Specific gut bacteria predict who responds to melanoma immunotherapy

Whether a melanoma patient responds to immunotherapy may depend, in part, on which bacteria are living in their gut. A new meta-analysis published in Communications Medicine by researchers at The George Washington University and Weill Cornell Medicine pooled data from 678 patients across seven cohorts to identify the microbial signatures most consistently linked to treatment success, a question that has been frustratingly hard to answer until now.

Faecalibacterium bacteria tied to immunotherapy response in melanoma

Patients who responded well to immune checkpoint inhibitors (ICIs), drugs that release the brakes on the immune system so it can attack tumors, were significantly more likely to carry a specific gut bacterium called Faecalibacterium SGB15346. That finding aligns with earlier work showing that gut bacteria influence immunotherapy response, but the GWU-Cornell team went further than most previous efforts by layering in two additional types of molecular data alongside species counts.

Along with the Faecalibacterium signal, responders showed enrichment in RiPP biosynthetic gene clusters and in metabolic pathways that produce short-chain fatty acids through fermentation. Short-chain fatty acids are metabolites that feed colon cells and help regulate immune activity, so their presence in responders fits a plausible biological story. Non-responders, by contrast, showed elevated dTDP-sugar biosynthesis, a biochemical pathway that the authors flag as a potential marker of treatment resistance.

"Our findings show that the gut microbiome may play an important role in how patients respond to immunotherapy," said Ali Rahnavard, one of the study's senior authors.

Why melanoma patients and oncologists should care

Immunotherapy has transformed treatment for advanced melanoma, but it still fails a substantial share of patients, and no reliable pre-treatment test exists to predict who will benefit. A validated microbiome signature could eventually guide treatment decisions or identify patients who might gain from microbiome-modifying interventions before they start therapy.

The practical angles are real. A Phase 1 trial published in Nature Medicine showed that fecal microbiota transplantation from ICI responders could re-sensitize some melanoma patients who had previously stopped responding to anti-PD-1 therapy, producing objective responses and favorable microbiome shifts. That proof-of-concept suggests the connection between gut flora and immunotherapy is not merely correlational. Diet is another lever: research published in Science found that high-fiber diets improved checkpoint blockade responses in melanoma patients, while over-the-counter probiotic use was associated with worse survival, a reminder that not all microbiome interventions are beneficial.

How the meta-analysis pooled seven cohorts into one signal

One of the persistent problems in microbiome-ICI research has been inconsistency: studies use different sequencing methods, enroll patients from different geographic regions with different dietary habits, and collect samples at different points in the treatment cycle. Individual cohorts have often contradicted each other, making it difficult to identify reproducible targets.

The GWU-Cornell team addressed this by reanalyzing stool metagenomic sequencing data from 15 melanoma cohorts covering 763 samples from 484 individuals, then applying a standardized computational pipeline across all of them. Crucially, they integrated three layers of information simultaneously: microbial species identities, biosynthetic gene clusters (BGCs), and functional metabolic pathways. This is the first microbiome meta-analysis for ICI response to incorporate BGC data, which tracks the genetic machinery bacteria use to produce bioactive compounds. The integration allowed the team to move beyond asking which species are present and begin asking what those species are actually doing chemically.

Open questions about reproducibility and microbiome causality

The study's own conclusions are appropriately cautious. The authors note that no single universal species captures the full microbiome signature of response; the relevant microbial profile appears to depend on treatment context, meaning the bacteria that predict response to one checkpoint inhibitor may not predict response to another. That context-dependence complicates efforts to build a one-size-fits-all diagnostic.

The field also carries a broader reproducibility burden. A state-of-the-art review of clinical microbiome studies found substantial methodological heterogeneity across cancer immunotherapy research, including inconsistent patient selection, variable sequencing platforms, and the likelihood that negative studies go underreported. The GWU-Cornell work is a step toward standardization, but the authors acknowledge that mechanistic studies, experiments that establish why specific microbes influence immune responses rather than just documenting that they do, remain necessary before microbiome data can inform clinical decisions.

Dietary and microbiome interventions as next steps

The most actionable near-term direction the authors point to is microbiome modification before or during immunotherapy: dietary changes that favor fiber-fermenting bacteria like Faecalibacterium, targeted probiotic formulations (designed with more specificity than over-the-counter blends), or FMT protocols. None of these are ready for routine clinical use, and the researchers call for randomized trials to test whether shifting the gut microbiome actually changes ICI outcomes rather than merely correlating with them.

What this meta-analysis contributes is a more precise target list. By identifying Faecalibacterium SGB15346, the RiPP biosynthetic class, and short-chain fatty acid fermentation pathways as consistent features of responders, it gives future intervention studies something concrete to aim at. For the roughly half of advanced melanoma patients who do not respond to first-line immunotherapy, that precision could eventually matter quite a lot.