Fecal transplant did no better than placebo for irritable bowel syndrome in a 450-person trial

The idea was attractive and for a while the evidence seemed to back it. People with irritable bowel syndrome have a different mix of gut bacteria from healthy people. Replace the bacteria with a healthy person’s, the reasoning went, and the symptoms might ease. Small trials in Norway reported exactly that.

The researchers behind the first of those trials have now run the large trial that the small ones called for, and published it in The Lancet in September. It found no benefit. People given a transplant of a healthy donor’s stool did no better than people given back their own.

Why fecal transplant was tried for irritable bowel syndrome

Irritable bowel syndrome brings recurring abdominal pain with diarrhea, constipation or both, and no visible damage to the bowel. Treatments help some people and leave many dissatisfied. The trial’s registry entry puts it bluntly: many patients with irritable bowel syndrome (IBS) do not experience adequate symptom relief with current treatments. The causes are disputed. In the same entry’s words, the pathophysiology of IBS is diverse, controversial and not completely understood, pathophysiology meaning the way a disease disturbs the body’s workings.

Gut bacteria became a leading suspect. The new paper begins from the view that gut microbiota composition might have a role in the pathogenesis of irritable bowel syndrome (IBS), the microbiota being the community of microbes in the bowel and pathogenesis meaning how a disease arises.

A fecal transplant, known formally as fecal microbiota transplantation or FMT, is the most direct way to change that community. Stool from a screened healthy donor is placed in the patient’s bowel. The technique works well against one bowel infection that keeps coming back, caused by a bacterium called C. difficile, and the hope was that it would carry over.

The first signal came from the group behind the new trial. In 2015 they recruited 90 participants and randomly assigned them to active treatment or placebo, with the transplant administered by a colonoscope deep into the large bowel. In the paper’s count, 36 (65%) of 55 participants receiving active treatment versus 12 (43%) of 28 receiving the placebo showed response at 3 months. The result only just cleared the usual threshold for ruling out chance.

A second Norwegian team then reported something far more dramatic. Using stool from a single donor, given through a tube into the upper gut, they tested two doses against placebo. Responses occurred in 23.6%, 76.9% and 89.1% of the patients who received placebo, 30 g FMT and 60 g FMT, respectively.

Reviewers who pooled all the trials were not convinced. One meta-analysis in 2019 concluded that there were no differences between FMT and control in improvement. Another appeared this May. Thirteen randomized trials involving 693 patients were eligible for this review, and its authors judged that there was very low certainty of evidence that FMT improved IBS symptoms. The American Gastroenterological Association (AGA) took the cautious line in its 2024 guideline. It suggests against the use of conventional fecal microbiota transplant as treatment for inflammatory bowel diseases or irritable bowel syndrome, except in the context of clinical trials.

How the fecal transplant trial was run

The new study was designed to settle the matter. Its registry entry describes the earlier pilot as the first indication of a possible benefit from treating IBS with fecal microbiota transplantation (FMT) and states the ambition plainly: this study is the first phase 3 trial of FMT for IBS worldwide. A phase 3 trial is the large, confirmatory kind. The hypothesis of the trial is that donor FMT is more effective than placebo FMT in treating IBS.

It was conducted at five hospitals in Norway, led by the University Hospital of North Norway. The researchers assessed 2,304 people for eligibility, and 450 participants were enrolled and randomly assigned to the intervention group (299) or placebo group (151). Participants were men and women aged 18-65 years with moderate-to-severe IBS. Two were left out of the results: two randomly assigned participants were excluded from analyses because their allocated treatments were switched. That left 448. About two-thirds were women, and the median age was 36 years, the median being the middle value.

Entry was guarded against other diagnoses. Recruitment ran from May 2021 to July 2022. Participants aged 50 or older needed a colonoscopy within the previous five years to rule out colorectal cancer, and those whose main symptom was diarrhea needed biopsies to exclude an inflammatory bowel condition that can pass for the syndrome.

Everyone had the same procedure. Treatments were delivered as a once-only rectal enema. What differed was the contents: stool from either healthy donors (intervention) or the participants themselves (placebo group). Using a person’s own stool as the placebo means that both groups experience an identical treatment, down to its look and smell, while only one receives new bacteria.

Blinding was thorough. All investigators, participants, and study personnel involved in treatment administration, patient care, and outcome assessment were masked to treatment allocation throughout the study.

Success was defined in advance as a drop of 75 points or more, 90 days after treatment, on a standard questionnaire of symptom severity. People had to score at least 175 points to enter.

At 450 participants, this one trial is about two-thirds the size of the 13 trials in that review combined.

Donor stool and placebo gave the same result

	Donor stool	Own stool (placebo)
Improved by 75 points or more at 90 days	119 (40%)	57 (38%)

In the donor group, 119 people, or 40%, met the target. The authors report that 57 (38%) in the placebo group showed an improvement of at least 75 points.

Before rounding, the difference was 1.9 percentage points, with a plausible range from 8.1 points in favor of placebo to 12.0 points in favor of donor stool. The range sits astride no effect.

The authors do not hedge. The trial did not show a clinical benefit of the transplant, they write.

Two features of the numbers are worth noticing. The first is how many people improved on placebo. Nearly four in ten felt meaningfully better after an enema of their own stool. Large placebo responses are typical of this condition, and they are the reason small trials without tight blinding can mislead.

The second is the pilot. A response rate of 65% with donor stool, seen in 55 people, became 40% in about 300. The placebo rate barely moved, from 43% to 38%.

Safety was unremarkable. To collect side effects, all participants were advised to report any adverse event during follow-up to their study contact, preferably by telephone. The proportion of participants with adverse events was similar in the two study groups.

Why earlier trials of fecal transplant looked better

There are two kinds of explanation, and the trial cannot choose between them.

One is that the earlier results were flukes or were inflated by the hazards of small trials. The pilot analyzed 83 people. With groups that small, a handful of responders either way changes the answer.

The other is that the method matters. The new trial used an enema, which reaches only the lower bowel. The pilot placed the transplant by colonoscope at the far end of the large bowel. The second Norwegian trial took material from one donor, frozen and administered via gastroscope into the small intestine, and its authors emphasized the choice of donor. The registry record for the new trial lists three donors.

The latest meta-analysis found hints that method might matter. Benefits appeared in some subgroups, including trials using a single dose, and its authors warned that this needs to be interpreted with caution. The 2019 review made a similar point: variations in FMT methods and patient factors may contribute to the heterogeneous results of the trials.

The authors of the new trial offer a broader reading. Their conclusion is that microbiota modulation alone might be insufficient for symptom improvement in IBS.

What the trial does not rule out for irritable bowel syndrome

Other routes and doses. One enema is the simplest version of the treatment. The trial does not test delivery higher in the gut, repeated doses or capsules.

Particular donors. If some donors’ stool works and most does not, a trial using a few donors could miss it. That idea remains unproven and would need its own large trial.

Subgroups. People were enrolled whatever their main symptom. A benefit confined to one type of the condition would be hard to see in the overall result.

Longer follow-up. The main result is at 90 days.

One country. All five hospitals were in Norway.

What the trial does establish is narrower and firm. A single transplant by enema, in a large and well-blinded study, did not improve symptoms. Funding came from a Norwegian program called KLINBEFORSK, and a patient representative was involved in the planning of the trial.

The choice of treatment for irritable bowel syndrome depends on a person’s symptoms and what they have already tried, which is assessed by their doctor.

In the first phase 3 trial of fecal transplant for irritable bowel syndrome, 40% of people given donor stool by enema and 38% of those given their own improved by the target amount at 90 days, a result that shows no benefit from that method while leaving other donors, doses and routes untested.