Immune checkpoint inhibitors (ICIs) have improved outcomes in NSCLC and melanoma, yet many patients still show primary resistance. The phase 2, multicenter, open‑label FMT‑LUMINate trial evaluated whether a single oral FMT from a healthy donor, administered before ICI initiation, could enhance first‑line treatment efficacy. 

Clinical activity

NSCLC cohort (n=20): The primary endpoint—objective response rate (ORR)—was 80% (16/20), exceeding the prespecified threshold for a positive study outcome.

Melanoma cohort (n=20; dual ICI: nivolumab + ipilimumab): ORR was 75% (15/20).

Safety

Safety outcomes differed by regimen. 

In NSCLC, no grade ≥3 adverse events were reported. 

In melanoma (dual ICI), 65% (13/20) experienced grade ≥3 toxicities, including myocarditis in 15% (3/20). The authors also observed associations between severe immune‑related events and engraftment of donor‑specific Prevotella species in the dual‑ICI setting. 

Mechanistic insights
A key finding was that clinical response was linked less to donor‑strain engraftment and more to selective loss of baseline gut microbes. Shotgun metagenomic sequencing showed that responders developed a distinct post‑FMT microbiome characterized by a significantly larger depletion of baseline taxa—commonly including Enterocloster citroniae, Enterocloster lavalensis, and Clostridium innocuum—regardless of donor‑recipient similarity or strain‑level engraftment. 

Mouse experiments reinforced this mechanism: transferring post‑FMT responder stool enhanced ICI efficacy, while reintroducing baseline species that had been depleted after FMT diminished the anti‑tumor effect. This supports the hypothesis that eliminating specific deleterious taxa may be necessary for optimal therapeutic benefit. 

Importance of robust microbiome measurement
Such mechanistic insights require microbiome profiling methods that are reproducible across timepoints and cohorts. Relative‑abundance sequencing is prone to compositionality effects, which can lead to false‑positive signals. To mitigate this, the authors complemented metagenomics with a targeted qPCR assay—Bio‑Me’s Precision Microbiome Profiling (PMP™)—a highly parallelized platform measuring absolute abundances of 108 microbial species and subspecies using standardized, calibration‑based quantification. 

Using PMP™, the authors observed strong correlation with metagenomic data, and importantly, that species depletion measured by PMP™ was associated with clinical response.