immunomodulatory therapy.173

The Working Party discussed their clinical experience of considering potential suitability of FMT for patients with non-CDI medical conditions associated with perturbation of the gut microbiome. They felt that if all below three criteria were fulfilled, there were potential grounds for consideration of administration of FMT on a compassionate use basis.

  • There was a reasonable case from published literature to support a contribution of the gut microbiome to pathogenesis of the condition, and at least some published data relating to safety and efficacy of FMT in either a preclinical or clinical setting for this condition.

  • The patient had been unresponsive to/was not suitable for a range of conventional treatment options for their condition and had very limited treatment alternatives, which had already been used. The scenario in which this is envisaged is one in which the limited ability to provide further effective treatment of the condition may cause significant ongoing symptoms, significantly impair the patient’s quality of life, and/or may risk progressive morbidity or even mortality for the patient.

  • The patient understood the treatment options that were available, including the potential risks and benefits of FMT (especially the potential for no benefit and/or complications related to the FMT), but was still willing to provide informed consent for FMT.

However, the Working Party emphasised that a few additional criteria merited consideration. First, it must be determined that a patient cannot be entered into ongoing relevant clinical trials and potentially receive FMT instead via this pathway. Second, such cases should be considered in an MDT setting (including senior clinical representation from the specialist team referring the patient and clinicians with experience in FMT, likely with a background in gastroenterology or microbiology/infectious diseases). The role of this MDT is to better clarify any prior experience of FMT within this setting, and/or the balance of risks and benefits from FMT versus alternative treatment options. Third, there should be agreement as to what should be defined as success or failure of FMT in this particular scenario. There must also be a plan prior to treatment initiation, for a strategy regarding potential further FMT based on the response to the initial therapy. Lastly, there should be comprehensive documentation/reporting of clinical data (and/or potentially stool and other biofluids collected from the patient for research, where such a resource exists) related to the outcome of this patient from FMT, to build knowledge and experience of the potential role for FMT within novel settings.

Self-banking of stool for potential future autologous FMT

The Working Party members reported that, in the past, they have been contacted by other clinicians and by patients enquiring about banking their own stool with a view to potential future autologous FMT. One such scenario might be a patient who has been informed about the imminent need for medical treatment which might be expected to significantly disrupt their gut microbiome, that is, a prolonged course of antibiotics that might risk CDI, or a patient due to undergo intestinal surgery, immunosuppression, etc. The Working Party discussed the published literature regarding this approach, including clinical evidence that stool collected from patients prior to their haematopoietic cell transplantation (HCT) could safely be given as FMT to them post-HCT, with associated restoration of pre-morbid microbiome diversity and composition.174 A further enquiry that the Working Party had received related to whether a person in entirely good health could be considered for stool banking in case the scenario arose whereby autologous FMT might become an appropriate treatment option at some point in the future based on changes of their health status. This conceptually might be considered to have a degree of comparability with cord blood banking, for which there is a Human Tissue Authority-regulated structure in the UK.175

The Working Party recognised some of the challenges related to this, which have already been discussed elsewhere.176 Firstly, there are uncertainties related to how much stool might optimally be stored (with associated resource issues, such as freezer capacity) and for how long (raising concerns about the long-term stability of a gut microbiome community when potentially frozen for a prolonged period). Given that many conventional potential healthy stool donors fail screening due to the stringency of the process, there is a reasonable likelihood that a significant proportion of those considering self-stool banking would also fail conventional screening. While the fact that the patients would be receiving autologous FMT may reduce health risks compared with unrelated donor stool, there are clear issues related to laboratory processing and storage of material, particularly from a regulatory perspective, if this does not reach the same status on pathogen screening as healthy donor faecal material conventionally prepared into FMT. Other outstanding issues related to the regulatory framework which might govern this process, and/or potential funding arrangements and cost-effectiveness of such an approach. As such, the Working Party concluded that while self-stool banking was of potential interest, it could not be currently advocated. However, this can be considered as a concept for further studies.

Regulation and oversight of FMT

There is no agreed definition as to what constitutes FMT, nor its active pharmaceutical ingredient(s), not its mechanism of action. This leads to variability in how and what is classified as FMT, and how it should be regulated. Briefly, FMT is either a biological product (eg, USA), human tissue product (eg, Italy), medicinal product (eg, UK) or medical procedure (eg, Denmark).177 In the UK, FMT is considered an unlicensed medicinal product that may be prepared, prescribed and administered to patients on a named basis under section 10 of the Medicines Act, 1968178 (‘pharmacy exemption’), provided that defined conditions are met. These include that the medicinal product is prepared or dispensed in a hospital or health centre by, or under the supervision of, a pharmacist, and in accordance with a doctor’s prescription. This process is overseen by regional Specialist Pharmacy Services Quality Assurance. If FMT is prepared as an unlicensed medicinal product and is to be shipped to another hospital or health centre for administration, this requires a licence to supply unlicensed medicinal products (‘specials’).169 Licensed facilities are regulated and audited by the Medicines and Healthcare Products Regulatory Agency (MHRA). If FMT is used as part of a clinical trial, it is considered an Investigational Medicinal Product (IMP) and must be manufactured in a Manufacturer’s/Importation Authorisation IMP-licensed facility adhering to Good Manufacturing Practice.179 Each batch should be released by a qualified person against an approved, trial-specific, Investigational Medicinal Product Dossier prior to participant administration. Licensed facilities are regulated and audited by the MHRA, and all trials must have received Clinical Trials Authorisation, among other approvals, prior to participant recruitment.

Further research

As highlighted above, there are gaps in the evidence for almost every topic presented in these guidelines. While the list is not exhaustive, the Working Party made some recommendations for research which they thought represented current research priorities.

Further considerations: next-generation FMT and novel microbiome therapeutics

The Working Party discussed several microbiome therapeutics, which have evolved from FMT, and are at various stages of development and clinical trials. There are several different approaches being used, including full spectrum microbiome products (which have the most direct comparability with conventional FMT), as well as products involving particular microbiome components (eg, spore-based therapies or defined microbial consortia). At the time of writing, two microbiome therapeutics have been approved by the US FDA for prevention of CDI relapses, namely RBX2660/Rebyota (Ferring; a rectally administered FMT-type product8) and SER-109/Vowst (Seres/Nestle; a purified spore-based product9); no such products have been licensed for the use in any non-CDI indication.

The Working Party discussed their expectation that several early and late-phase clinical trials involving such products were ongoing globally, and there was a reasonable expectation of applications for licensing for use within the UK within the lifespan of this guideline. If such licensing was granted, there would be clear implications for use of ‘conventional’ FMT within the UK. For instance, licensing of a microbiome therapeutic for use in recurrent CDI would potentially negate the ability to supply FMT under a UK specials licence, given that FMT is an unlicensed medicinal product. This may potentially also impact upon the ability to use FMT within a UK research setting, where there is currently highly active clinical and translational research activity.

The Working Party concluded that there was a clear need for ongoing dialogue between entities developing novel microbiome therapeutics, academic and hospital centres providing FMT, and regulators to ensure no interruption at any point in provision of therapy to eligible patients with CDI, and that clinical and translational FMT/microbiome therapeutics research in this field in the UK remains globally competitive.

The Working Party concluded that the following topics are now resolved and should not be included for an update in the future editions of the guidelines:

  1. Effectiveness of FMT for recurrent CDI versus anti-CDI antibiotics/placebo in the general population. This topic can be revisited if new therapies, more effective than current antibiotic treatment, become available. Topics in relation to patients with different conditions and factors related to CDI infections (eg, severity, first occurrence) should still be investigated.

  2. Non-modifiable recipient factors, for example, age. Current evidence suggests that these factors do not reduce the effectiveness of FMT to the point where recommendations would change. Future studies need to focus on identifying modifiable recipient and donor factors, optimising FMT administration and preventing CDI recurrence after FMT.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • BHM, BM and MNQ are joint first authors.

  • SDG and HRTW are joint senior authors.

  • X @bhmullish, @jonathansegal85

  • Presented at Published concurrently by both BMJ Publishing Group Limited and the British Society of Gastroenterology in GUT and by Elsevier Ltd on behalf of the Healthcare Infection Society in the Journal of Hospital Infection. The article is identical in each publication except for minor stylistic and spelling differences in keeping with each publication’s style. Citations from either of the two publications can be used when citing this article.

  • Contributors DJM helped design literature strategy. BHM, BMerrick, MNQ and AB screened the literature. AB conducted searches, performed initial data extraction and evidence syntheses, which were checked by BHM, BMerrick and MNQ. All authors except AB also provided advice. BHM, BMerrick, MNQ and AB wrote the first draft. BHM, BMerrick, MNQ, AB, CAG, DJM, RJP, NTE, JPS, NS, BMarsh, GK, SEM, ALH, CS, JJK, PH, THI, SDG and HRTW all attended Working Party meetings and contributed to rating of evidence and recommendations. BHM, BMerrick, MNQ, AB, CAG, DJM, RJP, NTE, JPS, NS, BMarsh, GK, SEM, ALH, CS, JJK, PH, THI, SDG and HRTW all read, reviewed, contributed to the writing of and approved the final documents.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests BHM has received consultation honoraria from Ferring Pharmaceuticals, Finch Therapeutics, Summit Therapeutics and speaker fees from Yakult. BMerrick received speaker fees from New Scientist. MNQ received speaker fees from BMS, Parapharm, Tillotts, Janssen and Takeda. NTE has received speaker fees from a company not related to the topic of this guideline. JPS received speaker fees from Takeda, AbbVie, Pfizer and BMS. NS received consultancy fees for advisory board for Pharmacosmos. ALH received fees for consultancy and speaker for AbbVie, Arena, Atlantic, Bristol-Myers Squibb, Celgene, Celltrion, Falk, Galapagos, Lilly, Janssen, MSD, Napp Pharmaceuticals, Pfizer, Pharmacosmos, Shire and Takeda; and also serves on the Global Steering Committee for Genentech. CS is a chair for the Healthcare Infection Society. JJK received consultancy fees from Vedanta Biosciences and Microviable Therapeutics and received a research grant from Vedanta Biosciences; and is a member of the scientific committee of European Helicobacter Microbiota Study group, member of the Scientific Advisory Board of 'Microviable Therapeutics', a member of a steering committee scientific study VE202 for ulcerative colitis (Vedanta Biosciences) and a founder and board member of Netherlands Donor Feces Bank (academic, non-profit). PH received consultancy and speaker fees from commercial companies and is a director of Modusmedica (since 2010 onwards). THI received consultancy fees from Ferring and speaker fees from Pharmacosmos and is a member of BSG/HIS microbiome for health expert panel and a director of University of Birmingham Microbiome Treatment Centre. SDG received consultancy honoraria from Enterobiotix and AstraZeneca, received speaker fees from Tillotts and has investments in biomedical company not related to the topic of this guideline. All other authors declared no conflicts of interest.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.