In January 2026, researchers publishing the phase 1 PERFORM trial in Nature Medicine reported something clinicians have suspected for nearly a decade but only now can act on: transplanting a healthy person’s gut bacteria into a kidney cancer patient before starting immunotherapy changed how that patient’s tumor responded to treatment. The trial tested healthy donor fecal microbial transplantation combined with immune checkpoint blockade in patients with previously untreated renal cell carcinoma, and found the treatment safe with an encouraging response signal. It’s the latest data point in one of oncology’s stranger and more promising storylines: that a substantial share of who lives and who doesn’t on checkpoint inhibitor therapy may be decided not by the tumor, but by trillions of bacteria in the patient’s colon.
The Scientific Foundation
Immune checkpoint inhibitors like pembrolizumab and nivolumab have transformed treatment for melanoma, lung cancer, and other advanced malignancies by releasing the brakes the immune system normally puts on T-cell activity. The catch is that roughly half of patients don’t respond, and until recently oncologists had no reliable way to predict who would benefit before starting a grueling, expensive course of treatment.
The first major clue came in 2018, when two independent teams — one led by Jennifer Wargo’s group at MD Anderson, the other by Bertrand Routy and Laurence Zitvogel at Gustave Roussy in France — published companion studies in Science. Both found that melanoma and lung cancer patients who responded to anti-PD-1 therapy had measurably more diverse gut microbiomes than non-responders, and that responders’ guts were enriched with specific bacteria, including Faecalibacterium and Akkermansia muciniphila. Patients who had taken antibiotics before starting immunotherapy — which can wipe out these same bacterial populations — fared notably worse.
The field then moved from correlation to causation. In 2021, two teams — Erez Baruch’s group in Israel and Diwakar Davar and Hassane Zarour’s group at the University of Pittsburgh — ran the first human trials testing whether transplanting stool from immunotherapy responders into non-responders could flip their treatment outcome. Both studies, published together in Science, found clinical benefit in a subset of treated patients, including increased abundance of the bacterial taxa associated with response, greater CD8+ T-cell activation, and reduced immunosuppressive myeloid cell activity. It was proof of concept that the microbiome wasn’t just a bystander marker — it was doing something mechanistic to the immune system’s ability to fight tumors.
Since then the results have gotten stronger. A 2023 Canadian phase 1 trial run by Routy’s team combined fecal transplants with frontline immunotherapy in previously untreated melanoma patients and reported an objective response rate of 65 percent, including four complete responses out of twenty patients — well above historical response rates for pembrolizumab or nivolumab alone. Follow-up results from that same MIMic trial, reported in 2025, found an objective response in three of fifteen treatment-naive patients with disease stability in another three, and a median progression-free survival of fourteen months among responders.
The Cross-Domain Connection
What makes this genuinely interdisciplinary is that it sits at the intersection of microbial ecology, immunology, and diagnostics — fields that historically didn’t talk to each other much. Oncologists spent decades optimizing checkpoint inhibitor molecules while largely ignoring the gut as anything other than a place drugs pass through. Microbiologists, meanwhile, were mapping the human microbiome’s role in metabolic and autoimmune disease without much contact with cancer immunotherapy trials.
The synthesis treats the gut microbiome as a diagnostic instrument and a therapeutic lever simultaneously. On the diagnostic side, researchers are now using machine learning models trained on 16S rRNA and shotgun metagenomic sequencing data to predict, before treatment even starts, which patients are likely to respond. One tumor-agnostic meta-analysis across nine cancer types used network analysis showing bacterial taxa are far more interconnected in responders than non-responders, suggesting the whole microbial community structure — not just individual bacteria — carries predictive signal. On the therapeutic side, that same biology becomes an intervention: reshape the microbiome first, through fecal transplant, targeted probiotics, or dietary change, and the immunotherapy that follows works better.
What Remains Undemonstrated
The honest caveat is that this field has a reproducibility problem. Different studies have implicated different bacterial taxa as the “good” responder-associated signature, and a 2020 meta-analysis of melanoma FMT data found a classifier that predicted response only modestly better than chance, with a fairly wide confidence interval. Findings from the newest PERFORM renal cell carcinoma trial suggest the relationship between specific bacteria like Segatella copri and treatment response depends heavily on whether the fecal donor is a stranger or a fellow patient — a wrinkle that complicates any simple “more of bacterium X is better” narrative. No fecal-transplant trial to date has been large enough or randomized against a placebo-transplant arm to prove definitively, rather than suggestively, that changing the microbiome causes better outcomes rather than simply correlating with patients whose immune systems were already going to respond well. Larger randomized trials are underway but results are still years out.
Why It Matters
If the microbiome-response link holds up in larger trials, the clinical payoff is substantial on two fronts. First, a stool sample taken before treatment could let oncologists avoid putting patients through months of toxic, expensive immunotherapy that was never going to work — sparing side effects and redirecting patients toward other options sooner. Second, and more provocatively, it suggests immunotherapy efficacy might be modifiable through something as comparatively low-tech as a fecal transplant or an engineered probiotic, rather than only through ever more expensive next-generation drugs.
The Human Dimension
There’s something almost humbling in the idea that a treatment engineered at extraordinary cost and sophistication — checkpoint inhibitor antibodies represent decades of immunology research — might work or fail partly depending on which bacteria happen to be living in someone’s gut. For patients facing a coin-flip response rate to a treatment that is their best remaining option, that humbling fact is also a hopeful one: it points to a lever, however imperfectly understood, that clinicians might eventually be able to pull.
Sources:
1. Gopalakrishnan et al., “Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients,” Science, 2018 — https://www.science.org/doi/10.1126/science.aan4236
2. Routy et al., “Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors,” Science, 2018 — https://pmc.ncbi.nlm.nih.gov/articles/PMC9295706/ (cited via secondary review)
3. Davar et al., “Fecal microbiota transplant overcomes resistance to anti–PD-1 therapy in melanoma patients,” Science, 2021 — https://www.science.org/doi/10.1126/science.abf3363
4. Baruch et al., “Fecal microbiota transplant promotes response in immunotherapy-refractory melanoma patients,” Science, 2021 — https://www.science.org/doi/10.1126/science.abb5920
5. Routy et al. (MIMic trial), “Fecal microbiota transplantation plus anti-PD-1 immunotherapy in advanced melanoma: a phase I trial,” Nature Medicine, 2023 — https://www.nature.com/articles/s41591-023-02453-x
6. “Improved survival in advanced melanoma patients treated with fecal microbiota transplantation… final results of the MIMic phase 1 trial,” 2025 — https://pmc.ncbi.nlm.nih.gov/articles/PMC12359429/
7. “Fecal microbiota transplantation plus immunotherapy in metastatic renal cell carcinoma: the phase 1 PERFORM trial,” Nature Medicine, January 2026 — https://www.nature.com/articles/s41591-025-04183-8
8. Limeta et al., “Meta-analysis of the gut microbiota in predicting response to cancer immunotherapy in metastatic melanoma,” JCI Insight, 2020 — https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7714408/
Idea originated at artificialideas.org. Article researched and written by Claude Sonnet 4.6. Published at artificialideas.org.
Artificial Ideas 