Engineered gut bacteria may replace fecal transplants

Microbiome therapy may be moving away from stool transfer and toward something more prosaic: precisely manufactured bacterial mixes.

In Nature Medicine, Bethlehem et al. tested a 15-strain live biotherapeutic product called MTC01 against same-donor fecal microbiota transplant in people with recurrent Clostridioides difficile infection. Its main finding was small but unusually direct. In the phase 1b randomized trial, the defined bacterial product met its primary safety endpoint and showed similar short-term engraftment to fecal transplant.

Read the caveat slowly. Only 18 patients were randomized. Follow-up for recurrence was eight weeks. A replacement claim would outrun the evidence. Still, the trial shows something narrower, and potentially more important for the field: a defined bacterial consortium can be made under a controlled manufacturing process and still behave enough like donor stool to deserve a larger test.

This is why fecal microbiota transplant, or FMT, remains both impressive and awkward. It can work dramatically well for recurrent C. difficile. By design, though, it is a treatment built from another person’s stool. Donor screening, batch variability, infectious risk, patient reluctance and regulatory oversight all sit inside the same package.

Jeremiah J. Faith, a co-senior and co-corresponding author at Mount Sinai, framed the problem as one of translation rather than biology.

“Microbiome therapies have shown enormous potential, but manufacturing challenges have slowed broader clinical use.”
Jeremiah J. Faith, Mount Sinai

Read that way, this is a manufacturing paper with clinical data attached. That may sound less exciting than a cure story. It is probably the more honest version.

How the trial changed the microbiome question

Clinically, recurrent C. difficile infection is one of the few places where microbiome therapy has a clear anchor. Antibiotics can suppress the infection, but they can also leave the gut community damaged enough for recurrence. FMT tries to solve that by restoring a more resilient microbial community.

The Mount Sinai investigators screened 20 patients and randomized 18 into four arms: low-dose FMT, high-dose FMT, low-dose MTC01 and high-dose MTC01. Each MTC01 product contained the same 15 bacterial strains derived from the donor material used for the matched FMT arm. Such a design makes the comparison useful. Rather than asking whether some generic probiotic blend could help, it asked whether a defined, manufactured version of donor-sourced bacterial strains could reproduce key features of stool transfer.

Lukas Bethlehem and colleagues reported no treatment-related adverse events across all four groups. Across MTC01 and FMT recipients, 10 adverse events occurred in eight patients. For an early efficacy signal, seven of nine patients who received the live biotherapeutic product were free of recurrent C. difficile at eight weeks, compared with eight of nine in the FMT group.

Encouraging, yes. Definitive, no. Shift one patient from success to recurrence and the apparent gap between groups changes. Safety, feasibility and engraftment came first in this study, not a final efficacy verdict.

Fecal transplant remains the benchmark

FMT earned its place because conventional treatment often fails people with recurrent disease. Van Nood et al., in a 2013 New England Journal of Medicine trial indexed on PubMed, found donor feces infusion significantly more effective than vancomycin-based therapy for recurrent C. difficile. That paper helped turn FMT from an odd procedure into a serious therapeutic model.

Ari Grinspan, another co-senior author of the 2026 trial, is not arguing that stool transfer never worked. His point is that a therapy can be clinically powerful and still difficult to standardize.

“Our goal was to move beyond stool-based therapies toward something more precise and reproducible.”
Ari Grinspan, Mount Sinai

Precision is carrying a lot of weight in that sentence. A donor stool sample contains a living microbial community, not a neatly labeled drug product. Composition can vary from donor to donor and from sample to sample. Some of that complexity may be part of why FMT works. Some of it may be noise. Some of it may be risk.

Defined consortia try to keep the useful parts while stripping away the unknowns. MTC01 is not a single strain sold as a wellness probiotic. It is a mixture of 15 selected bacterial strains produced under a process meant to be documented, repeated and inspected. Drug Discovery News, in an analysis of microbiome manufacturing, described the broader challenge plainly: microbiome therapeutics are becoming a manufacturing problem as much as a discovery problem.

Manufacturing may be the main result

Regulators tend to prefer therapies whose identity can be defined. Clinicians tend to prefer products with predictable handling. Many patients would probably prefer a therapy that does not require the psychological hurdle of donor stool. A manufactured live biotherapeutic product potentially speaks to all three audiences.

Mount Sinai’s team emphasized that it had produced regulatory documentation and manufacturing protocols alongside the trial. That part of the paper could matter beyond MTC01 itself. A field built on bespoke academic preparations cannot easily scale into routine medicine. One built on reproducible strain banks, quality-control testing and dosing protocols has a clearer path.

Grinspan made the quality-control case directly.

“By defining exactly which bacterial strains are included, we can better understand how these therapies work and potentially improve safety, quality control, and scalability.”
Ari Grinspan, Mount Sinai

None of that means defined products will automatically outperform stool-based FMT. They may not. The gut microbiome is not a pill bottle with 15 missing ingredients waiting to be replaced. It is a dense, adaptive microbial network shaped by diet, antibiotics, immune activity, bile acids, pathogens and host genetics. A narrower product may be easier to manufacture precisely because it leaves out some biology.

Reproducibility still has scientific value. If a defined consortium succeeds or fails in a larger trial, investigators can ask which strains engrafted, which metabolites changed and which patients responded. With stool transfer, the causal map is harder to draw.

What patients should not conclude yet

Patients with recurrent C. difficile should not read this trial as a reason to delay established care or seek unregulated microbiome products. This study used a specific live biotherapeutic product in a clinical-trial setting, not an over-the-counter probiotic and not a supplement protocol.

That distinction is especially important because “gut health” has become a marketing language of its own. A 15-strain investigational product for recurrent C. difficile is not evidence that general-purpose probiotics can reset the microbiome, improve immunity or treat unrelated digestive complaints. Here, the claim is narrower and stronger because the disease, endpoint and product were clearly defined.

CNN’s recent reporting on the gut microbiome and cancer care shows why the wider medical interest is growing. Researchers are testing whether microbial communities shape immune responses, drug metabolism and treatment outcomes across conditions. Real as that context is, it can also tempt readers to blur early signals from one disease area into general microbiome optimism.

The Mount Sinai trial does not justify that leap. It suggests that engineered bacterial consortia may become more testable than stool transfer, and therefore easier to improve or reject. That is useful precisely because it is disciplined.

Where microbiome medicine is heading

Near term, FMT is unlikely to disappear suddenly. Larger trials will ask whether defined live biotherapeutics can match or exceed stool-based treatment while offering better consistency. Useful data would include more patients, longer follow-up, recurrence rates by dose, strain-level engraftment, adverse-event detail and comparisons against existing approved microbiome products where appropriate.

Manufacturing will shape those trials before the first patient is dosed. Can the same strains be grown reliably? Do they remain stable? Can a capsule or other oral formulation deliver them to the right part of the gut? Mount Sinai’s release says investigators have developed an oral formulation for upcoming studies, which would make the therapy easier to test and potentially easier to use than procedures requiring direct intestinal delivery.

Bethlehem et al. have not solved microbiome medicine. They have made one of its central problems more concrete. Instead of asking whether “the microbiome” can be repaired, the field can ask whether specific, named microbes, grown under defined conditions, can safely reconstitute enough function to prevent recurrence in a clearly diagnosed disease.

That is a smaller promise than the wellness industry usually sells. It is also the one more likely to survive contact with clinical trials.

Bottom line

For now, engineered gut bacteria are not replacing fecal transplants. In an 18-patient phase 1b trial, a 15-strain live biotherapeutic product looked safe in the short term and produced an early recurrence-prevention signal similar to same-donor FMT.

The more important result is logistical. If microbiome therapy can be defined, manufactured and quality-controlled without losing its clinical effect, it may move from a messy donor-based procedure toward a drug-like platform. Larger trials will decide whether that promise is real.

References

1. Bethlehem L, Bartu L, Marke G, et al. 15-strain live biotherapeutic product or same donor fecal microbiota transplant for recurrent Clostridioides difficile infection: a randomized phase 1b trial. Nature Medicine. 2026. https://www.nature.com/articles/s41591-026-04442-2
2. van Nood E, Vrieze A, Nieuwdorp M, et al. Duodenal infusion of donor feces for recurrent Clostridium difficile. New England Journal of Medicine. 2013. https://pubmed.ncbi.nlm.nih.gov/23323867/