Fecal microbiota transplantation plus immune checkpoint inhibitor rechallenges in patients with advanced non-small cell lung cancer: a single-arm exploratory study

Introduction

Immunotherapy targeting the programmed death ligand 1 (PD-L1)/programmed death 1 (PD-1) pathway has become the standard of care for advanced non-small cell lung cancer (NSCLC) patients without driver gene mutations (1-3). However, the majority of patients will ultimately develop resistance, leading to disease progression and a dire need for effective subsequent therapies (4). While chemotherapy or anti-angiogenic therapy are commonly used post-progression, they offer limited efficacy accompanied by cumulative toxicity (5,6). One potential strategy for these patients is the “rechallenge” with or “cross-line application” of immune checkpoint inhibitor (ICI), particularly for those who initially derived clinical benefit but discontinued treatment due to immune-related adverse events (irAEs) or transient progression. However, current clinical data indicate that the overall response rate (ORR) to rechallenge was generally low, typically ranging from 10% to 20%, as the underlying immunosuppressive mechanisms may even be reinforced upon disease progression (PD) (7-9). Therefore, there is an urgent, unmet need for novel therapeutic strategies capable of effectively reversing the immunosuppressive tumor microenvironment and safely enhancing the efficacy of ICI rechallenge.

The gut microbiome, a large collection of micro-organisms that co-evolve with their host, has recently been confirmed as a pivotal regulator of cancer immunotherapy (10-12). Modulating the composition of intestinal microbiota appears to be easier than altering host genes. Fecal microbiota transplantation (FMT), a contemporary approach to restore gut microbiota homeostasis by transferring fecal material from donors to recipients, has been widely applied to treat recurrent Clostridioides difficile infections and has gained widespread use as a novel treatment for several other diseases, such as Parkinson’s disease, obesity, multiple sclerosis, and irritable bowel syndrome (13-15). These new insights have generated interest in the application of FMT in anti-tumor immunotherapy. Preclinical studies established a causal relationship, demonstrating that FMT from ICI-responding patients could reconstitute the gut microbiome and enhance ICI efficacy in germ-free or antibiotic-treated mouse models, primarily by promoting the infiltration and activation of cytotoxic T cells within the tumor microenvironment (16,17). Building on this compelling rationale, translational research has rapidly advanced to early-phase clinical trials. A pioneering phase I clinical trial by Baruch et al. reported that three of 10 ICI-refractory metastatic melanoma patients showed an objective response after undergoing FMT from responding patients (18). Similarly, Davar et al. reported that six of 15 melanoma patients showed restored ICI activity after undergoing FMT from responding patients (19). Furthermore, the application of FMT is being explored in gastrointestinal cancers and for managing severe irAEs, such as colitis (20). A recent narrative review also summarized the efficacy and regulatory strategies of gut microbiota in immunotherapy, highlighting the translational potential of modulating the microbiome (21). While these findings position FMT from responders as a promising strategy to modulate the host immune system and improve oncological outcomes, this approach faces significant limitations in donor availability, standardization, and long-term safety profiling. In contrast, the use of rigorously screened healthy donors offers a more feasible and potentially safer alternative. The selection of healthy donors for FMT in oncology is supported by extensive experience in other immune-related conditions. For instance, in inflammatory bowel disease, the use of rigorously screened healthy donors has been a cornerstone of FMT strategies, balancing safety and efficacy (22).

Jin et al. demonstrated that higher baseline gut microbial diversity was significantly associated with improved objective response rate (ORR) and survival outcomes in a Chinese cohort (23). Beyond the well-documented detrimental effects of antibiotics, emerging evidence suggests that the neoplastic disease state itself, along with prior cytotoxic therapies such as chemotherapy and radiation, can induce a profound and persistent dysbiosis. As comprehensively reviewed by Zitvogel et al., these interventions disrupt intestinal microbial diversity and deplete key immunostimulatory bacterial species, thereby eroding the foundation for effective anti-tumor immunity (24). Thus, we have designed a single-arm, exploratory trial to evaluate the safety and feasibility of healthy donor FMT combined with ICI rechallenge in patients with advanced NSCLC who have progressed on prior PD-1/PD-L1 inhibitor therapy grounded in the principle of “ecological restoration” of a dysbiotic gut environment. We present this article in accordance with the TREND reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-973/rc).

Methods

This was a single-center, single-arm, prospective study (Chinese Clinical Trial Registry, ChiCTR2100043472). The primary endpoints were the safety and tolerability of combining FMT with PD-1 blockade rechallenge therapy in patients with NSCLC. Adverse events (AEs) were evaluated according to the Common Terminology Criteria for Adverse Events, version 5.0 (CTCAE v5.0). The secondary endpoint was efficacy, including progression-free survival (PFS) (defined as the time from the first day of study intervention to the date of PD or death from any cause, whichever occurred first), and overall survival (OS) (defined as the time from the first day of study intervention to the date of death from any cause). Tumor response was assessed using the Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST v1.1). The exploratory objective was to examine the effects of ICI rechallenge and FMT administration on the composition and function of the gut microbiota.

Patients

Recruitment for this trial was mainly carried out at outpatient clinics of the West China Hospital of Sichuan University. Patients were included in the study if they met the following inclusion criteria: (I) were aged 18 years or older; (II) had a diagnosis of advanced NSCLC; (III) had wild-type EGFR and no ALK/ROS1 gene rearrangements; (IV) had achieved a complete response, partial response (PR), or stable disease (SD) during previous treatment with PD-1 or PD-L1 inhibitors; (V) experienced PD as shown on radiographic imaging during previous PD-1 or PD-L1 inhibitor treatment; (VI) had an Eastern Cooperative Oncology Group performance status (ECOG-PS) score of 0–2; (VII) had at least one evaluable lesion according to the RECIST v.1.1; (VIII) had a minimum washout period of 4 weeks from the previous treatment line; and (IX) had recovered from all toxicities or complications of previous treatments to ≤ grade 1 or to baseline. Patients were excluded from the study if they met any of the following exclusion criteria: (I) had an inability to swallow capsules; (II) had a history of inflammatory bowel disease; (III) had a history of major abdominal surgery; (IV) had a history of gastrointestinal perforation, hemorrhage, or obstruction within the 6 months before study enrollment; (V) had an active or previously documented autoimmune disorder; (VI) had current exposure to high dose oral or intravenous corticosteroids (>10 mg of prednisone daily or equivalent) (patients who intermittently required the use bronchodilators or local steroid injections were not excluded); (VII) had an infection requiring systemic therapy within 28 days before the initiation of the study treatment; (VIII) had a history of another malignancy, either previous or concurrent; (IX) had a serious or unstable symptomatic cardiovascular disease; (X) had symptomatic brain or leptomeningeal metastasis; and/or (XI) had a psychiatric or substance abuse disorder, or any other clinically relevant physical or mental condition that would interfere with compliance with the trial requirements.

This study was conducted in accordance with the principles of the Declaration of Helsinki and its subsequent amendments, and was approved by the Ethics Committee on Biomedical Research, West China Hospital of Sichuan University (approval No. 2021121). All patients provided written informed consent prior to the study and were allowed to withdraw consent at any point.

Intervention procedures

FMT administration can be performed via upper gastrointestinal methods (e.g., nasoduodenal tubes and oral capsules) or lower gastrointestinal methods (e.g., colonoscopy, enema, and rectal tubes). Given that patients with NSCLC have a lower demand for endoscopy compared to those with intestinal diseases, and the higher operational risks associated with endoscopy, this study selected the more convenient and acceptable method of oral capsules to standardize the administration route and enhance patient compliance. Three eligible fecal donors were selected through a rigorous and standardized screening process based on previously established criteria to ensure consistency and safety (25). All FMT capsules were prepared uniformly by Xiamen Treat-gut Biotechnology Co., Ltd. (Xiamen, China) from the stool of these pre-screened healthy donors. Donors were excluded from the study of they met any of the following major exclusion criteria: (I) had a history of human immunodeficiency virus (HIV), hepatitis A virus (HAV), hepatitis B virus (HBV), or hepatitis C virus (HCV) infection; (II) had a history of smoking or alcohol consumption; (III) had received antibiotics or proton pump inhibitor (PPI) in the previous 3 months; (IV) had a history of intrinsic gastrointestinal disease (e.g., inflammatory bowel disease, irritable bowel syndrome, chronic constipation, gastrointestinal malignancy, and/or a prior major gastrointestinal surgery or procedure); (V) had a history of another disease (e.g., a malignancy, malnutrition, an autoimmune or atopic disease, a cardiovascular or metabolic disease, diabetes, hypertension, stroke, and/or a psychiatric disorder); (VI) had an oral disease (e.g., a periodontal disease, a mucosal disease, or oral cancer); (VII) had positive results for gastrointestinal pathogens (e.g., Helicobacter pylori); and/or (VIII) had positive results for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).

The PD-1 blockade used in this study was camrelizumab (200 mg every 3 weeks). Routine re-examination and evaluation were conducted after every two cycles of camrelizumab. Safety data were collected from the time of enrollment until 30 days post-treatment discontinuation. Treatment was discontinued due to either PD or unacceptable toxicity.

Recipients were required to ingest 90 capsules (“initial FMT”) before the first dose of anti-PD-1 rechallenge therapy. To optimize bacterial colonization by donor flora, these 90 capsules were administered in three divided doses with a one-day interval between each administration. Additionally, a dose of 12 capsules, termed “consolidation FMT”, was administered before each of the subsequent five cycles of camrelizumab. Each immunotherapy treatment was administered two days after the FMT treatment (Figure 1). All the trial medications were administered by a study nurse in an inpatient setting to facilitate the close monitoring of AEs and the timely management of any emergent incidents. The study flow chart is shown in Figure 1.

Figure 1 Study flow diagram. FMT, fecal microbiota transplantation; PD-L1, programmed death ligand 1.

The patients were instructed to consume a low-fiber diet 1 day before capsule administration and to fast for at least 4 hours before taking the capsules. They were permitted to resume a liquid diet two hours after dosing and were advised to maintain a light diet, avoiding spicy and strongly flavored foods.

Data and sample collection

Demographic and clinical information, including tumor diagnosis (e.g., pathologic type, tumor, node, metastasis (TNM) stage, site/s of distant metastasis, PD-L1 expression level, and mutation status) and previous anti-tumor therapy (e.g., number of therapy lines, regimen, start and end dates, and best response) were recorded for all patients. Hematological test results, imaging assessments, and AE records were collected at each scheduled visit. Fecal samples for the microbial analysis and peripheral blood samples for the T cell receptor (TCR) repertoire analysis were collected at the baseline (S1), at the first assessment of efficacy (S2), after six cycles of immunotherapy (S3), and at the time of PD (S4). Feasibility was assessed by the accrual rate (number of patients enrolled per month) and the proportion of patients completing both the initial FMT and at least one cycle of consolidation FMT alongside the planned ICI therapy. Patients were followed up every 3 weeks during treatment for safety and efficacy assessments. After treatment discontinuation, survival follow-up was conducted every 3 months until death or the data cut-off date.

16S ribosomal ribonucleic acid (16S rRNA) sequencing

16S rRNA sequencing was conducted by Xiamen Treat-gut Biotechnology Co., Ltd. (Xiamen, China). Bacterial DNA was extracted from the fecal samples using the PowerSoil DNA Isolation Kit (Mo Bio Laboratories, Inc., Carlsbad, CA, USA; now part of Qiagen, Hilden, Germany). The V3–4 region of the bacterial 16S rRNA gene was amplified using primers 341F/515F and 806 R. The forward primer sequence of 16S was 5'-GTGCCAGCMGCCGCGGTAA-3', and the reverse primer sequence was 5'-GGACTACNVGGGTWTCTAAT-3'. The polymerase chain reaction (PCR) products were purified and evaluated using the Qubit 3.0 Fluorometer (Thermo Fisher Scientific, Waltham, US). The PCR products were further purified using the Qiagen Gel Extraction Kit (Qiagen). Sequencing libraries were generated using the TruSeq^® DNA PCR-Free Sample Preparation Kit (Illumina, San Diego, CA, USA) according to the manufacturer’s recommendations with index codes added. The quality of the library was assessed using the Qubit 3.0 Fluorometer (Thermo Fisher Scientific) and the Agilent Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA). Finally, the library was sequenced on an Illumina NovaSeq platform (Illumina, Inc., San Diego, CA, USA), generating 250 bp paired-end reads.

Metabolomic profiling

Metabolomic profiling was conducted using the Vanquish UHPLC system (Thermo Fisher Scientific). A loading plot based on orthogonal projections to latent structures-discriminant analysis (OPLS-DA) was generated to demonstrate the contribution of variables to differences between two groups. The loading plot highlighted important variables situated far from the origin, but it appeared complex due to the numerous variables. To refine this analysis, the first principal component of variable importance in the projection (VIP) was obtained. Metabolites with VIP values exceeding 1.0 were initially selected as significantly altered. Subsequently, the remaining variables were assessed using the Student’s t-test, and those with a P value >0.05 were discarded. Additionally, commercial databases such as the Kyoto Encyclopedia of Genes and Genomes (http://www.genome.jp/kegg/) and MetaboAnalyst (http://www.metaboanalyst.ca/) were used to identify the metabolic pathways involved.

TCR sequencing

Antigen recognition by T cells requires the interaction between major histocompatibility complex (MHC) complexes and TCRs. The diversity of TCRs is primarily conferred through the third hypervariable region of the receptor chains, known as the complementarity-determining region 3 (CDR3), which is encoded by the V(D)J genes. To better understand the immunologic changes occurring in patients in this trial, CDR3 sequencing in peripheral blood was conducted.

Genomic DNA was extracted from frozen peripheral blood mononuclear cell (PBMC) specimens using a High Pure PCR Template Preparation Kit (Roche Diagnostics, Basel, Switzerland). Multiplex PCR amplification of the CDR3 region of the TCR β chain was performed in two rounds, PCR1 and PCR2. The first round, PCR1, used specific primers against each V and J gene, with the primer sequences filed as part of a Chinese patent (CN105087789A). The second round, PCR2, employed universal primers. Paired-end sequencing of samples was conducted with a read length of 151 bp using the Illumina HiSeq 3000 platform (Illumina, San Diego, CA, USA). The raw sequencing data were processed and analyzed as follows: (I) the raw data were filtered using CutAdapt to remove sequences that did not contain the primers for multiplex PCR; (II) high-quality paired reads were merged to obtain contigs using the Paired-End Read Merger (PEAR) program; and (III) Mixcr was used to align the reference TCR β chain V(D)J gene sequences and determine the TCR β chain V(D)J gene segments in each contig.

Statistical analysis

The present study was designed as an exploratory investigation to assess the preliminary feasibility of combining FMT with ICI rechallenge and to identify potential problems to guide the design of future experiments. Given the exploratory nature, no formal sample size estimation was performed; a sample size of 5–10 patients was deemed sufficient for an initial assessment of feasibility and safety parameters. Raw reads obtained from the 16S rDNA sequencing platform were processed using Flash software to obtain clean reads, followed by chimera filtering to obtain Chimera_check, and finally, operable taxonomic units (OTUs) were clustered using QIIME software. The species diversity analysis included both α-diversity and β-diversity analyses. A linear discriminant analysis of effect size was conducted to identify species with significantly different abundances between groups. Data are presented as the mean ± standard deviation, or the median (range). The survival analysis was conducted using the Kaplan-Meier method. For all the statistical analyses, the P values were two-tailed, and the alpha level was set at 0.05. Given the exploratory nature of this pilot study and the small sample size (N=7), all statistical comparisons, particularly those involving subgroup analyses, are severely underpowered. Therefore, P values are provided for descriptive purposes only, and emphasis is placed on the observed effect sizes and trends rather than on statistical significance. The results should be interpreted as generating hypotheses for future research. All statistical analyses were performed using R (version 3.5.3) or GraphPad Prism 9.0 statistical software.

Results

Baseline characteristics of patients

From May 2021 to August 2022, 10 NSCLC patients were screened, of whom seven were enrolled in the study and three were excluded from the study due to failed screening. The baseline demographic and clinical characteristics of the patients are presented in Table 1. All the patients were male with a median age of 55 (range, 43–73) years, and most patients (5/7, 71.4%) had a history of smoking. The ECOG-PS score of all patients was 0–1. In terms of their nutritional status, all the patients had a normal weight with a body mass index ranging from 21.3 to 24.5 kg/m². All the patients were pathologically diagnosed with adenocarcinoma (n=4, 57.1%) or squamous cell carcinoma (n=3, 42.9%). At enrollment, all the patients suffered from intrathoracic metastasis (lung and/or pleural), and four patients had extrathoracic metastasis, of whom, all had bone metastasis, one had brain metastasis, and one had liver and peritoneal metastasis, simultaneously. All the patients underwent genetic testing and PD-L1 (22C3) expression assays on lung tissue before enrollment. Targetable driver mutations, such as EGFR/ROS-1 mutations and ALK rearrangements, were not detected. In terms of PD-L1 expression, one patient was PD-L1 negative, five had low PD-L1 expression (1–49%), and one had strong PD-L1 expression (≥50%).

All the patients had received prior treatment of PD-L1/PD-1 inhibitor in combination with chemotherapy, of whom, three had achieved SD and four had achieved PR as the best response according to the RECIST v1.1. The PFS of all the patients ranged from 1.5–29.9 months. The majority of patients had received multiple lines of anti-tumor therapy, with a median of three lines of prior therapy, and 5 (71.4%) patients had undergone three or more lines of therapy (for further details, see Table S1).

Safety

All the patients completed the trial and were included in the intention-to-treat analysis. During the study period, all the patients experienced at least one AE. A total of 66 AEs were reported, and categorized as grade 1 (69.7%), grade 2 (22.7%), and grade 3 (7.6%), with no AEs above grade 3 observed (Table 2). Grade 3 AEs were reported in two cases. One patient (recipient #3), who had asymptomatic brain metastases at the baseline, developed grade 3 dizziness and decreased appetite during the intervention. These symptoms improved after dehydration and resolved within 15 days, allowing the patient to continue the trial treatment and to subsequently achieve a PR. Another patient (recipient #4) developed Grade 3 pneumonia, hyponatremia, and decreased appetite simultaneously after two cycles of camrelizumab administration. This necessitated a 1-month interruption in treatment, leading to the withdrawal of the patient from the study. Although the patient’s symptoms eventually improved, the patient later died due to PD (Table 2).

As set out in Table 3, treatment-related AEs were primarily associated with camrelizumab. The most frequent irAEs were asymptomatic hypothyroidism (n=3, 42.9%) and reactive cutaneous capillary hyperplasia (n=3, 42.9%). Overall, five patients (71.4%) experienced irAEs, but none experienced any previously unreported irAEs. Four patients (57.1%) experienced FMT-related AEs, which were all Grade 1 events, including nausea, diarrhea, bloating, and constipation. These reactions were mild, transient, and self-limiting.

Response and survival

As of the cut-off date on January 30, 2024, the median duration of follow-up was 26.1 months [95% confidence interval (CI): 19.4–32.8]. All the enrolled patients received the initial FMT and the first dose of consolidation FMT from one of the three available donors. Five of the seven patients (recipients #2, #4, #5, #6, and #7) discontinued protocol treatment during the first evaluation due to PD. The remaining patients (recipients #1 and #3) derived a clinical benefit and continued the treatment. Both recipients #1 and #3 proceeded with four additional doses of consolidation FMT as planned and received maintenance single-agent camrelizumab until PD. Ultimately, recipient #1 underwent a total of 20 cycles of immunotherapy, while recipient #3 underwent 10 cycles. And recipient #1 achieved a PR and recipient #3 achieved SD as best response, which was consistent with their best response to prior immunotherapy.

At the data cut-off time, three patients were still alive. The ORR was 14.3%, and the disease control rate (DCR) was 28.6%. The median PFS was 1.5 months (95% CI: 1.24–1.75), and the median OS was 12.1 months (95% CI: 0.3–23.9) for all patients (Figure 2). For the two responders, the PFS was 14.6 months (recipient #1) and 8.1 months (recipient #3), while the OS was 30.6 months (recipient #1) and 12.1 months (recipient #3).

Figure 2 Survival analysis (n=7). Progression-free survival (A) and overall survival (B).

Changes in gut microbiota composition

A 16S rRNA sequencing analysis was conducted to analyze the baseline information of the recipients and donors. The OTU abundance statistics are shown in Table S2. The community structures of the gut flora are shown in Figure S1. The diversity indices showed that the donors appeared to have higher richness than the recipients; however, there was no statistical significance between the donors and recipients in terms of alpha diversity (Shannon’s index, P=0.055). Escherichia-Shigella, Klebsiella, un_f_Muribaculaceae, Dialister, Pseudomonas, Odoribacter, and Fusobacterium were more frequent in the donor group, while Fusicatenibacter, Christensenellaceae_R-7_group, and Ruminococcaceae UCG-005 were more frequent in the recipient group as indicated in Figure S2 (P<0.05).

The gut microbiota structure of the recipients altered after treatment. Recipient #4 was not included in this analysis due to constipation and an inability to provide a sample at the indicated time. Figure 3 shows a heatmap of the microbiological profile of each patient, pre- and post-initial FMT, and their donor. The gut microbiota compositions of each group at the different levels are displayed in Figure S3. The analysis of similarities indicated a slight but non-significant difference after treatment (R=0.006, P=0.41); however, while most species decreased, two original species (Prevotella_9 and Veillonella) increased, and two new species (Megamonas and Alloprevotella) appeared at the genus level. However, no statistically significant differences in the diversity analysis or differential analysis were found between the groups.

Figure 3 Cluster heat map of species abundance in recipients pre- and post-treatment and their donors at genus level.

Gut microbiota and clinical response

The patients were stratified into responsive (R) and non-responsive (NR) groups based on whether or not they experienced PD at the first evaluation. Before treatment, there was no significant difference in the gut flora diversity between the two groups, and no signature bacteria distinguished the groups (P>0.05). However, after receiving FMT from the standardized healthy donors, distinct patterns of changes in diversity emerged between the R and NR groups. Specifically, the NR patients had lower Chao1 and ACE indices but higher Shannon’s and Simpson’s indices than the R patients at the baseline; this pattern was reversed after FMT (Figure 4A,4B). The top 20 dominant gut microbiota at the genus level after the intervention in both groups are illustrated in Figure S4. Patients in the R group had higher levels of Prevotella_9, Phascolarctobacterium, Veillonella, Prevotella_2, unclassified_f_Muribaculaceae, Alloprevotella, Lachnospira, Paraprevotella, Parabacteroides, and Sutterella, [Eubacterium]_eligens_group compared to the NR group. Additionally, f_Eggerthellaceae was significantly enriched in the NR group after treatment (P<0.05).

Figure 4 Alpha diversity analyses between R group and NR group pre-treatment (A) and post-treatment (B). The observed species, ACE, and Chao1 indicate richness; the Shannon index reflects diversity; the Simpson and Pielou’s (J) indices assess dominance and evenness. Blue, pre-treatment; Orange, post-treatment. ACE, abundance-based coverage estimator; NR, non-responsive; R, responsive.

Changes in host metabolism

The OPLS-DA was used to identify inter-group differential metabolites after FMT. A total of 9,138 metabolites were identified in the positive ion mode and 7,578 metabolites in the negative ion mode in the fecal samples. Based on the criteria of P<0.05 and VIP >1, the expression of 114 metabolites was found to differ significantly between pre- and post-treatment. Of these, 73 metabolites increased, and 41 metabolites decreased in the stool specimens of the recipients after FMT (Figure 5A). Eleven of these compounds were identified as known metabolites, including L-pyroglutamic acid, citrulline, indole-3-acetic acid, 2-azetidinecarboxylic acid, O-acetyl-L-serine, D-maltose, traumatic acid, L-norleucine, 3-methyl-2-oxovaleric acid, 1-pyrroline, and 1-methylguanosine. The hierarchical clustering results for these metabolites are presented in a heatmap, illustrating the relationships between the samples and the differential expression of the metabolites among the samples (Figure 5B).

Figure 5 Changes in metabolites after FMT (n=6). (A) Volcano plots for differential metabolites. The abscissa represents the fold change of each metabolity, and the ordinate represents P value of Student’s t-test. (B) Hierarchical clustering analysis of 11 known differential metabolites. (C) Correlation matrix of interaction in 11 known differential metabolites. Blue presents negative correlation; Red presents positive correlation. (D) Pathway analysis for pre- and post-FMT groups. *, P<0.05. FMT, fecal microbiota transplantation.

Additionally, a correlation heatmap was generated using the “corrplot” package in R to explore the correlations between different metabolites. The Pearson correlation analysis demonstrated that the highest correlation was observed between L-pyroglutamic acid and 2-azetidinecarboxylic acid, with a correlation coefficient of 1. Strong correlations were also observed between 1-pyrroline and 2-azetidinecarboxylic acid, L-pyroglutamic acid and citrulline, O-acetyl-L-serine, and 3-methyl-2-oxovaleric acid, and D-maltose, with correlation coefficients of 0.88, 0.88, 0.88, and 0.81, respectively (Figure 5C). These metabolites were subsequently subjected to metabolic pathway analysis using the MetaboAnalyst web-based platform (https://www.metaboanalyst.ca/). As Figure 5D shows, the following five pathways were screened: sulfur metabolism, selenoamino acid metabolism, cysteine and methionine metabolism, arginine and proline metabolism, and tryptophan metabolism. The metabolites related to these pathways were O-acetyl-L-serine, citrulline, and indole-3-acetic acid. Sulfur metabolism emerged as the most significantly differentially expressed pathway following the intervention (P=0.051, approaching statistical significance).

Changes in hematological indices

The clinical laboratory test results of recipients were collected before each immunotherapy cycle to explore the potential clinical implications of the FMT-ICI treatment. As illustrated in Figure S5, the blood routine examination results showed a transient rise in neutrophil counts after the initial FMT, which quickly returned to the baseline levels. This transient increase was likely related to the immune response triggered by exposure to microbiota from donors. Conversely, the lymphocyte counts initially declined in all patients; however, they continued to decline in patients in the R group and increased in the patients in the NR group. Additionally, the number and proportion of CD cell sub-clusters was monitored. The NR group showed a decrease in the CD3, CD4, and CD8 cell counts, while the R group showed an increase in these counts after treatment. In relation to cytokines, the levels of interleukin (IL)-8 and tumor necrosis factor α gradually decreased after treatment. Further, the level of IL-6 initially increased in all patients but subsequently declined only in the NR group.

A TCR analysis was conducted to assess changes in TCR diversity and clonotype distribution. Before treatment initiation, no significant differences in the number of gene clone types were observed between the R group (range, 8,438–14,490) and NR group (range, 13,152–56,119) as determined by the Mann-Whitney U-test. The Shannon entropy analysis also showed no statistically significant difference between the pre- and post-intervention samples, or between the R and NR groups at the baseline (Wilcoxon paired signed-rank test, P>0.05). Next, the quantity of reads of specific T cell receptor beta variable (TRBV) clonotypes was summarized. As shown in Figure 6, significant changes were observed in two V gene fragments (TRBV7-8 and TRBV14), and 13 paired V-J genes (TRBV14_TRBJ1-1, TRBV14_TRBJ1-2, TRBV20-1_TRBJ1-1, TRBV20-1_TRBJ2-5, TRBV7-2_TRBJ1-3, TRBV7-2_TRBJ1-5, TRBV7-2_TRBJ2-1, TRBV7-2_TRBJ2-3, TRBV7-6_TRBJ2-7, TRBV7-8_TRBJ1-1, TRBV7-9_TRBJ2-5, TRBV7-9_TRBJ2-7, and TRBV9_TRBJ1-2) after FMT (P<0.05). Moreover, no differences in the expression of any TRBV clonotype (V/J/V-J gene) were found between the efficacy subgroups. However, due to the small sample size and multiple comparisons, these results should be considered exploratory.

Figure 6 TCR repertoire changes after initial FMT. Bar plots display the normalized frequency of (A) specific V gene segments and (B) paired V-J gene segments. FMT, fecal microbiota transplantation; TCR, T-cell receptor.

Case presentation

Recipient #1 was a 53-year-old, male, ex-smoker, diagnosed with clinical stage T3N2M1a, IVA NSCLC, without EGFR/ROS-1 mutations and ALK rearrangements, and a negative expression level of PD-L1 based on the 22C3 immunohistochemistry assay. He had received three kinds of systemic therapies at enrollment; his drug regimens, treatment response, time to progression, and RT treatment (yes/no) are detailed in Table S1. After signing the informed consent form, he was treated with FMT and ICI rechallenge as described above. His response assessment after two cycles of ICI was SD, and after four cycles, the efficacy was evaluated as a PR. After 20 cycles of ICI therapy, he had developed PD; a computed tomography scan showed the increased size of the lesion in the perihilar and left upper lobe of the lung. The PFS of this patient was 14.6 months (Figure 7).

Figure 7 Computed tomography of the chest of Recipient #1.

The sequencing analysis of the 16S rDNA gene showed that the patient’s intestinal flora composition changed after treatment, but the species diversity was not significantly altered. Peripheral blood samples were dynamically collected from the patient at the baseline, after two cycles of immunotherapy, after six cycles of immunotherapy, and at PD for the TCR sequencing analysis. Compared to the baseline peripheral blood TCR repertoire, new V-J genes were observed post-treatment. Meanwhile, at the baseline and PD time points, the TCR V-J gene clonotypes appeared more similar to each other, whereas the two samples collected during the treatment process showed greater resemblance in their clonal composition. suggesting that the treatment process may reflect the activation or remodeling of immune responses (Figure S6).

Discussion

To our knowledge, this study represented the first exploratory evaluation of the combination of healthy-donor FMT with ICI rechallenge in patients with advanced NSCLC who progressed after initial immunotherapy. The safety profile observed in our study compares favorably with historical data on both FMT and ICI rechallenge safety. In our cohort, only self-limiting grade 1 FMT-related AEs were reported, consistent with the established safety profile of FMT in other indications. Large systematic reviews have shown that serious AEs related to FMT are rare, with a major AE rate of <2% in immunocompetent populations (26-28). Similarly, the immune-related toxicity profile in our study was manageable, with only one patient experiencing grade 3 irAEs. This aligns with the known safety profile of camrelizumab monotherapy, where grade ≥3 irAEs occurred in 9.1% of patients in the phase III CameL trial (29). Importantly, no previously unreported safety signals emerged from the combination approach, suggesting that adding FMT to ICI rechallenge does not appear to exacerbate the toxicity profile of either intervention alone.

The observed clinical outcomes—including an ORR of 14.3% and a median OS of 12.1 months—occurred in a heavily pretreated population, most of whom had low or negative PD-L1 expression. While these outcomes appear numerically favorable compared to historical data on ICI rechallenge alone, the lack of a direct comparator group limit the ability to distinguish the specific effects of FMT from those of ICI rechallenge or other unmeasured variables (30,31).

Previous studies have demonstrated that FMT from ICI-responding donors can overcome resistance in melanoma (18,19), and our study extends this concept to NSCLC using healthy donors—a more scalable and standardized approach. The observed microbial shifts post-FMT, including the increase in Prevotella_9—a genus previously associated with improved ICI responses in some studies (24,32,33)—suggest potential mechanisms worthy of further exploration. Additionally, the metabolic changes observed, particularly in citrulline and indole-3-acetic acid, align with previous reports linking these metabolites to immunotherapy outcomes (34,35). TCR sequencing and peripheral immune cell subset analysis indicated changes in the TCR repertoire composition and in the counts of CD3⁺, CD4⁺, and CD8⁺ T cells. These immunodynamic alterations suggest that the combination of FMT and ICI rechallenge may have modulated systemic immune status. However, due to the small sample size and the lack of a clear correlation with clinical efficacy, the biological significance and functional consequences of these changes remain uncertain, and whether they genuinely reflect the reinstatement of anti-tumor immunity requires validation in future studies.

Several important limitations constrain the interpretability of our findings. First, the most significant limitation is the small sample size and the single-arm design without a control group (e.g., patients receiving ICI rechallenge alone). This prevents causal inference regarding the specific contribution of FMT to the observed safety profile and any potential efficacy signals, such as PD-L1 expression level, which is a key factor influencing ICI response. The results should therefore be interpreted as preliminary, highlighting the feasibility and safety of the combination, and justifying future randomized controlled trials. Second, although patients were advised to follow a specific diet around the time of FMT administration, the content and duration of meals were not strictly controlled or recorded throughout the study. Unrecorded dietary variations remain an uncontrolled confounding factor that may have influenced gut microbiota composition and clinical outcomes. Third, methodological uncertainties remain regarding key aspects of FMT administration. The use of oral capsules was selected for patient convenience, but comparative data on engraftment efficiency relative to other routes are lacking. Similarly, the FMT dosing regimen was empirically defined, and optimal criteria for maximizing donor microbiota colonization require systematic evaluation. Although donors were rigorously screened for safety, their microbiota profiles were not characterized for features potentially linked to ICI response. We did not implement rigorous bowel preparation before FMT, such as the use of broad-spectrum antibiotics and enemas, as studies have indicated that antibiotic use was associated with worse response rates and survival in patients receiving ICI therapy (36,37). However, the impact on donor microbiota engraftment remains unknown.

Several critical questions remain unaddressed and warrant investigation in future studies. These include identifying the specific microbial taxa or metabolic pathways responsible for reversing ICI resistance; establishing criteria for optimal donor-recipient matching; evaluating the long-term safety of FMT in oncology populations; and determining whether immune changes are directly attributable to FMT. The ongoing phase III trials evaluating FMT in combination with ICIs (NCT04116775, NCT04758507) may provide more definitive answers.

Conclusions

This pilot study indicates that combining FMT with ICI rechallenge is feasible and exhibits a tolerable safety profile in advanced NSCLC. The promising clinical activity observed in a subset of patients and the associated biological changes warrant further investigation. These preliminary findings justify subsequent larger, controlled studies to determine the efficacy of this approach.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TREND reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-973/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-973/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-973/prf

Funding: This work was supported by the West China Hospital of Sichuan University (No. 2021HXFH055).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-973/coif). J.H., L.L., and B.S. are from Xiamen Treatgut Biotechnology Co., Ltd., Xiamen, China. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the principles of the Declaration of Helsinki and its subsequent amendments, and was approved by the Ethics Committee on Biomedical Research, West China Hospital of Sichuan University (Approval No. 2021121). All patients provided written informed consent prior to the study and were allowed to withdraw consent at any point.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Zhou C, Chen G, Huang Y, et al. Camrelizumab plus carboplatin and pemetrexed versus chemotherapy alone in chemotherapy-naive patients with advanced non-squamous non-small-cell lung cancer (CameL): a randomised, open-label, multicentre, phase 3 trial. Lancet Respir Med 2021;9:305-14. [Crossref] [PubMed]
2. Xu J, Kato K, Raymond E, et al. Tislelizumab plus chemotherapy versus placebo plus chemotherapy as first-line treatment for advanced or metastatic oesophageal squamous cell carcinoma (RATIONALE-306): a global, randomised, placebo-controlled, phase 3 study. Lancet Oncol 2023;24:483-95. [Crossref] [PubMed]
3. Qin S, Chan SL, Gu S, et al. Camrelizumab plus rivoceranib versus sorafenib as first-line therapy for unresectable hepatocellular carcinoma (CARES-310): a randomised, open-label, international phase 3 study. Lancet 2023;402:1133-46. [Crossref] [PubMed]
4. Sharma P, Hu-Lieskovan S, Wargo JA, et al. Primary, Adaptive, and Acquired Resistance to Cancer Immunotherapy. Cell 2017;168:707-23. [Crossref] [PubMed]
5. Leighl NB, Redman MW, Rizvi N, et al. Phase II study of durvalumab plus tremelimumab as therapy for patients with previously treated anti-PD-1/PD-L1 resistant stage IV squamous cell lung cancer (Lung-MAP substudy S1400F, NCT03373760). J Immunother Cancer 2021;9:e002973. [Crossref] [PubMed]
6. Reckamp KL, Redman MW, Dragnev KH, et al. Phase II Randomized Study of Ramucirumab and Pembrolizumab Versus Standard of Care in Advanced Non-Small-Cell Lung Cancer Previously Treated With Immunotherapy-Lung-MAP S1800A. J Clin Oncol 2022;40:2295-306. [Crossref] [PubMed]
7. Gandara DR, von Pawel J, Mazieres J, et al. Atezolizumab Treatment Beyond Progression in Advanced NSCLC: Results From the Randomized, Phase III OAK Study. J Thorac Oncol 2018;13:1906-18. [Crossref] [PubMed]
8. Tian T, Yu M, Yu Y, et al. Immune checkpoint inhibitor (ICI)-based treatment beyond progression with prior immunotherapy in patients with stage IV non-small cell lung cancer: a retrospective study. Transl Lung Cancer Res 2022;11:1027-37. [Crossref] [PubMed]
9. Hakozaki T. Re-administration of immune checkpoint inhibitors for patients with non-small cell lung cancer. Transl Lung Cancer Res 2022;11:2170-4. [Crossref] [PubMed]
10. Thomas S, Izard J, Walsh E, et al. The Host Microbiome Regulates and Maintains Human Health: A Primer and Perspective for Non-Microbiologists. Cancer Res 2017;77:1783-812. [Crossref] [PubMed]
11. Sepich-Poore GD, Zitvogel L, Straussman R, et al. The microbiome and human cancer. Science 2021;371:eabc4552. [Crossref] [PubMed]
12. Routy B, Le Chatelier E, Derosa L, et al. Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors. Science 2018;359:91-7. [Crossref] [PubMed]
13. Drekonja D, Reich J, Gezahegn S, et al. Fecal Microbiota Transplantation for Clostridium difficile Infection: A Systematic Review. Ann Intern Med 2015;162:630-8. [Crossref] [PubMed]
14. Paramsothy S, Kamm MA, Kaakoush NO, et al. Multidonor intensive faecal microbiota transplantation for active ulcerative colitis: a randomised placebo-controlled trial. Lancet 2017;389:1218-28. [Crossref] [PubMed]
15. Xu MQ, Cao HL, Wang WQ, et al. Fecal microbiota transplantation broadening its application beyond intestinal disorders. World J Gastroenterol 2015;21:102-11. [Crossref] [PubMed]
16. Sivan A, Corrales L, Hubert N, et al. Commensal Bifidobacterium promotes antitumor immunity and facilitates anti-PD-L1 efficacy. Science 2015;350:1084-9. [Crossref] [PubMed]
17. Gopalakrishnan V, Spencer CN, Nezi L, et al. Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients. Science 2018;359:97-103. [Crossref] [PubMed]
18. Baruch EN, Youngster I, Ben-Betzalel G, et al. Fecal microbiota transplant promotes response in immunotherapy-refractory melanoma patients. Science 2021;371:602-9. [Crossref] [PubMed]
19. Davar D, Dzutsev AK, McCulloch JA, et al. Fecal microbiota transplant overcomes resistance to anti-PD-1 therapy in melanoma patients. Science 2021;371:595-602. [Crossref] [PubMed]
20. Wang Y, Wiesnoski DH, Helmink BA, et al. Fecal microbiota transplantation for refractory immune checkpoint inhibitor-associated colitis. Nat Med 2018;24:1804-8. [Crossref] [PubMed]
21. Sang Y, Zheng K, Zhao Y, et al. Efficacy and regulatory strategies of gut microbiota in immunotherapy: a narrative review. Transl Cancer Res 2024;13:2043-63. [Crossref] [PubMed]
22. Bi Y, Cheng B, Zou B, et al. The current landscape of fecal microbiota transplantation in treating inflammatory bowel disease. Transl Gastroenterol Hepatol 2025;10:55. [Crossref] [PubMed]
23. Jin Y, Dong H, Xia L, et al. The Diversity of Gut Microbiome is Associated With Favorable Responses to Anti-Programmed Death 1 Immunotherapy in Chinese Patients With NSCLC. J Thorac Oncol 2019;14:1378-89. [Crossref] [PubMed]
24. Zitvogel L, Ma Y, Raoult D, et al. The microbiome in cancer immunotherapy: Diagnostic tools and therapeutic strategies. Science 2018;359:1366-70. [Crossref] [PubMed]
25. He J, He X, Ma Y, et al. A comprehensive approach to stool donor screening for faecal microbiota transplantation in China. Microb Cell Fact 2021;20:216. [Crossref] [PubMed]
26. Qiao X, Biliński J, Wang L, et al. Safety and efficacy of fecal microbiota transplantation in the treatment of graft-versus-host disease. Bone Marrow Transplant 2023;58:10-9. [Crossref] [PubMed]
27. Wang JG, Liang Q, Dou HH, et al. The global incidence of adverse events associated with fecal microbiota transplantation in children over the past 20 years: A systematic review and meta-analysis. J Gastroenterol Hepatol 2022;37:2031-8. [Crossref] [PubMed]
28. Baxter M, Colville A. Adverse events in faecal microbiota transplant: a review of the literature. J Hosp Infect 2016;92:117-27. [Crossref] [PubMed]
29. Zhou C, Chen G, Huang Y, et al. Camrelizumab Plus Carboplatin and Pemetrexed as First-Line Treatment for Advanced Nonsquamous NSCLC: Extended Follow-Up of CameL Phase 3 Trial. J Thorac Oncol 2023;18:628-39. [Crossref] [PubMed]
30. Zhu J, Yu Y, Mei J, et al. Efficacy and safety of camrelizumab combined with albumin-bound paclitaxel as third- or later-line regimen in patients with advanced non-small cell lung cancer. Front Immunol 2023;14:1278573. [Crossref] [PubMed]
31. Watanabe H, Kubo T, Ninomiya K, et al. The effect and safety of immune checkpoint inhibitor rechallenge in non-small cell lung cancer. Jpn J Clin Oncol 2019;49:762-5. [Crossref] [PubMed]
32. Peng Z, Cheng S, Kou Y, et al. The Gut Microbiome Is Associated with Clinical Response to Anti-PD-1/PD-L1 Immunotherapy in Gastrointestinal Cancer. Cancer Immunol Res 2020;8:1251-61. [Crossref] [PubMed]
33. Pietrzak B, Tomela K, Olejnik-Schmidt A, et al. A Clinical Outcome of the Anti-PD-1 Therapy of Melanoma in Polish Patients Is Mediated by Population-Specific Gut Microbiome Composition. Cancers (Basel) 2022;14:5369. [Crossref] [PubMed]
34. Ouaknine Krief J, Helly de Tauriers P, Dumenil C, et al. Role of antibiotic use, plasma citrulline and blood microbiome in advanced non-small cell lung cancer patients treated with nivolumab. J Immunother Cancer 2019;7:176. [Crossref] [PubMed]
35. Tintelnot J, Xu Y, Lesker TR, et al. Microbiota-derived 3-IAA influences chemotherapy efficacy in pancreatic cancer. Nature 2023;615:168-74. [Crossref] [PubMed]
36. Pinato DJ, Howlett S, Ottaviani D, et al. Association of Prior Antibiotic Treatment With Survival and Response to Immune Checkpoint Inhibitor Therapy in Patients With Cancer. JAMA Oncol 2019;5:1774-8. [Crossref] [PubMed]
37. Pushalkar S, Hundeyin M, Daley D, et al. The Pancreatic Cancer Microbiome Promotes Oncogenesis by Induction of Innate and Adaptive Immune Suppression. Cancer Discov 2018;8:403-16. [Crossref] [PubMed]