Concomitant medication reporting should accompany fecal microbiota transplantation plus anti-PD-1 therapy in gastric cancer

Abstract

This commentary discusses the phase I study of fecal microbiota transplantation plus anti-programmed cell death protein 1 therapy in refractory microsatellite-stable gastric cancer. We suggest that this strategy should be treated as a pharmacomicrobiomic intervention rather than only as an immunotherapy combination. Evidence from fecal microbiota transplantation trials and microbiome immunotherapy studies indicates that antibiotics, proton pump inhibitors, corticosteroids and other microbiome-modifying exposures may affect donor strain engraftment, immune activation, response assessment and safety. In China and other Asian settings, where acid suppression, Helicobacter pylori history, perioperative antibiotic use and nutritional interventions are common, standardized concomitant medication reporting would make future studies more interpretable and transferable.

Manuscript

Gastric cancer remains a major setting in which immune checkpoint blockade has substantial but uneven clinical activity, particularly in microsatellite-stable disease. The phase I study by Zhang et al is therefore an important early report because it tests whether fecal microbiota transplantation can be combined with anti-programmed cell death protein 1 (PD-1) therapy to improve treatment responsiveness in patients with refractory microsatellite-stable gastric cancer.¹ Early trials in melanoma and other solid tumors have shown that fecal microbiota transplantation from immunotherapy-responsive or healthy donors can reshape the gut microbiome, support donor-derived microbial engraftment and restore or enhance anti-PD-1 sensitivity in a subset of patients.2,4

As this approach moves from proof of concept to larger studies, we believe that fecal microbiota transplantation plus anti-PD-1 therapy should be viewed not only as a biologic immunotherapy strategy but also as a pharmacomicrobiomic intervention. Its apparent efficacy may be influenced by medications and supportive care measures that alter microbial ecology, mucosal immunity or systemic immune tone. If these exposures are incompletely recorded, donor strain engraftment, immune response and safety findings may be difficult to interpret. This concern is supported by cross-disease metagenomic evidence showing that donor strain engraftment after fecal microbiota transplantation is variable and is shaped by clinical and treatment-related factors rather than by donor material alone.⁵

This issue is particularly relevant in gastric cancer. Patients often receive acid suppression for reflux symptoms, ulcer risk, post-gastrectomy symptoms, dyspepsia or chemotherapy-related gastrointestinal discomfort. Many also have a history of Helicobacter pylori testing or eradication, perioperative antibiotic exposure, biliary infection, nutritional support or treatment for diarrhea. These are not minor background variables in a microbiome-directed trial. The evidence base of the influence of concomitant medications in patients with gastric cancer receiving fecal microbiota transplantation and anti-PD-1 therapy is still in development, but available studies already show why careful covariate control is needed. In melanoma, fecal microbiota transplantation combined with anti-PD-1 rechallenge or first-line anti-PD-1 therapy has produced clinical benefit in a subset of patients, and the 2023 phase I trial by Routy et al reported donor strain engraftment in treated patients, with donor-patient microbiome similarity increasing over time in responders.² In the FMT-LUMINate phase II trial, fecal microbiota transplantation plus immunotherapy was extended to non-small cell lung cancer and melanoma, further supporting the clinical relevance of microbiome modulation in immune checkpoint blockade.³ In a later solid tumor study, clinical benefit was again observed in some patients, including gastrointestinal cancers, and response was linked to sustained microbial changes and immune activation.⁴ These findings support biological plausibility, but they also indicate that donor selection alone is unlikely to explain all interpatient variation. In small early trials, even well-selected donors, standardized fecal microbiota transplantation procedures and immune checkpoint blockade may produce heterogeneous outcomes, which makes documentation of background exposures important for interpreting both positive and negative results.

Several lines of evidence support this concern. In epithelial tumors, Routy et al showed that antibiotic exposure was associated with impaired clinical benefit from PD-1 blockade and that microbiome transfer from responding patients could improve anti-PD-1 effects in mouse models.⁶ In gastrointestinal cancer, Peng et al reported associations between gut microbial features and clinical response to anti-PD-1 or anti-programmed death-ligand 1 (PD-L1) therapy.⁷ Antibiotics are the most obvious example. They may be necessary for infection, cholangitis, perioperative management or complications of advanced disease, but they can reduce microbial diversity and interfere with donor strain engraftment. Retrospective studies and meta-analyses suggest that proton pump inhibitor use may be associated with poorer outcomes in patients receiving immune checkpoint inhibitors, although confounding by comorbidity and indication remains difficult to remove.⁸ Corticosteroids and other immunosuppressive agents are also important because they may dampen the immune activation that fecal microbiota transplantation is intended to support. Baseline corticosteroid exposure has been associated with worse outcomes after PD-1 or PD-L1 blockade in some clinical settings.⁹ Probiotics, enteral nutrition, dietary changes, antidiarrheal agents and traditional medicines are less consistently studied, but dietary fiber and probiotic exposure have been linked to microbiome composition and immunotherapy outcomes in melanoma, supporting the need to record these exposures in microbiome-immunotherapy studies.¹⁰

For future trials, concomitant exposures should be reported with the same care as donor screening, stool processing and immune monitoring. A minimal reporting framework could include the exposure window before fecal microbiota transplantation, changes during anti-PD-1 therapy, indication, agent class, dose or intensity, duration and temporal relationship to stool sampling, infusion or capsule administration, radiographic assessment and adverse events. For antibiotics, reporting should include class, spectrum, route and indication. For acid suppression, proton pump inhibitors and histamine 2 receptor antagonists should be separated. For corticosteroids, dose equivalence and indication should be distinguished, since steroids used for brain metastases, nausea, appetite or immune-related adverse events may have different implications. This does not reduce the value of the study by Zhang et al. Rather, it highlights how to make the next generation of microbiome immunotherapy trials more reproducible. In Chinese and Asian gastric cancer practice, medication exposure is closely tied to tumor location, surgery, H. pylori history, supportive care and access to microbiome testing. Without systematic reporting, it may be difficult to know whether a change in microbial composition reflects donor engraftment, a medication-driven shift or natural fluctuation during treatment. It would also make adverse-event interpretation more reliable, because diarrhea, infection, appetite loss or steroid treatment may be recorded as isolated clinical events unless their timing is linked to stool sampling and anti-PD-1 dosing. Such temporal annotation is particularly useful when microbiome endpoints are used as exploratory biomarkers rather than as purely descriptive data. Table 1 summarizes a concise practical set of exposures that could be reported without making trial publications unnecessarily complex.

Table 1. Medication and supportive-care exposures to report in fecal microbiota transplantation plus anti-PD-1 studies.

Exposure	Why it matters	Minimum reporting items	Source or supporting evidence
Antibiotics	Can reduce diversity and confound donor engraftment or response assessment	Class, spectrum, route, indication, dates, timing relative to stool sampling and FMT	Antibiotics and treatment-related factors are linked to impaired PD-1 benefit and variable donor strain engraftment.5 6
Acid suppression	Common in gastric cancer care and may alter gastrointestinal microbiota	PPI or H2 receptor antagonist, dose, indication, duration, timing	PPI use has been associated with poorer ICI outcomes, although confounding remains possible.8
Corticosteroids and immunosuppressants	May suppress immune activation and complicate response or adverse-event interpretation	Prednisone-equivalent dose, indication, duration, baseline or on-treatment exposure	Baseline corticosteroids have been associated with poorer PD-1/PD-L1 blockade outcomes.9
Probiotics, prebiotics and nutritional interventions	May reshape microbial composition and obscure donor-derived effects	Product or regimen, dose, duration, diet or enteral nutrition changes	Dietary fiber and probiotic exposure have been linked to microbiome and immunotherapy outcomes.10
Antidiarrheal and gastrointestinal supportive drugs	May reflect toxicity, infection or altered transit and affect stool interpretation	Agent, indication, timing, symptom severity	Timing of symptoms and supportive drugs should be recorded when interpreting microbial endpoints.5
Traditional or over-the-counter medicines with possible microbiome effects	May introduce unmeasured microbial or immune-modifying exposure in Asian practice	Name, formulation, duration, timing, reason for use	Direct evidence is limited, but reporting is justified because donor engraftment is treatment-context dependent.5

PPI, proton pump inhibitor.FMT, fecal microbiota transplantation; H2, histamine 2; ICI, immune checkpoint inhibitor; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1.

Conclusions

Fecal microbiota transplantation plus anti-PD-1 therapy is an important direction for refractory microsatellite-stable gastric cancer. Its clinical translation will depend not only on identifying favorable bacterial taxa, but also on controlling and reporting the medications that may determine whether those taxa can function in patients. A standardized concomitant medication table in future protocols and publications would improve interpretation, safety assessment and regional transferability, especially in China and other Asian countries where gastric cancer is common and fecal microbiota transplantation plus anti-PD-1 therapy is being studied.