Stool Transplants Revived Immunotherapy in Patients Whose Liver Cancer Had Stopped Responding

When a patient's tumor stops responding to immunotherapy, the standard assumption is that the immune system has hit a wall. A phase 2 clinical trial led by J Ma, Y Zheng, Z Wang, and X Wang suggests the wall may actually be in the gut. The team tested whether transplanting fecal microbiota from immunotherapy-responsive donors could reactivate anti-tumor immunity in patients with unresectable hepatocellular carcinoma (HCC) who had already failed PD-1 inhibitor therapy. The results are modest in scale but scientifically striking.

Fecal Transplant Achieves Disease Control in 46% of Refractory HCC Patients

The trial enrolled patients with unresectable HCC who had confirmed disease progression on nivolumab monotherapy; roughly half had primary resistance and half had secondary resistance. After receiving a fecal microbiota transplant (FMT) via colonoscopy from donors with durable responses to PD-1 blockade, patients resumed anti-PD-1 therapy. The objective response rate was 7.7%, with one patient achieving a partial response. More notably, the disease control rate reached 46.2%: six of thirteen patients stopped progressing. Earlier work on FMT reversing ICI resistance in solid cancers had hinted this was possible, but this trial specifically targets HCC and pairs FMT with tyrosine kinase inhibitor combinations.

One patient, designated R7, offers the clearest illustration of the mechanism at work. After initially progressing following a first FMT, R7 received a second transplant from a different donor and achieved a partial response. Researchers then isolated a specific bacterial species from R7's stool: Prevotella merdae Immunoactis.

Why Liver Cancer Patients Have So Few Options After Immunotherapy Fails

Unresectable HCC is among the most treatment-resistant cancers in systemic oncology. Immune checkpoint inhibitors represented a major advance, but a significant share of patients never respond or lose their response after initial benefit. Once a PD-1 inhibitor fails, the options narrow sharply: subsequent lines of therapy offer diminishing returns, and most patients in this trial had already received a median of three prior systemic regimens. The median age was 60, and 77% were male, consistent with the broader HCC population.

The gut microbiome has emerged as a credible explanation for why some patients respond to checkpoint inhibitors and others do not. The immune system and the microbial communities in the colon are in constant dialogue, and disrupting or restoring that dialogue appears to shift the balance of anti-tumor immunity in ways that drugs alone cannot.

How Colonoscopy-Delivered Donor Stool Reshaped the Tumor Microenvironment

The trial's design was methodologically careful for its size. Donors were selected specifically because they had achieved durable complete or partial responses (lasting at least six months) to anti-PD-(L)1 monotherapy for advanced solid tumors. Six donors were used in total, four with HCC backgrounds, one with gastric cancer, and one with esophageal squamous cell carcinoma. FMT was delivered by colonoscopy, and repeated transplants from the same or different donors were permitted.

Microbiome sequencing revealed that responders' gut flora shifted significantly toward the donor's composition, while non-responders showed less colonization. Alpha diversity, a measure of the variety of bacterial species present, was significantly higher in responders. Differential bacterial species were enriched in responders, along with changes in biological process functions including the dTMP biosynthetic process.

The isolated bacterium Prevotella merdae Immunoactis proved particularly active in preclinical experiments. It activated both CD4+ and CD8+ T cells in human samples, driving increased secretion of interferon-gamma. In a syngeneic mouse model, P. merdae Immunoactis suppressed tumor growth by enhancing cytotoxic T cell infiltration into tumors. When combined with anti-PD-1 antibody in mice, tumor volume shrank more effectively than with anti-PD-1 alone.

Small Sample Size and Single-Arm Design Leave Key Questions Open

Thirteen patients is a small cohort, and the trial's single-arm design means there is no comparator group to rule out spontaneous disease stabilization or other confounders. The 7.7% objective response rate, while encouraging for a population that had exhausted prior lines of therapy, cannot be confidently generalized without larger, randomized confirmation. FMT in cancer treatment also carries procedural and safety challenges that will need to be addressed at scale: standardized donor selection protocols, pathogen screening, and long-term safety monitoring are all areas requiring further development.

It also remains unclear which patients are most likely to benefit from FMT-based microbiome modulation. The heterogeneity among the six donors, and the fact that R7 required a second transplant from a different donor to achieve a response, suggests donor matching may be critical but is not yet well understood.

What Comes Next for Microbiome-Based Immunotherapy in Liver Cancer

The identification of Prevotella merdae Immunoactis as a biologically active species opens a more targeted research path. Rather than whole-stool transplants, future approaches might deliver specific bacterial strains or their metabolites as defined therapeutics, reducing variability and safety concerns. The preclinical evidence that P. merdae Immunoactis synergizes with anti-PD-1 blockade in mice gives investigators a mechanistic hook for designing the next generation of studies.

A larger randomized trial is the obvious next step, ideally with microbiome profiling at baseline to test whether donor-recipient matching improves outcomes. The trial ran from January 2019 to August 2020, meaning the data have had years to mature, and the bacterium isolation work represents a concrete translational lead. For patients with advanced HCC who have failed two or more prior regimens, even a 46% disease control rate in a refractory population is a signal worth following.