Article Text

Safety and efficacy of faecal microbiota transplantation for active peripheral psoriatic arthritis: an exploratory randomised placebo-controlled trial
Free
  1. Maja Skov Kragsnaes1,2,
  2. Jens Kjeldsen3,
  3. Hans Christian Horn1,
  4. Heidi Lausten Munk1,
  5. Jens Kristian Pedersen4,
  6. Søren Andreas Just4,
  7. Palle Ahlquist5,
  8. Finn Moeller Pedersen3,
  9. Maarten de Wit6,
  10. Sören Möller2,7,
  11. Vibeke Andersen8,9,
  12. Karsten Kristiansen10,11,
  13. Dorte Kinggaard Holm12,
  14. Hanne Marie Holt13,
  15. Robin Christensen2,14,
  16. Torkell Ellingsen1,2
  1. 1 Rheumatology Research Unit, Department of Rheumatology, Odense University Hospital, Odense, Denmark
  2. 2 Department of Clinical Research, University of Southern Denmark, Odense, Denmark
  3. 3 Department of Medical Gastroenterology, Odense University Hospital, Odense, Denmark
  4. 4 Section of Rheumatology, Department of Medicine, Svendborg Hospital, Svendborg, Denmark
  5. 5 Reumaklinik Fyn, Odense, Denmark
  6. 6 Patient Research Partner, Amsterdam, The Netherlands
  7. 7 OPEN – Open Patient data Explorative Network, Odense University Hospital, Odense, Denmark
  8. 8 IRS-Center Sønderjylland, University Hospital of Southern Denmark, Aabenraa, Denmark
  9. 9 Institute of Molecular Medicine, University of Southern Denmark, Odense, Denmark
  10. 10 Laboratory of Genomics and Molecular Biomedicine, Department of Biology, University of Copenhagen, Copenhagen, Denmark
  11. 11 Institute of Metagenomics, BGI-Shenzhen, Shenzhen, China
  12. 12 Department of Clinical Immunology, Odense University Hospital, Odense, Denmark
  13. 13 Department of Clinical Microbiology, Odense University Hospital, Odense, Denmark
  14. 14 Section for Biostatistics and Evidence-Based Research, the Parker Institute, Bispebjerg and Frederiksberg Hospital, Copenhagen, Denmark
  1. Correspondence to Professor Torkell Ellingsen, Rheumatology Research Unit, Department of Rheumatology, Odense University Hospital, Odense, Denmark; torkell.ellingsen{at}rsyd.dk

Abstract

Objectives Although causality remains to be established, targeting dysbiosis of the intestinal microbiota by faecal microbiota transplantation (FMT) has been proposed as a novel treatment for inflammatory diseases. In this exploratory, proof-of-concept study, we evaluated the safety and efficacy of FMT in psoriatic arthritis (PsA).

Methods In this double-blind, parallel-group, placebo-controlled, superiority trial, we randomly allocated (1:1) adults with active peripheral PsA (≥3 swollen joints) despite ongoing treatment with methotrexate to one gastroscopic-guided FMT or sham transplantation into the duodenum. Safety was monitored throughout the trial. The primary efficacy endpoint was the proportion of participants experiencing treatment failure (ie, needing treatment intensification) through 26 weeks. Key secondary endpoints were change in Health Assessment Questionnaire Disability Index (HAQ-DI) and American College of Rheumatology (ACR20) response at week 26.

Results Of 97 screened, 31 (32%) underwent randomisation (15 allocated to FMT) and 30 (97%) completed the 26-week clinical evaluation. No serious adverse events were observed. Treatment failure occurred more frequently in the FMT group than in the sham group (9 (60%) vs 3 (19%); risk ratio, 3.20; 95% CI 1.06 to 9.62; p=0.018). Improvement in HAQ-DI differed between groups (0.07 vs 0.30) by 0.23 points (95% CI 0.02 to 0.44; p=0.031) in favour of sham. There was no difference in the proportion of ACR20 responders between groups (7 of 15 (47%) vs 8 of 16 (50%)).

Conclusions In this first preliminary, interventional randomised controlled trial of FMT in immune-mediated arthritis, we did not observe any serious adverse events. Overall, FMT appeared to be inferior to sham in treating active peripheral PsA.

Trial registration number NCT03058900.

  • arthritis
  • psoriatic
  • therapeutics
  • inflammation

Data availability statement

Data are available upon reasonable request. Requests on data sharing can be made by contacting the corresponding author. Data will be shared after review and approval by the trial scientific board, and terms of collaboration will be reached together with a signed data access agreement.

Statistics from Altmetric.com

Key messages

What is already known about this subject?

  • Psoriatic arthritis (PsA) is a systemic immune-mediated disease associated with subclinical gut inflammation and dysbiosis of the intestinal microbiota.

  • Faecal microbiota transplantation (FMT) has demonstrated local therapeutic immune-modulating abilities in patients with chronic inflammatory bowel disease.

What does this study add?

  • In this first preliminary, randomised controlled trial of FMT in immune-mediated arthritis, transfer of donor microbiota was safe, but appeared inferior to sham in reducing disease activity in patients with active peripheral PsA concomitantly treated with methotrexate.

How might this impact on clinical practice or future developments?

  • Whether microbial dysbiosis or specific bacteria are common or decisive mediators of disease activity in PsA and whether this proposed relation can be modified without exacerbating the disease will be crucial to clarify to determine the future role of microbiota-targeted interventions in the management of PsA.

Introduction

For a century, the link between enteric infections and reactive arthritis1 has motivated investigation into the proposed gut–joint axis implicating intestinal micro-organisms in the aetiology of immune-mediated arthritic disease.2 Recently, this theory has gained renewed interest due to accumulating evidence of disease-related imbalance (dysbiosis) in the composition and function of the intestinal microbiota in chronic disorders.3–5 Among these, psoriatic arthritis (PsA)6 has been associated with decreased intestinal bacterial diversity, displaying both disease-specific patterns7 and microbial abnormalities similar to those seen in other subtypes of spondyloarthritis, rheumatoid arthritis and inflammatory bowel disease (IBD).8 These findings have encouraged research into the host–microbiota interplay in the dysregulated immunological cascade underlying immune-mediated arthritis and the prospects of microbiota-targeted therapies.9

Faecal microbiota transplantation (FMT) is currently considered the most efficient method to restore a healthy diversity of the gastrointestinal microbiota.10 11 Indeed, the transfer of faeces containing minimally manipulated communities of micro-organisms from a donor to a recipient has revolutionised the treatment of Clostridioides difficile infection.12 FMT may also induce beneficial responses in patients with IBD, thereby demonstrating local therapeutic immune-modulating abilities.13 However, whether manipulation of the intestinal microbiota can treat extraintestinal, immune-mediated disorders remains to be established.14 This is the first exploratory, randomised trial to assess the safety and efficacy of FMT in patients with active, peripheral PsA.

Methods

Trial design

This is a proof-of-concept, 26-week, 1:1 randomised, parallel-group, double-blind, placebo-controlled, single-centre superiority trial. In 2015, the Regional Committees on Health Research Ethics for Southern Denmark (DK-S-20150080) and the Danish Data Protection Agency (15/41684) approved the trial protocol (see online supplemental appendices S1 and S1A). Although not required by the Danish Health and Medicines Authority, we fulfilled the requirements of documentation, monitoring and reporting according to the principles of Good Clinical Practice. We registered the trial with ClinicalTrials.gov. Our trial protocol paper was published in 2018.15 The study was conducted at one Danish tertiary referral hospital with nationwide inclusion. All participants gave written informed consent. The trial was temporarily suspended from March to May 2020 due to the COVID-19 pandemic (see online supplemental appendix S4). A statistical analysis plan (see online supplemental appendix S2) was developed with subsequent closure (2 April 2020) before unmasking and analysis (10 June 2020).

Participants

We included adults who were between 18 and 75 years of age, fulfilled the Classification for Psoriatic Arthritis criteria,16 and had active peripheral disease, defined as three or more swollen joints, despite ongoing treatment with methotrexate at the maximal tolerable dose (≥15 mg/week) for at least 3 months prior to study inclusion. A washout period of 12 weeks (26 weeks for biologic agents) was required in patients previously treated with intra-articular or systemic glucocorticoids, and non-methotrexate conventional synthetic and biologic disease-modifying antirheumatic drugs. Key exclusion criteria were immune-mediated arthritis other than PsA, IBD, cancer, severe chronic infection, and history of food allergy, severe food intolerance or coeliac disease.15

Donor selection and stool preparation

The transplants were obtained from four healthy stool donors recruited from a non-profit, public stool bank located at the local blood and tissue transplant service.17 Every step of the donation process and the laboratory handling were in agreement with the requirements of the European Union’s regulative directives on human cells and tissues (2004/23/EC). Donors did not receive any compensation and had to pass an extensive screening programme (see online supplemental table S3) before and after the 30-day donation cycle.15 Stool donations were transported to the stool bank facility within 1 hour after defaecation in an airtight container placed in a cooling bag. The donation was processed at normal room temperature within 2 hours of delivery under aerobic conditions, including 10–15 s of blending, before storage at −80°C (median storage time: 20 months; range: 2.5–30).18

Interventions

The transplant consisted of a single stool donation (50 g) mixed with saline (0.9%) and glycerol (10%) to a total volume of 250 mL. Before transplantation, we thawed the material to 36°C. The sham transplant consisted of 250 mL saline (0.9%) mixed with three drops of food colouring (E150c). We performed the allocated treatment within 14 days of the baseline visit. Treatment preparation included a 6-hour fast and one dose of oral proton-pump inhibitor. The transplant suspension was transferred into the third part of the duodenum via a closed system of tubes under gastroscopic guidance.

Outcomes

Safety was monitored by open assessment of adverse events (AEs) and evaluated before unmasking. The National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0 was used to grade the severity.19 We evaluated causality for expected AEs of grade 2 and above and for all unexpected AEs regardless of severity. The primary efficacy endpoint was a composite outcome on the proportion of participants who experienced treatment failure through 26 weeks, defined as need for at least one of the following: more than one intra-articular glucocorticoid injection, and non-methotrexate conventional synthetic and/or biologic disease-modifying antirheumatic drugs. This endpoint covered disease activity and shared decision-making between the patient and the rheumatologist in accordance with the European PsA recommendations.20 Three key secondary endpoints were evaluated at week 26: change from baseline in Health Assessment Questionnaire Disability Index (HAQ-DI),21 proportion of participants fulfilling the American College of Rheumatology (ACR20) response22 and change from baseline in the Spondyloarthritis Research Consortium of Canada (SPARCC) Enthesitis Index.23 Additional secondary outcomes addressed all essential domains in the PsA core outcome set24 (see table 1).

Table 1

Comparison of efficacy endpoints at week 26

Randomisation and masking

We assigned participants to FMT or sham transplantation using permuted blocks with varying sizes of four and six, according to computer-generated random numbers. These lists were generated by the trial statistician and uploaded to a concealed area of a password-protected trial database (REDCap25) by an independent and otherwise trial agnostic data manager.15 The trial coordinator implemented the randomisation. The randomisation record and the signed transfusion journals were stored in the database with restricted access separate from the patient record and other study data so the participants and the treating rheumatologists (ie, care providers and outcome assessors) remained unaware of treatment allocation and treatment.

Sample size and power considerations

Conceptually guided by the idea that at least twice as many participants in the sham group would be treatment failures, compared with the FMT group if the procedure should potentially be considered clinically relevant, we wanted to randomly assign 80 patients with PsA to two groups (40 patients to each), providing a good statistical power (90%) to detect a difference between two proportions (35% vs 70%) with a significance level of 5% (see further details in online supplemental appendices S1 and S2). In April 2019, we decided to stop recruitment by 31 December 2019, thereby adhering to the original planned trial completion date of 1 July 2020 (see online supplemental appendix S3). The main reason for this was that essential funding would stop following this date. Due to a slower than expected recruitment rate, only 31 participants were enrolled.

Patient involvement

Patients were directly involved in the design, funding, recruitment, conduct, reporting and dissemination of the study (see description at the end of the manuscript).

Statistical analysis

Analyses were based on the intention-to-treat (ITT) population including all randomised individuals, independent of subsequent adherence to the trial protocol. To assess the effect of FMT on the risk of treatment failure during the 26-week trial, we compared groups using risk ratios with 95% CIs based on an unadjusted (crude) model. Time-to-treatment failure was analysed based on a Kaplan-Meier plot from baseline to week 26 using Cox regression to estimate the HR with 95% CI and p value. Primary analyses in the ITT population at week 26 were based on a conservative treatment failure imputation default option for binary outcomes and a mixed-effects repeated measures model for continuous variables. For continuous outcome measures, we modelled between-group differences in outcomes at 0, 1, 2, 3, 4, 12 and 26 weeks with mixed-effects models, using time and group as categorical fixed-effect factors, interactions between time and group, random intercepts, and an unstructured covariance matrix; these models are reported as least squares means (and the difference between them) with 95% CIs.

All p values and 95% CIs were two-sided. We did not apply adjustments for multiplicity, rather we analysed the three key secondary outcomes in a prioritised order with a gatekeeping rule for serial testing. The other secondary outcomes are presented without conducting any formal statistical testing. Safety data are summarised descriptively in the full analysis set. In addition, we conducted a per-protocol analysis (see online supplemental table S1 and online supplemental figure S1) which included participants in whom the assigned transplant was successfully transferred into the third part of the duodenum. In addition, participants were excluded (censored) from the analysis following the day where they were categorised as treatment failures. Results of additional sensitivity analyses (including analyses based on the ITT population with missing data handled with multiple imputation) are presented in online supplemental table S2. We performed analyses using StataSE-64 V.16.1.

Results

Participants

We enrolled participants to the trial between 16 May 2017 and 11 December 2019 at one Danish tertiary referral hospital (nationwide recruitment) with 26-week follow-up until 2 June 2020. Of 97 patients screened, 31 (32%) fulfilled the study requirements and were randomised to FMT or sham transplantation. All received the assigned intervention and 30 (97%) completed the clinical evaluation at week 26 (figure 1). Demographics and clinical characteristics of the two groups were comparable at baseline (table 2), except for some imbalance in sex and disease duration. Of the 31 participants, 27 (87%) had a personal history of skin psoriasis. A complete list of medication is presented in online supplemental table 4.

Table 2

Baseline demographics and disease characteristics

Figure 1

Patient disposition. Reasons for not meeting study criteria (n=48): not diagnosed with psoriatic arthritis (n=5), not ≥3 swollen joints (n=11), treated with other csDMARD (n=2) or bDMARD (n=5), methotrexate (≥15 mg/week) toxicity (n=6), age below or above limit (n=3), systemic inflammatory comorbidity (n=1), living abroad (n=14) and closed inclusion 2 days after initial contact (n=1). bDMARD, biologic disease-modifying antirheumatic drug; csDMARD, conventional synthetic disease-modifying antirheumatic drug; FMT, faecal microbiota transplantation; ITT, intention-to-treat.

Safety

We observed no serious AEs or deaths in any of the groups. Forty-seven AEs occurred in 14 participants (93%) receiving FMT, and 53 AEs occurred in 14 participants (88%) receiving sham. No participants withdrew from the trial due to AEs. In the FMT group, AEs were mainly related to the gastrointestinal tract and included abdominal discomfort, flatulence, nausea and vomiting. One case of diverticulitis was deemed unrelated to FMT because the participant had a history of diverticulitis before trial enrolment and the episode occurred 24 weeks after FMT.

Three AEs were deemed related to the gastroscopic procedure. Of those, facial capillary rupture and uncontrolled defaecation both occurred within an hour of the procedure. The third procedure-related AE was exacerbation of known asthmatic disease (grade 2) 3 days after the procedure. This participant vomited during the gastroscopy. We did not suspect pulmonary aspiration, and the clinical evaluation (including measures of C reactive protein) performed by the participant’s general practitioner 4 and 7 days after the procedure did not suggest an underlying bacterial aetiology. A complete list of AEs, routine laboratory findings and metabolic changes is presented in table 3.

Table 3

Adverse events through 26 weeks

Efficacy

During the entire 26 weeks of observation, the rate of the primary outcome (treatment failure) was significantly higher in the FMT than in the sham group (HR, 4.87 (95% CI 1.31 to 18.18); p=0.018) (see figure 2). After 26 weeks, treatment had failed in more FMT-treated participants (9 of 15, 60%) than sham-treated participants (3 of 16, 19%) (crude relative risk, 3.20 (95% CI 1.06 to 9.62); p=0.018) (see table 1). Starting biologic disease-modifying antirheumatic drug was the main reason for being categorised as treatment failure (8 of 15 (53%) vs 3 of 16 (19%)). In the FMT group, eight patients (53%) had initiated biologic therapy at the 12-week visit compared with two (13%) in the sham group. The median time from trial intervention to starting biologics was 32 days (IQR 23–64) in the FMT group and 99 days (IQR 60–175) in the sham group. The median time from starting biologics to evaluation at week 26 was 155 days (IQR 118–173; total group exposure time, 995 days) in the FMT group and 70 days (IQR 0–126; total group exposure time, 196 days) in the sham group.

Figure 2

Event curves of treatment failure by intervention group from baseline to week 26. Time-to-treatment failure was analysed using survival analysis as Kaplan-Meier curves from baseline to week 26. Cox regression was used to provide an estimate of the HR with 95% CI and p value. The number of participants remaining at risk is displayed below the horizontal axis. Treatment failure was defined as need for at least one of the following: more than one intra-articular glucocorticoid injection, and non-methotrexate conventional synthetic and/or biologic disease-modifying antirheumatic drugs. FMT, faecal microbiota transplantation.

The HAQ-DI decreased more (indicating better physical function) in the sham group than in the FMT group (least squares means, −0.30 (95% CI −0.44 to −0.15) vs −0.07 (95% CI −0.22 to 0.09), difference 0.23 (0.02 to 0.44); p=0.031). The trajectories for HAQ-DI by treatment group from baseline to week 26 are presented in figure 3. Hierarchical statistical testing failed with regard to the proportion of ACR20 responders when comparing the FMT group with sham (7 of 15 (47%) vs 8 of 16 (50%); crude relative risk, 0.93 (0.45 to 1.94)).

Figure 3

HAQ-DI scores by treatment group from baseline to week 26. Least squares means calculated from the repeated measures linear mixed model: intention-to-treat population. Bars indicate 95% CI. HAQ-DI scores range from 0 to 3, with higher scores indicating greater disability. FMT, faecal microbiota transplantation; HAQ-DI, Health Assessment Questionnaire Disability Index.

Discussion

Modification of the immunogenic, intestinal microbial communities and their metabolites associated with immune-mediated diseases has been highlighted as a possible way to either directly or indirectly modulate a dysregulated immune response in the recipient. In this first FMT trial in PsA, we performed one upper, single-donor FMT to evaluate safety and efficacy in patients with active, peripheral PsA treated with steady state dose methotrexate (≥15 mg/week). Most importantly, one FMT appeared to be safe in this patient population. Contrary to our hypothesis, the rate of treatment failure (ie, patients needing treatment intensification) was significantly higher in the FMT group than in the sham group. Likewise, the HAQ-DI improved more in sham-treated (indicating better physical function) than in FMT-treated patients.26 The event curve (figure 2) demonstrates that treatment failure occurred very quickly after the procedure in patients receiving FMT. Because of the comparable disease activity between groups at baseline, our findings suggest that FMT from selected donors can worsen the symptoms of PsA. This contrasts a case report of a patient with PsA where FMT seemed to assert beneficial effects on the arthritic disease.27

FMT for other immune-mediated conditions such as IBD has demonstrated local therapeutic immune-modulating abilities, and disease flares following FMT seem to occur in similar rates among control group patients.13 Although transient increase in C reactive protein and self-limiting fever are well-known side effects associated with an immunological response in patients receiving FMT for C. difficile infection and/or IBD,28 based on the existing evidence, thoroughly screened stool for FMT is in general considered safe and has not been related to severe immune responses. Hence, our findings add to the growing body of evidence suggesting a gut–joint axis in the pathogenesis of PsA.29 30

A strength of our study is that we designed the trial to provide results relevant to clinical practice. This included the use of FMT products from routine treatment of recurrent C. difficile infections, the timing of the intervention, the long follow-up, the allowance of antibiotics and other medication during follow-up (except for non-methotrexate disease-modifying antirheumatic drugs), and the lack of diet restrictions. Due to the randomised design, allocation concealment, masking of patients and treating rheumatologists/outcome assessors, high treatment adherence, low attrition, no missing data on the primary outcome, and only few cases of protocol violations (see online supplemental text S1), we deem the risk of bias to be low. Furthermore, because our decision to stop the trial before reaching 80 patients was made independently of the trial findings, we do not expect that this decision has biased the results.31 Nevertheless, the small study population did affect the precision of the trial estimates, making the conclusions less robust.

Limitations include the initiation of additional disease-modifying antirheumatic drugs in patients experiencing treatment failure, which may likely have exerted positive effects on the secondary outcomes (HAQ-DI, ACR20 and SPARCC Enthesitis Index) evaluated at week 26. Although this may especially be true for the FMT group, where the majority of participants received biologics within the first 12 weeks of the trial, this explanation is not valid in the sham group, where less than one in five received additional treatment. Hence, the significant clinical improvement observed in this group suggests the presence of trial participation effects.32 Findings from a qualitative study nested within the trial support this notion.33 Nevertheless, based on the double-blind, randomised design and the fairly comparable demographics and disease characteristics of FMT-treated and sham-treated patients at baseline, we have no reason to believe that exposure, intensity and susceptibility of these effects differed between groups.34

The compositional nature of the primary endpoint combining both the patient’s values, preferences and needs in relation to the outcome domains that were important to him/her (eg, pain, physical function, fatigue and social participation) with the physician’s assessment of disease (eg, musculoskeletal disease activity and systemic inflammation), in addition to the very high ACR50/70 response in the sham group, complicated the interpretation of the trial results. For example, we cannot rule out that parts of patients’ perceptions of the disease, and the resulting motivation for receiving additional treatment, could have been affected by elements of the disease not related to active inflammation (which was our hypothesised target of the FMT) such as central sensitisation and structural damage. In addition, while measures of disease activity appeared comparable between groups at baseline, the (random) imbalance in sex and disease duration could hypothetically have had an effect on the between-group differences in the primary and secondary outcomes.

Furthermore, the participants of our trial constituted primarily adults with active, polyarticular PsA, which is a relatively rare condition in clinical practice.20 Consequently, although the spectrum of patients with PsA that are enrolled in pharmacological trials is skewed towards this study population, the ability to generalise our findings to the majority of patients with PsA is limited. Moreover, because only 10 participants (32%) had active skin psoriasis at the time of inclusion, this trial was not suited for assessment of the potential of microbiota modulation in cutaneous inflammation. Finally, our study was neither large enough nor long enough to evaluate uncommon serious AEs and long-term risks.35

Previous findings in patients with ulcerative colitis indicate that one FMT regimen performed within 1 week is insufficient to maintain long-lasting (12 months) local anti-inflammatory response in the majority of patients with a beneficial response after 8 weeks.36 Hence, we could have missed early clinical significant changes in the secondary outcomes that abated before the 26-week evaluation. In addition, pooled donor batches and high-intensity induction of FMT followed by frequent administration of donor transplant seem to enhance the chances for clinical remission in ulcerative colitis.36–38 Perhaps as importantly, previous successes of FMT in IBD appear to have been driven by ‘super-donors’ characterised by the presence or absence of specific bacteria species.36 39 To further complicate this picture, matching of donor and recipient could be another important factor to consider.

In conclusion, further investigation is needed to explore whether extrinsic factors related to the FMT method, such as single versus multiple donor batches, fresh versus frozen products, aerobic versus anaerobic environment, type of stool preparation protocol, storage time, freeze–thaw cycles, pretreatment preparation such as bowel lavage and antibiotics, delivery form, and overall treatment strategy (dose and frequency), may influence the outcome of FMT in PsA.40 Future FMT trials could pursue an approach that is similar to the ones used in IBD studies.41 Moreover, evaluation of dysbiosis in patients (or donors) prior to trial entry could hypothetically enhance FMT efficiency. Other mechanisms that should be thoroughly investigated in future studies include the degree and durability of donor microbiota engraftment, changes in patients’ microbiota following FMT and comparator intervention, and characterisation of ‘good’ and ‘bad’ donations. Indeed, the lack of comprehensive microbiota analyses is a limitation of the current study.

In this preliminary randomised controlled trial with focus on safety, we did not observe any serious AEs. Although no firm conclusions can be drawn from this small trial and despite the similar proportions of ACR20 responders between groups at 26 weeks, our findings indicate that FMT may lead to worsening of PsA, suggesting a role of the intestinal microbiota in downstream immune effects of this disease. Larger, randomised trials of FMT where a sufficient amount of participants will be included combined with exploration of immunological effects and indepth analyses of the composition and functional potential of the microbiota in donor and recipients should be undertaken to further investigate the safety and potential benefits of therapeutic targeting of the gut–joint axis in immune-mediated arthritis.

Patient and public involvement

Patients and the public played an important part in all elements of the research process beyond the conception of the trial. In the design phase, we asked patients who attended the outpatient clinic to provide input on ethical challenges and trial logistics, especially regarding donor selection and method of FMT administration. Patients also gave feedback to the wording and design of patient information material and recruitment flyers. During the undertaking of the trial, we asked patient organisation and the public to assist the refinement of our recruitment strategy and to help with publicity and funding. Following the last trial visit and before unmasking, we interviewed 10 participants about their trial experiences, which provided insight into their motivation for participation, impact on everyday life, FMT acceptability, and factors related to the trial that may have promoted trial participation effects (the results of this qualitative study will be presented elsewhere). In the dissemination phase of the trial, we invited a patient research partner (MW), who did not participate in the trial, to become coauthor. He made valuable suggestions for improving the reporting of the study and helped us clarify the main findings as seen from a patient’s perspective. Finally, we disseminated the results of the trial in a letter to all study participants and invited them to attend an online meeting, where we further explained and discussed the findings of the trial.

Data availability statement

Data are available upon reasonable request. Requests on data sharing can be made by contacting the corresponding author. Data will be shared after review and approval by the trial scientific board, and terms of collaboration will be reached together with a signed data access agreement.

Ethics statements

Ethics approval

The study was approved by the Regional Committees on Health Research Ethics for Southern Denmark (DK-S-20150080).

Acknowledgments

We thank all participants for their contribution. We thank CS Klinkby, trial nurse, for assistance in relation to the conduct of the trial visits. We also thank L Albjerg (biomedical laboratory technologist), AC Nilsson (consultant), KF Rasmussen (consultant) and J Georgsen (consultant) at the Department of Clinical Immunology, Odense University Hospital, Denmark, for assisting in the implementation of the FMT stool bank.

References

Supplementary materials

Footnotes

  • Handling editor Josef S Smolen

  • Twitter @JustSoren

  • Contributors TE, MSK, RC and JK designed the study. MSK and TE were responsible for funding. MSK, DKH and HMH were responsible for donor recruitment, screening and FMT product manufacture. JK and FMP were responsible for the FMT procedure. MSK, TE, HCH, HLM, JKP, PA and SAJ were responsible for patient recruitment. TE, HCH and HLM acquired the clinical data. MSK, SM and RC analysed the clinical data. TE, MSK, RC, JK, HLM, HCH, KK and MW interpreted the results. MSK, TE and RC drafted the report. All authors critically reviewed the report and approved the final version. MSK and TE are guarantors. The corresponding author (TE) attests that all listed authors meet the authorship criteria and that no others meeting the criteria have been omitted.

  • Funding This study was supported by the Danish Rheumatism Association, the Danish Psoriasis Research Foundation, the University of Southern Denmark Research Fund, the Research Fund of Odense University Hospital, the Danish Regions (Medicinpuljen), the Region of Southern Denmark Research Fund and Novartis Healthcare (unrestricted grant).

  • Competing interests VA declares personal fees from Merck (MSD) and personal fees from Janssen, outside the submitted work. RC declares a core grant to his institution (Parker Institute, Bispebjerg and Frederiksberg Hospital) from the Oak Foundation (OCAY-18-774-OFIL) and honorariums paid to his institution in relation to the following activities: lecture, research methods (Pfizer, DK; 2017); lecture, GRADE lecture (Celgene, DK; 2017); ad board lecture, CAM (Orkla Health, DK; 2017); project grant: 'GreenWhistle' (Mundipharma, 2019); lecture: diet in RMD (Novartis, DK; 2019); consultancy report, Network MA’s (Biogen, DK; 2017); ad board lecture, GRADE (Lilly, DK; 2017); consultancy report, GRADE (Celgene, 2018); and lecture, Network MA’s (LEO; 2020).

  • 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.

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.