FMT to reverse immunotherapy resistance: a trial found no tumour responses

Transplanting stool from cancer patients who responded to immunotherapy into patients whose tumours had stopped responding did not shrink a single tumour in a 2026 clinical trial. The MITRIC basket trial gave faecal microbiota transplants plus checkpoint inhibitors to 12 patients with advanced, treatment-resistant cancer and recorded no objective responses, though the approach was safe and five patients had temporarily stable disease [s1].

That result is worth reporting precisely because the idea behind it is genuinely promising and has been over-sold in coverage that leans on the early, smaller studies. The gut microbiome does shape how tumours respond to immunotherapy. Whether deliberately changing it can rescue patients who have already failed treatment is a separate question, and MITRIC is a sobering data point on it.

What the trial did

The microbiome influences immune checkpoint inhibitors (ICIs) — the drugs, such as anti–PD-1 antibodies, that release the brakes on anti-tumour immunity. The hypothesis MITRIC tested is that resistance to ICIs might be reversed by installing the gut community of a patient in whom the same drugs worked.

MITRIC (NCT05286294) was a single-arm phase IIa basket trial: one intervention, no control group, several tumour types enrolled together [s1]. It recruited 12 patients whose cancer was progressing on ICIs — nine with melanoma, and one each with head and neck squamous cell carcinoma, renal cell carcinoma and microsatellite instability-high pancreatic cancer [s1]. The donors were long-term ICI responders [s1]. Patients received two transplants by colonoscopy before their first scan at six weeks, with up to three later transplants allowed by enema, all alongside continued ICI therapy [s1]. The co-primary endpoints were transplant-related adverse events and the objective response rate [s1].

What it found

On safety, the procedure held up: FMT was well tolerated [s1]. Immune-related toxicity — the expected hazard of the checkpoint drugs, not the transplant — occurred in 6 of 12 patients [s1]. All 12 received the first transplant and 10 received the second [s1].

On efficacy, the headline is stark. There were no objective responses [s1]. Five of the 12 patients had stable disease [s1]. Per protocol, clinical benefit — stable disease beyond six months — was achieved in one patient, a melanoma case who had regression of some lesions and remained alive after 33 months with no further systemic treatment [s1]. Two other patients, one with melanoma and one with head and neck cancer, had mixed responses, with some lesions shrinking while others grew [s1]. Median progression-free survival was 1.5 months and median overall survival 10.1 months [s1].

Sequencing confirmed that the donor microbes engrafted in most patients after the first transplant, so the intervention did what it mechanically set out to do [s1]. Blood immune profiling suggested that an activated, differentiated T-cell signature tracked with longer survival, while a naïve T-cell and myeloid-dominant profile tracked with worse outcomes [s1]. Those are correlations in 12 people, not a validated biomarker.

Why this differs from the trials it will be compared to

The excitement around microbiome-plus-immunotherapy rests largely on two 2021 studies in Science. In one, responder-derived FMT plus anti–PD-1 provided clinical benefit in 6 of 15 patients with PD-1-refractory melanoma, reprogramming the tumour microenvironment and, notably, reducing interleukin-8-expressing myeloid cells [s2]. MITRIC's own signal — that a myeloid-dominant environment was unfavourable — echoes that mechanism, but its clinical yield was lower [s1].

The MITRIC authors' conclusion is measured: FMT with ICIs was safe and feasible but had limited clinical activity, and further work is needed to clarify any benefit, identify the right patients and define how to select donors [s1]. A single-arm trial of 12 heterogeneous patients cannot prove a treatment works, and this one did not suggest it does at the response level.

What it means for readers

The gap between mechanism and cure is the recurring lesson of microbiome medicine. That the gut community regulates immune and tumour biology is well supported; that a stool transplant reliably changes a hard clinical outcome is established only for Clostridioides difficile infection, and remains unproven for most other indications. MITRIC extends that pattern into oncology.

For patients, the practical point is that FMT for cancer is investigational. It belongs in trials, with screened donors, not in clinics or at home — donor stool carries real infection risk, which is exactly why formal donor selection is one of the open problems the authors flag [s1]. The more durable route to matching treatments to tumours remains through tested strategies such as genomically guided and vaccine-based approaches now in larger trials.

MITRIC does not close the microbiome-immunotherapy question. It narrows it: engraftment is achievable and safe, but installing a responder's microbiome after resistance has set in was not, on its own, enough to make the drugs work again.

This article describes an investigational cancer treatment tested in a small trial. It is not medical advice, and nothing here should be read as guidance on treatment.

Sources

1. [s1] Safety and efficacy of fecal microbiota transplantation in solid cancers resistant to immune checkpoint inhibitors: results of the MITRIC trial. Journal for ImmunoTherapy of Cancer, published online July 20, 2026. https://doi.org/10.1136/jitc-2026-015122
2. [s2] Fecal microbiota transplant overcomes resistance to anti–PD-1 therapy in melanoma patients. Science, published online February 4, 2021. https://doi.org/10.1126/science.abf3363