Immune-checkpoint inhibitors transformed oncology, but they only work for some patients — and one of the strongest predictors of who responds turns out to be the gut microbiome. The obvious question: if you change the microbiome, can you change the odds?

A 2023 phase I trial (published in Nature Medicine) was the first to test healthy-donor fecal microbiota transplantation (FMT) alongside anti–PD-1 therapy as first-line treatment in 20 previously untreated advanced-melanoma patients. The objective response rate was 65% (13 of 20), including four complete responses.

In January 2026, the phase 2 FMT-LUMINate trial extended the idea to two cancers and two immunotherapy backbones. In non-small-cell lung cancer (n=20, anti–PD-1), the response rate was 80%; in melanoma (n=20, dual anti–PD-1 + anti–CTLA-4), it was 75% — both above the historical rates for immunotherapy alone.

The counterintuitive part

You’d expect the mechanism to be “transplant good bacteria, which help the immune system.” The data pointed elsewhere. Patients who responded showed greater loss of their own pre-existing gut species, while the degree to which donor strains engrafted didn’t separate responders from non-responders.

The benefit may come less from the microbes you add than from the resident ones you clear out of the way.

The honest caveats: these are small, single-arm, early-phase studies — 20 patients per group. They establish a strong, repeatable signal, not proof of efficacy, and they need randomised confirmation. But the direction is consistent across cancers and years.

Sources

  1. Routy et al., phase 2 FMT-LUMINate trial, Nature Medicine (Jan 28, 2026)
  2. First-line FMT + anti–PD-1 in melanoma, phase I, Nature Medicine (2023)
Editorial content for general information only — not medical advice, diagnosis or treatment. Talk to a qualified clinician about your own health.