Faecal microbiota transplantation (FMT) shows no evidence of clinical benefit in the treatment of irritable bowel syndrome (IBS) in the phase III REFIT2 trial.
At 90 days after treatment, the primary endpoint of a ≥75-point reduction in the IBS-Severity Scoring System (IBS-SSS) score occurred in 40 percent of participants who underwent FMT using stool samples from healthy donors (intervention) vs 38 percent of those who underwent FMT using their own stool (placebo; adjusted risk ratio [aRR], 1.09, 95 percent confidence interval [CI], 0.79–1.50; p=0.76). [Lancet 2026;doi:10.1016/S0140-6736(26)01424-8]
Additionally, there was no significant between-group difference in the proportion of participants who had a ≥14 point-improvement in the IBS-QoL score at 90 days (difference, 2.8 percentage points), had adequate relief according to the combined Global Improvement Scale (score 1 or 2) and abdominal pain measure (≥30-percent decrease) (difference, 1.7 percentage points) on at least two of four timepoints in the period between 69 days and 90 days after treatment, or had a ≥75-point reduction in IBS-SSS score at 1 year after treatment (difference, –1.5 percentage points).
The frequency of adverse events (AEs) was similar between the treatment groups. Twenty-four AEs, including six serious AEs, occurred in the intervention group over 1 year of follow. In the placebo group, eight AEs were reported, and none were serious. Four participants in the intervention group and three in the placebo group had AEs that were possibly or probably related to the study treatment. There were no deaths reported during the study.
“The results of the REFIT2 trial do not support the use of FMT in patients with IBS,” said Dr Peter Holger Johnsen from the University Hospital of North Norway, Harstad, Norway, and colleagues.
“The absence of clinical effects despite donor-dependent microbial engraftment, confirmed by directional shifts of recipient microbiota towards assigned donor profiles, suggests that altering global gut microbiota composition through FMT is not sufficient for symptom relief in IBS,” added Johnsen and colleagues.
What doesn’t work, what is needed
In an accompanying editorial, Drs Jacqueline Liua and Andrea Shin from David Geffen School of Medicine at the University of California Los Angeles, Los Angeles, California, US, discussed how the REFIT2 trial, despite its negative findings, provides important insights that could guide the next stage of microbiome-directed therapies for IBS. [Lancet 2026;doi:10.1016/S0140-6736(26)01815-5]
“Importantly, the lack of efficacy with healthy donor-derived FMT might reflect the limitations of nonselective microbiota therapy, in which donor selection is based on clinical health rather than defined microbial features,” Liua and Shin wrote.
“Rather than implying that the gut microbiome has only a minor role in IBS pathogenesis, the findings [from REFIT 2] further underscore the need for targeted microbiota modulation based on disease-relevant microbial characteristics,” they continued.
Liua and Shin also raised the uncertainty of whether administering FMT as a single rectal enema in REFIT2 achieves microbial delivery similar to other methods such as administration via colonoscopy. Moreover, REFIT2 used stool samples from three distinct donors, possibly introducing biological heterogeneity within the intervention, they added.
Moving forward, “improved characterization of donor–recipient compatibility, optimized donor–recipient pairing, standardized methods for confirming microbial engraftment, incorporation of dietary co-interventions, harmonized microbiota delivery protocols, and longitudinal symptom assessments” may be needed to advance the future development of microbiome therapeutics in IBS, Liua and Shin concluded.
REFIT2
The trial included 448 adult patients (median age 36 years, 65 percent female) with moderate-to-severe IBS. Participants aged ≥50 years had to have undergone a colonoscopy within 5 years before enrolment to rule out colorectal cancer, and those with IBS with predominantly diarrhoea were required negative biopsies to exclude microscopic colitis.
The participants were randomly allocated to the intervention group (n=298) or the placebo group (n=150). Treatments were administered as a once-only rectal enema.