Randomised Clinical Trial

A 12-Strain Bacterial Mixture Versus Faecal Microbiota Transplantation Versus Vancomycin for Recurrent Clostridioides difficile Infections

Anne Abildtrup Rode; Mahtab Chehri; Laura Rindom Krogsgaard; Kristine Klysner Heno; Anna Tølbøll Svendsen; Iben Ribberholt; Morten Helms; Jørgen Engberg; Kristian Schønning; Michael Tvede; Christian Østergaard Andersen; Ulrich Stab Jensen; Andreas Munk Petersen; Peter Bytzer


Aliment Pharmacol Ther. 2021;53(9):999-1009. 

In This Article



We recruited participants from May 2017 to December 2018 at Zealand University Hospital and from June 2017 to March 2019 at Hvidovre University Hospital. We assessed 1020 individuals with a positive C difficile test within 90 days of an earlier positive test (Figure 1).

Figure 1.

Flow diagram according to CONSORT 2010.40 RBT, rectal bacteriotherapy; FMT, faecal microbiota transplantation; ITT, intention-to-treat analysis, including all randomised participants; mITT, modified ITT, excluding participants not receiving any of the allocated intervention; PP, per protocol, excluding participants not receiving all three rectal bacteriotherapy instillations, not receiving a repeat faecal microbiota transplantation if indicated, not receiving 7–14 days of vancomycin prior to rectal bacteriotherapy/faecal microbiota transplantation and not discontinuing vancomycin at least 8 h prior to rectal bacteriotherapy or at least 24 h before faecal microbiota transplantation

As planned, we conducted an interim analysis for the first 90 participants. When the reported results were apparent, including the mortality data, the study was terminated due to futility and ethical concerns—even though the Haybittle-Peto boundary was not met. Participants enrolled between randomisation of participant 90 and the time of the interim analysis (90 days follow-up) were included in the final analyses, yielding a study population of 98 participants.


The 98 participants (intention-to-treat) were equally allocated to the three treatments. Two participants in the rectal bacteriotherapy group did not receive any infusions, one because of concomitant non-C difficile-related disease and the other withdrew consent. Thus, 96 participants received the allocated intervention (or at least one infusion of rectal bacteriotherapy/FMT) and were followed until 180 days after treatment or death (modified intention-to-treat) (Figure 1).

Baseline characteristics for the participants are shown in Table 3. There were no significant or clinically important differences between the groups. Of note, approximately 60% of the participants in each group were included at their first recurrence, while the remaining had multiple (≥2) recurrences.

Ten different donors were used. They were mostly women (60%), aged between 21 and 63 years (median 41 years).

Primary Outcome

Clinical cure, defined as absence of C difficile infection within 90 days after treatment, was observed in 14/31 (45%) participants receiving vancomycin, in 16/31 (52%) participants receiving rectal bacteriotherapy and in 26/34 (76%) participants receiving 1–3 FMTs (modified intention-to-treat) (Figure 2). FMT was more effective than both vancomycin (OR 3.9 [1.4–11.4], P < 0.01) and rectal bacteriotherapy (OR 3.0 [1.1–8.8], P = 0.04) with numbers needed to treat (NNT) of 3 and 4, respectively. There was no difference in effect between rectal bacteriotherapy and vancomycin (OR 1.3 (0.5–3.5), P = 0.61). Of note, the duration of pre-treatment with vancomycin was similar between the FMT and rectal bacteriotherapy groups (median days 9.5 [IQR 8–12] vs 10.0 [IQR 9–12.5], P = 0.31). Furthermore, results were unchanged after adjusting for the number of recurrences (FMT vs vancomycin: P = 0.01; FMT vs rectal bacteriotherapy: P = 0.03; rectal bacteriotherapy vs vancomycin: P = 0.61).

Figure 2.

Clinical cure rate at day 90 in each treatment group (in proportions). *Chi-squared test. FMT, faecal microbiota transplantation; RBT, rectal bacteriotherapy; NNT, number needed to treat

In the FMT group, 10/34 participants had clinical treatment failure within the first 14 days fulfilling the criteria for repeating the transplantation. Seven received one additional infusion, and one had two additional infusions, all with a different donor. Two participants did not receive a repeat FMT within the predefined 14 days for logistic reasons. The median time between the first and second FMT was eight days (range: 6–14 days).

The cure rate after a single FMT was 56% (19/34), rising to 74% for 1–2 infusions (25/34). A single FMT was thus not superior in efficacy to either vancomycin (OR 1.5 [0.6–4.1]) or rectal bacteriotherapy (OR 1.2 [0.4–3.2]), while 1–2 FMTs (composite effect) were more effective than vancomycin (OR 3.4 [1.2–9.5], NNT = 4) and tended to be better than rectal bacteriotherapy (OR 2.6 [0.9–7.4]).

We observed no clinical treatment failures in the rectal bacteriotherapy group (defined as ongoing, worsened or new-onset diarrhoea before the last of three infusions), and thus, all failures in this group were classified as recurrences with microbiological verification. In the vancomycin group, three participants experienced worsened diarrhoea during tapering of vancomycin. We treated these participants as clinical treatment failures without testing for C difficile, increased the dose of vancomycin again and planned FMT as a rescue-treatment.

The intention-to-treat analysis included the two participants excluded from modified intention-to-treat analysis since they did not receive any infusions of rectal bacteriotherapy. These two participants were considered not cured, which left only 16 (48%) in the rectal bacteriotherapy group as cured. Thus, intention-to-treat analysis confirmed that FMT performed better than rectal bacteriotherapy (P = 0.02). In per-protocol analysis, FMT was also superior in efficacy to both other treatments (Table S2).

Subgroup Analyses

The cure rate substantially differed between the two strata in the FMT group with a cure rate of 90% for participants with their first recurrence, yet only 57% for participants with multiple (ie two or more) recurrences. In an exploratory post hoc analysis, we found that also participants with only their second recurrence of C difficile infection benefitted from FMT (cure rate of 70%). The very limited number of participants with three or more recurrences had a more disappointing response to FMT (cure rate 25% (1/4), Figure 3). Notably, there were no differences in effect across strata or number of recurrences for either rectal bacteriotherapy or vancomycin (Figure 3 and Figure S1), despite participants with multiple recurrences receiving tapering of vancomycin in the vancomycin group.

Figure 3.

Cure rates after 90 days follow-up in pre-specified subgroups. A, Cure rates according to strata—first recurrences vs two or more recurrences. B, Cure rates after faecal microbiota transplantation at 90 days follow-up stratified according to the number of recurrences (post hoc analysis). C, Cure rates after 90 days follow-up according to the toxin profile of the Clostridioides difficile strain. Toxin B: Infection with C difficile strains with the toxin B-gene; Binary toxin: Infection with C difficile strains containing both the toxin B-gene and the binary toxin gene; CD027: Infection with C difficile strains of presumptive CD027 with the characteristic Δ117-deletion in the tcdC-gene and with both the toxin B-gene and the binary toxin gene. FMT, faecal microbiota transplantation; RBT, rectal bacteriotherapy

The presence of the binary toxin gene did not seem to affect the cure rates compared with C difficile strains carrying only the toxin B-gene. Only six participants (6%) were infected with CD027. Among these, only two (33%) were cured by the intervention (Figure 3).


Thirteen participants (14%) died during the 180 days of follow-up with a predominance in the vancomycin group (Figure 4). The all-cause mortality rate was 79% lower in the FMT group than in the vancomycin group with a NNT of 6 (OR 0.21 [0.04–1.12], P = 0.07). The rates of possibly C difficile-related mortality were 0% in the FMT group, 6% (n = 2) in the rectal bacteriotherapy group and 13% (n = 4) in the vancomycin group, in which one person died from toxic megacolon after a new recurrence on day 101 after ended treatment.

Figure 4.

Mortality rates: Proportions of deaths during follow-up (180 days) in each group. *Fisher's exact test. FMT, faecal microbiota transplantation; RBT, rectal bacteriotherapy; NNT, number needed to treat


Gastrointestinal symptoms were common (82%) during the procedure and the 1-hour observation period following FMT and to less extent during and following rectal bacteriotherapy (35%). Typical symptoms were defecation urge, incontinence and abdominal pain or discomfort. Mild self-limiting gastrointestinal complaints were also frequent in the first days and weeks after both FMT and rectal bacteriotherapy (in 73% and 84% within 48 hours and in 32% and 55% after 48 hours). See Table S3 for a full list of adverse events.

Ten participants (10%) had serious adverse events (Table S3) within the first 14 days after treatment, including nine hospital admissions and one death. This death in the vancomycin group was unrelated to either treatment or C difficile infection. None of the hospital admissions was related to either FMT or rectal bacteriotherapy, except a potentially related case of suspected urosepsis in the rectal bacteriotherapy group.

Timing of Recurrences of C difficile Infection

The majority of recurrences occurred in the first 30 days of follow-up in all groups, yet a substantial part of the failures in the vancomycin group occurred with a delay before the risk stabilised after approximately 60 days (Figure S2).

Since other randomised controlled trials on this subject evaluated efficacy at 8 weeks[29] or 10 weeks,[7,8] we conducted a post hoc analysis of the clinical cure rates at these time-points. At 8 weeks, the cure rates were 79% (27/34) in the FMT group, 55% (17/31) in the rectal bacteriotherapy group and 48% (15/31) in the vancomycin group. At 10 weeks, they were 76% (26/34), 55% (17/31) and 45% (14/31), respectively.

Only two participants experienced late recurrences after 90 days. The cure rates at 180 days after ended treatment were 74% for FMT, 52% for rectal bacteriotherapy and 42% for vancomycin.