Prognostic value of gut microbiota-derived metabolites in non-small cell lung cancer: a retrospective cohort study of overall survival

Introduction

Lung cancer remains the leading cause of cancer-associated mortality globally and constitutes a substantial global health burden, with non-small cell lung cancer (NSCLC) accounting for nearly 85% of all cases (1). Despite remarkable advances in immune checkpoint inhibition and molecular targeted therapy for NSCLC, the long-term survival of patients with NSCLC remains unsatisfactory, particularly for those with advanced-stage disease. The prognosis of NSCLC is affected by a multitude of tumor-derived and host-related factors, including tumor stage, performance status, metastatic tumor burden, treatment regimens, systemic inflammatory status, and nutritional reserve capacity. Although tumor stage is one of the best-established determinants of survival, patients with comparable clinicopathological profiles frequently display divergent clinical outcomes, implicating unrecognized host-derived biological drivers of prognostic heterogeneity. Accumulating data confirm that systemic inflammatory biomarkers carry clinically meaningful prognostic utility in NSCLC; however, the upstream biological regulators of these biomarkers and their crosstalk with metabolic dysregulation remain incompletely characterized (2). Accordingly, the identification of robust prognostic biomarkers that integrate both tumor biological features and host systemic status is essential to optimize risk stratification and guide the development of individualized therapeutic strategies for NSCLC (3).

Systemic inflammation and nutritional status are well-validated host-related prognostic determinants in patients with malignant tumors. The neutrophil-to-lymphocyte ratio (NLR) is a canonical systematic indicator that mirrors the balance between neutrophil-driven inflammatory activation and lymphocyte-mediated antitumor immune responses. Elevated NLR has been linked to adverse survival outcomes in NSCLC, underscoring that systemic inflammatory activation may contribute to tumor progression and dismal prognosis. In parallel, the prognostic nutritional index (PNI), calculated from serum albumin levels and peripheral lymphocyte counts, serves as a composite biomarker quantifying nutritional reserve and immune competence. Of note, a recent study demonstrated that inflammatory and nutritional indices, including NLR and PNI, were correlated with the clinical outcomes of patients with advanced NSCLC who were treated with immune checkpoint inhibitors (ICIs) (4). However, NLR and PNI predominantly reflect downstream alterations in host inflammatory and immunonutritional status and fail to capture the upstream microbial and metabolic perturbations that may fundamentally modulate systemic inflammation, antitumor immunity, and nutritional deterioration in NSCLC.

Beyond these inflammatory and nutritional indices, gut microbiota are also implicated in tumor initiation and progression. The intestinal microbial ecosystem sustains host metabolic homeostasis and modulates tumor development via complicated immune-inflammatory network interactions (5). Importantly, gut microbiota exert systemic effects predominantly by producing bioactive metabolites, among which bile acids (BAs) and short-chain fatty acids (SCFAs) stand out as key mediators of host-microbiota interactions (6,7). In the setting of lung cancer, the gut-lung axis functions as a critical biological framework that bridges intestinal microbial activity with pulmonary immunity, systemic inflammation, tumor immune regulation, and therapeutic responsiveness. Recent reviews have unveiled that gut microbial metabolites may affect the progression and immunotherapy efficacy of lung cancer via immune and metabolic pathways, supporting their potential as pivotal biomarkers and mechanistic mediators in NSCLC (8).

The gut microbiota can ferment dietary fiber to generate SCFAs, primarily including acetate, propionate, and butyrate, which play essential roles in regulating immune responses, preserving intestinal barrier integrity, and suppressing pro-inflammatory cascades (9). Mechanistically, SCFAs exert regulatory effects on tumor development through several mechanisms, such as induction of regulatory T cell differentiation, repression of histone deacetylase (HDAC) activity, and modulation of cellular metabolism (10,11). Butyrate is an SCFA of particular biological interest because of its potent immunomodulatory functions via HDAC inhibition, G-protein-coupled receptor signaling, and inflammatory pathway modulation. As reported, SCFAs exert modulatory effects on innate and adaptive immunity, including effects on T-cell differentiation, cytokine secretion, epithelial barrier integrity, and inflammatory signaling (12). The gut microbiota and their derived metabolites are increasingly recognized as important modulators of host immunity via the “gut-lung axis,” thereby regulating the progression and therapeutic responsiveness of lung cancer (13).

BAs represent another pivotal category of microbiota-related metabolites involved in the regulation of host metabolism and immune homeostasis. Primary BAs (PBAs) synthesized in the liver undergo extensive microbial biotransformation in the intestine to yield secondary BAs (SBAs). Acting as key signaling molecules, these metabolites activate nuclear and membrane receptors, such as the Farnesoid X receptor (FXR) and the G-protein-coupled BA receptor TGR5, therefore regulating metabolic homeostasis, inflammation, and immune responses (14). Dysregulated BA metabolism has been implicated in the pathogenesis of multiple malignancies, such as hepatocellular carcinoma, colorectal cancer, and lung cancer (15). Moreover, BAs may orchestrate tumor progression by modulating immune cell activity and remodeling the inflammatory tumor microenvironment. Importantly, SBAs, including deoxycholic acid (DCA) and lithocholic acid (LCA), may exert context-dependent immunomodulatory effects that potentially facilitate pro-tumorigenic phenotypes. It has been recently evidenced that microbiota-derived BAs drive the growth of colorectal cancer by impairing the effector function of CD8+ T-cells, reinforcing the notion that elevated SBAs attenuate antitumor immune responses and foster an immunosuppressive tumor microenvironment (16). A recent mechanistic review further emphasized the intricate crosstalk among BAs, gut microbiota, and tumor immune microenvironment (17). While these biological mechanisms have not been fully clarified in NSCLC, they provide a robust biological rationale for analyzing the association between SBA-related metabolites and dismal survival outcomes.

Despite mounting evidence linking systemic inflammation, nutritional status, and gut microbiota-derived metabolites to cancer prognosis, several critical research gaps still exist. First, prior studies have evaluated inflammatory markers, nutritional indices, and microbial metabolites in isolation, rather than comprehensively exploring their interactive and synergistic prognostic values within the same clinical framework. Second, it remains poorly elucidated about the potential association among butyrate depletion, SBA enrichment, systemic inflammatory activation, and impaired immunonutritional homeostasis in NSCLC. Third, although the prognostic implication of the gut-lung axis has been increasingly acknowledged, the prognostic significance of specific microbiota-derived metabolites in relation to overall survival (OS) have been largely understudied in NSCLC patients.

Against this background, the present study characterized the profiles of SCFAs and BAs in NSCLC patients and evaluated their correlations with systemic inflammation and host nutritional status. Furthermore, with OS as the primary clinical endpoint, this study further ascertained the prognostic significance of gut microbiota-derived metabolites, with a particular focus on butyrate and SBA-related indicators, and explored their potential associations with NLR and PNI, which serve as indicators of systemic inflammation and immunonutritional status, respectively. We present this article in accordance with the STROBE reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0681/rc).

Methods

Study design

This single-center retrospective cohort study enrolled 256 consecutive patients with pathologically confirmed NSCLC who were diagnosed and treated at our institution from January 2019 to April 2024. These patients were followed up until April 2025. OS was defined as the time from baseline sampling to death from any cause or the last follow-up. Patients who survived until the last follow-up were defined as censored cases. For baseline and clinical characteristic comparisons, all enrolled patients were stratified into long-OS (patients with OS greater than or equal to the mean) and non-long-OS (patients with OS shorter than the mean) cohorts according to the mean OS. A total of 342 consecutive patients with pathologically confirmed NSCLC were initially screened for enrollment during the study period. Of these, 86 patients were excluded per the predefined exclusion criteria: 24 patients with antibiotic use within 4 weeks before baseline sampling, 15 patients with inflammatory bowel disease or severe gastrointestinal disorders, 13 patients with significant hepatobiliary disease, 12 patients with acute infections or severe inflammatory conditions, and 22 patients with missing key clinical data or incomplete follow-up information. Ultimately, 256 eligible patients were included in the final analysis. This retrospective study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study protocol was approved by the Human Ethics Committee of Affiliated Jinling Hospital, Medical School of Nanjing University (Ethics Approval Number: DZQH-KYLL-25-08). Given the retrospective nature of the study and the use of fully anonymized clinical data, the requirement for informed consent was waived.

Inclusion criteria: patients who met all of the following criteria were enrolled in the present study: (I) age of at least 18 years; (II) pathologically confirmed diagnosis of NSCLC; (III) complete baseline clinical data; (IV) valid baseline fecal samples collected for microbial metabolite analysis; (V) accessible and complete follow-up information.

Exclusion criteria: individuals presenting with any of the following conditions were excluded: (I) antibiotic administration within 4 weeks prior to baseline sampling; (II) a history of inflammatory bowel disease or severe gastrointestinal disorders; (III) clinically significant hepatobiliary disease; (IV) acute infections or severe inflammatory conditions; (V) missing key clinical data.

Data collection

For all enrolled patients, baseline clinical information was retrieved, including demographic details (age and sex), body mass index (BMI), and smoking status through standardized review of electronic medical records. Tumor-related variables were also recorded, including the Eastern Cooperative Oncology Group (ECOG) performance status, histological subtype, tumor-node-metastasis (TNM) stage, presence of liver metastasis, and treatment regimens. Baseline hematological parameters were measured through routine laboratory tests before the initiation of antitumor treatment to evaluate systemic inflammation and nutritional status. Subsequently, NLR was computed as the ratio of peripheral absolute neutrophil count to absolute lymphocyte count. PNI was utilized to assess the nutritional and immune status of patients, which was calculated with the Onodera formula: PNI = 10 × albumin (g/dL) + 0.005 × lymphocyte count (/mm³) (18), where serum albumin mirrors nutritional status and lymphocyte count underpins host immune function. Treatment-related information was obtained from the electronic medical records. All patients were stratified according to their primary therapeutic regimens, including chemotherapy, surgery, immunotherapy, targeted therapy, and radiotherapy. For patients receiving immunotherapy, the type of ICIs, including anti-PD-1 and anti-PD-L1 agents, was recorded when clinical data were available. Combination therapeutic strategies, particularly chemo-immunotherapy, were also documented.

The concentrations of SCFAs and BAs were measured. The analyzed SCFA profiling encompassed acetate, propionate, and butyrate, with total SCFAs levels calculated for subsequent analysis. BA detection covered PBAs [cholic acid (CA) and chenodeoxycholic acid (CDCA)] and SBAs (DCA and LCA). Subsequently, the total concentration of SBA and the ratio of SBA/PBA were calculated.

Follow-up information was attained via regular outpatient reexaminations, inpatient medical record review, telephone follow-up, and review of the institutional electronic medical record system. Survival status and exact death dates were validated against archived medical records or telephone follow-up records. The cutoff date for final follow-up was April 2025. Patients who remained alive without documented death at the final follow-up date were censored at the date of last confirmed contact. Any patients lost to follow-up were censored at the time of their last retrievable follow-up documentation.

Fecal sample collection and metabolite measurement

Baseline fecal samples were obtained prior to the initiation of any systemic anticancer therapy under a standardized protocol for gut microbiome/metabolite sample acquisition established at our center, which was formulated by referring to previously published recommendations and standardized procedures for fecal sampling and metabolomic analysis (19,20). All participants provided fresh fecal samples using sterile fecal collection containers, and the samples were processed in strict accordance with laboratory standard operating procedures to avoid exogenous contamination and metabolite degradation and were properly preserved until subsequent metabolite analysis.

SCFAs and BAs were measured using baseline fecal samples. For metabolite extraction, fecal aliquots were precisely weighed and homogenized in extraction solvent. After vortexing and centrifugation, the resultant supernatant was harvested for targeted metabolite analysis. The concentrations of acetate, propionate, and butyrate were quantitatively determined using an Agilent 7890B gas chromatography system coupled to an Agilent 5977B mass-selective detector (Agilent Technologies, Santa Clara, CA, USA) as per a previously described gas chromatography-mass spectrometry-based method with minor modifications (21). The concentrations of CA, CDCA, DCA, and LCA were detected using a Waters ACQUITY ultra-performance liquid chromatography system coupled to a Waters Xevo TQ-S micro triple-quadrupole mass spectrometer (Waters Corporation, Milford, MA, USA) based on validated targeted LC-MS/MS metabolomic protocols (22). The SBA/PBA ratio was calculated as (DCA + LCA)/(CA + CDCA) × 100%. Metabolite concentrations were normalized to the weight of fecal samples.

Statistical analysis

Statistical analysis was performed with R version 4.5.2 and IBM SPSS Statistics version 26.0 (IBM Corp., Armonk, NY, USA). All tests were two-sided, and statistical significance was set at P<0.05. The study outcome was defined as a binary variable (alive =0, death =1). Categorical variables, which were summarized as counts (percentages), were compared between groups with the chi-square test or Fisher’s exact test as appropriate. Continuous variables were displayed as mean ± standard deviation for normally distributed data or median (interquartile range) for non-normally distributed data; between-group comparisons were implemented with the independent samples t-test or the Mann-Whitney U test accordingly.

Differences in the concentrations of SCFAs and BA-related metabolites between groups were analyzed. Spearman or Pearson correlation analysis was performed to assess the associations between gut microbial metabolite concentrations and systemic inflammatory-nutritional biomarkers (NLR and PNI).

Considering the retrospective design of this study, no formal a priori sample size calculation was conducted. The final sample size was determined based on all consecutive, eligible patients with pathologically confirmed NSCLC who received diagnosis and treatment during the predefined study timeframe and had complete baseline clinical data, available baseline fecal specimens, and complete follow-up records.

In total, 256 patients were included in this study, and 100 deaths were observed during follow-up. Following the restriction of the number of covariates and elimination of redundancy among correlated variables, the observed number of clinical outcome events was adequate to support the primary multivariable Cox regression analysis. Nevertheless, given that the study was not prospectively powered for subgroup or interaction analysis, the findings of these exploratory subgroup comparisons should be interpreted with caution and warrant further validation in large-scale prospective cohorts.

Cox proportional hazards regression models were constructed to evaluate the associations of OS with clinical variables, inflammatory and nutritional indicators, and gut microbiota-derived metabolites. All clinically meaningful variables and OS-associated variables in the univariate analysis were enrolled for subsequent multivariate Cox regression analysis. Variance inflation factor (VIF) was calculated to assess multicollinearity among candidate variables. In the case of significant multicollinearity, representative indicators were retained to eliminate statistical redundancy (specifically, butyrate was selected as the representative marker of SCFAs, the SBA/PBA ratio was adopted for among BA indicators, and tumor stage was prioritized, instead of including liver metastasis simultaneously). The Cox proportional hazards regression model was then constructed with the selected variables, and the results were presented as hazard ratios (HRs) and corresponding 95% confidence intervals (95% CIs).

Kaplan-Meier survival curves were generated based on patient stratification by key metabolic indicators, including butyrate, the SBA/PBA ratio, and the integrated metabolic profile. Survival differences were compared with the log-rank test.

Results

Baseline clinical characteristics of the two study cohorts

In total, 256 eligible NSCLC patients were included in this study, with 112 patients allocated to the long-OS cohort and 144 patients assigned to the non-long-OS cohort. Most baseline demographic and clinical characteristics were comparable between the two cohorts, including age, sex, BMI, smoking history, ECOG performance status, histological type, liver metastasis, and treatment regimens (Table 1).

Conversely, significant differences were observed between the two cohorts in tumor stage and inflammatory and nutritional indicators. Compared with the cohort of long-OS patients, the cohort of non-long-OS patients had a higher proportion of patients with advanced stages III-IV [101/144 (70.14%) vs. 65/112 (58.04%), χ²=4.048, P=0.04], higher NLR levels [3.09 (2.09, 4.44) vs. 2.38 (1.94, 3.01), Z=−4.285, P<0.001], and higher neutrophil counts [4.29 (3.53, 5.26) vs. 3.84 (3.19, 4.30), Z=−3.524, P<0.001].

In contrast, compared to the cohort of long-OS patients, the cohort of non-long-OS patients exhibited markedly lower PNI (45.40±6.21 vs. 48.48±4.97, t=4.402, P<0.001), serum albumin levels (38.35±5.16 vs. 40.59±4.47, t=3.654, P<0.001), and lymphocyte counts (1.41±0.43 vs. 1.58±0.38, t=3.220, P=0.001).

Comparisons of gut microbiota-derived metabolites between the two study cohorts

The concentrations of several SCFA-related metabolites were higher in the cohort of long-OS patients than in the cohort of non-long-OS patients. Specifically, the concentrations of propionate and butyrate and the total concentrations of SCFAs were significantly higher in the cohort of long-OS patients than in the cohort of non-long-OS patients [13.86±4.87 vs. 12.37±4.78, P=0.02; 9.81 (7.43, 11.97) vs. 7.85 (5.84, 9.89), P<0.001; 65.27±13.16 vs. 60.74±12.51, P=0.005, respectively]. Meanwhile, the concentration of acetate was insignificantly higher in the cohort of long-OS patients than in the cohort of non-long-OS patients (39.90±11.62 vs. 38.40±10.44, P=0.28).

On the contrary, SBA-related indicators, including DCA, LCA, and the SBA/PBA ratio, were markedly elevated in the cohort of non-long-OS patients compared with the cohort of long-OS patients [1.33 (0.91, 1.72) vs. 1.12 (0.85, 1.51), P=0.04; 0.59 (0.37, 0.83) vs. 0.50 (0.30, 0.67), P=0.003; 1.32 (0.89, 1.73) vs. 1.07 (0.73, 1.45), P=0.007, respectively]. Nevertheless, the concentrations of CA and CDCA (PBAs) did not differ substantially between the two study cohorts (both P>0.05; Table 2).

Correlations between metabolites and inflammatory/nutritional indicators

Significant correlations were observed between the inflammatory marker NLR and gut microbiota-derived metabolites. Specifically, NLR was negatively correlated with butyrate (r=−0.241, P<0.001) and total SCFAs (r=−0.241, P<0.001) but positively correlated with DCA (r=0.191, P=0.002), LCA (r=0.203, P=0.001), and the SBA/PBA ratio (r=0.181, P=0.004).

Marked correlations also existed between the nutritional indicator PNI and gut microbiota-derived metabolites. In detail, PNI displayed positive correlations with butyrate (r=0.298, P<0.001) and total SCFAs (r=0.213, P<0.001) but negative correlations with DCA (r=−0.188, P=0.003), LCA (r=−0.193, P=0.002), and the SBA/PBA ratio (r=−0.140, P=0.03) (Table 3 and Figure 1).

Figure 1 Correlation analysis between gut microbiota-derived metabolites and inflammatory/nutritional indicators. (A) Correlations of butyrate and the SBA/PBA ratio with the inflammatory indicator NLR and the nutritional indicator PNI. (B) Heatmap displaying the correlations among metabolites, inflammatory markers, and nutritional indicators. Colors indicate the strength and direction of the correlation coefficients. CA, cholic acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; LCA, lithocholic acid; NLR, neutrophil-to-lymphocyte ratio; PNI, prognostic nutritional index; SBA/PBA, secondary-to-primary bile acid ratio; SCFAs, short-chain fatty acids.

Cox regression analysis for OS

As observed in the results of the univariate Cox regression analysis, OS was markedly associated with age, ECOG performance status, advanced tumor stage, liver metastasis, NLR, PNI, and several gut microbiota-derived metabolites. Among SCFA-related metabolites, acetate, propionate, butyrate, and total SCFAs were associated with a reduced risk of death. In contrast, DCA, LCA, and the SBA/PBA ratio were associated with an increased risk of death.

Considering clinical relevance and potential multicollinearity, tumor stage, treatment regimens, NLR, PNI, butyrate, and the SBA/PBA ratio, alongside selected clinical covariates, were included into the multivariate Cox regression model for later analysis. Following adjustment, advanced tumor stage remained significantly associated with inferior OS (HR =2.612, 95% CI: 1.284–5.314, P=0.008). Higher NLR was also independently associated with an elevated risk of death (HR =1.502, 95% CI: 1.327–1.700, P<0.001), whereas higher PNI was associated with a decreased risk of death (HR =0.894, 95% CI: 0.858–0.932, P<0.001).

Importantly, butyrate, after multivariable adjustment, remained independently associated with favorable OS (HR =0.775, 95% CI: 0.716–0.838, P<0.001), whilst a higher SBA/PBA ratio was independently associated with dismal OS (HR =1.794, 95% CI: 1.453–2.215, P<0.001). Immunotherapy-containing and targeted therapy-based treatment regimens showed a trend toward prolonged survival when compared with chemotherapy-based treatment regimens; however, these associations failed to reach statistical significance in the multivariate regression model (Table 4).

Kaplan-Meier survival analysis

The results of the Kaplan-Meier survival analysis revealed significant survival differences under metabolite-based stratification. Patients with high butyrate levels had favorable OS than those with low butyrate levels (log-rank P<0.001; median OS: not reached vs. 26.58 months; Figure 2A). The results demonstrated that patients with a favorable metabolic profile exhibited superior survival outcomes, while patients with an unfavorable metabolic profile displayed worse survival outcomes (log-rank P<0.001; median OS: not reached, 67.84 months, and 13.63 months for favorable, intermediate, and unfavorable profiles, respectively; Figure 2B). Conversely, patients with a high SBA/PBA ratio showed poorer OS than those with a low SBA/PBA ratio (log-rank P<0.001; median OS: 44.78 months vs. not reached; Figure 2C). The integrated metabolic profile further stratified patients into distinct prognostic subgroups.

Figure 2 Kaplan-Meier survival curves according to gut microbiota-derived metabolite profiles. (A) Overall survival stratified by butyrate levels according to the cohort median. Patients with high butyrate levels showed significantly superior overall survival than those with low butyrate levels. (B) Overall survival according to the integrated metabolic profile based on butyrate and the SBA/PBA ratio. (C) Overall survival stratified by the SBA/PBA ratio according to the cohort median. Patients with a high SBA/PBA ratio exhibited markedly dismal overall survival. The favorable profile was defined as high butyrate levels plus a low SBA/PBA ratio, the unfavorable profile as low butyrate levels plus a high SBA/PBA ratio, and the intermediate profile as all other combinations. Shaded areas correspond to 95% confidence intervals. Censored observations are marked on the curves. P values were calculated with the log-rank test. SBA/PBA, secondary-to-primary bile acid ratio.

Discussion

In the present study, gut microbiota-derived metabolites were markedly correlated with OS in patients with NSCLC. Compared with the cohort of long-OS patients, the cohort of non-long-OS patients presented with lower concentrations of propionate, butyrate, and total SCFAs, accompanied by higher levels of SBA-related metabolites, including DCA, LCA, and the SBA/PBA ratio. Butyrate was correlated negatively with NLR and positively with PNI, whereas SBA-related metabolites showed opposite correlations. The multivariate Cox regression analysis further identified advanced tumor stage, NLR, PNI, butyrate, and the SBA/PBA ratio as independent factors associated with OS in NSCLC patients. Notably, the integrated metabolic profile of butyrate and the SBA/PBA ratio showed marked survival disparities in the Kaplan-Meier survival analysis, highlighting the potential involvement of gut microbiota-associated metabolic dysregulation in the prognosis of NSCLC via interactions among inflammation, immunity, and nutritional status.

The prognostic implication of fecal butyrate in NSCLC constitutes a key finding of the present study. Accumulating evidence has demonstrated the significant depletion of butyrate-producing bacteria in patients with lung cancer, indicating impaired butyrate synthesis as a hallmark of lung cancer-associated gut microbial dysbiosis. For example, Gui et al. found that the relative abundance of several typical butyrate-producing bacteria, including Faecalibacterium prausnitzii and Roseburia, were reduced in patients with NSCLC, which may provide a plausible microbial ecological mechanism underlying the diminished butyrate levels observed in the cohort of non-long-OS patients in the present study (23). As a key microbial metabolite, butyrate exerts multifaceted biological effects. Beyond serving as a primary energy substrate for colon epithelial cells, butyrate exerts multifaceted effects on host physiology, including maintenance of intestinal barrier integrity, modulation of immune responses, and suppression of inflammatory pathways (24,25). Mechanistically, butyrate can modulate gene expression via epigenetic mechanisms, particularly via HDAC inhibition. By remodeling chromatin structure and transcriptional activity, HDAC inhibition can orchestrate fundamental cellular processes, such as cell proliferation, differentiation, and apoptosis, ultimately facilitating antitumor immune responses and restraining malignant progression (26). Preclinical experimental studies have reported that butyrate reduces HDAC activity and induces cell-cycle arrest and apoptosis in cancer cells via epigenetic regulation of downstream gene expression.

Notably, butyrate is tightly correlated with systemic inflammatory and immunonutritional conditions. The present study elucidated that butyrate was negatively correlated with NLR and positively correlated with PNI. Typically, elevated NLR indicates pro-tumor inflammatory activation, myeloid cell predominance, and relative lymphocyte suppression, whereas reduced PNI reflects decreased albumin levels and insufficient immune reserve capacity. Both markers have frequently been linked to the treatment responses and survival outcomes of NSCLC patients.

Meanwhile, accumulating evidence has unraveled that gut microbiota-derived SCFAs modulate host immune responses and systemic inflammatory status through multiple regulatory mechanisms, including HDAC inhibition, G-protein-coupled receptor activation, and inflammatory pathway modulation (12,27). Collectively, these findings support that butyrate represents a pivotal metabolic mediator that bridges gut microbial activity with host inflammatory and immunonutritional status in NSCLC patients.

In contrast to the protective effect of SCFAs, elevated SBAs represent an adverse metabolic feature associated with poor prognosis. In our cohort, the levels of SBAs, including DCA, LCA, and the SBA/PBA ratio, were substantially higher in the cohort of non-long-OS patients, while no discernible differences were found between the two cohort regarding the concentrations of PBAs (CA and CDCA), underscoring that poor outcomes are not attributed to alterations in total BA levels, but rather to BA transformation altered by gut microbial metabolism, leading to the accumulation of SBAs.

BAs function not only as the products of lipid metabolism but also as critical signaling molecules that orchestrate immune responses, intestinal barrier homeostasis, and metabolic processes via receptor-mediated pathways, such as FXR and TGR5 (17,28,29). Notably, SBAs, including LCA, have been elaborated to impair effector T-cell function in certain experimental models, pointing to their potential in tumor-promoting immunosuppression (30). Such mechanisms partially explain the association between elevated SBAs and dismal outcomes of NSCLC patients observed in this study.

Another key strength of the present study lies in the integration of gut microbiota-derived metabolites with routine clinical indicators within a unified prognostic analytical framework. Traditional prognostic evaluation for NSCLC mainly depends on tumor stage, performance status, and certain hematological indicators. Of note, our results unveiled that after adjustment for tumor stage, NLR, and PNI, butyrate and the SBA/PBA ratio remained independently associated with OS, suggesting that microbial metabolites may offer complementary prognostic information beyond conventional clinical indicators in NSCLC patients.

Furthermore, our multivariate Cox regression model exhibited that butyrate was still independently associated with favorable OS in NSCLC patients, while a higher SBA/PBA ratio was independently associated with poorer OS. The Kaplan-Meier analysis further showed that the integrated metabolic profile of butyrate and the SBA/PBA ratio was associated with diverse OS patterns. These findings demonstrate that integrating microbial metabolites with host inflammatory and nutritional indicators may better recapitulate the complex biological landscape underlying NSCLC prognosis compared with single markers alone. It has been growingly evidenced that cancer prognosis is determined by complicated crosstalk among tumor biology, host immunity, metabolism, and the tumor microenvironment (31,32). Accordingly, integrated biomarker profiles may offer a complementary strategy for OS-related risk stratification in NSCLC. Nevertheless, external validation is required to verify the generalizability of the findings.

Our findings are consistent with and further expand existing evidence regarding the prognostic and functional significance of gut microbiota-derived metabolites in human malignancies. SCFAs, particularly butyrate, have been extensively studied in colorectal cancer and other malignancies, where they generally exert protective effects by preserving epithelial barrier integrity, repressing HDAC, and regulating inflammatory signaling and antitumor immune responses. Beyond its anti-inflammatory and HDAC-inhibitory properties, butyrate also directly modulates cytotoxic CD8+ T-cell function. He et al. (33) reported that gut microbial metabolites, especially butyrate, enhanced the efficacy of anticancer therapy by modulating CD8+ T-cell function in the tumor microenvironment. Similarly, Zhu et al. (34) found that butyrate supplementation potentiated the antitumor efficacy of anti-PD-1 treatment by augmenting antitumor cytokine expression and regulating T-cell receptor signaling in cytotoxic CD8+ T cells. More recent studies further displayed that gut microbiota-derived metabolites were associated with the long-term efficacy of ICIs and the incidence of immune-related adverse events in patients with lung cancer (35), and that gut microbiota-derived butyrate sustained CD8+ T-cell immune competence via a FOXO1-related stemness regulatory program (36). These findings provide a plausible biological explanation for our key observation that elevated fecal butyrate levels were independently associated with prolonged OS in patients with NSCLC. Butyrate also showed a favorable trend in patients receiving immunotherapy-containing treatment regimens. Given the limited sample size of this subgroup analysis, however, this finding should be interpreted with caution.

Conversely, SBAs exhibit context-dependent and potentially tumor-promoting functions in NSCLC progression. BAs are extensively modified by the gut microbiota and participate in metabolic regulation, inflammation, immune homeostasis, and neoplastic diseases as signaling molecules (28). Additionally, the dysregulation of the gut microbiota-BA axis has been implicated in the initiation and progression of hepatocellular carcinoma through chronic inflammation, BA homeostasis disruption, intestinal barrier dysfunction, and immune microenvironment remodeling (37,38). Similar disturbances in the gut–liver axis have also been detected in cholangiocarcinoma, wherein gut microbiota dysbiosis, aberrant BA metabolism, and immune microenvironment remodeling can drive tumor progression (39). An experimental study on colorectal cancer elucidated that gut microbiota-modified BAs facilitated tumor growth by impairing the effector function of CD8+ T-cells (16). In the present study, a higher SBA/PBA ratio was independently associated with poorer OS in NSCLC, illustrating that enhanced microbial conversion of PBAs to SBAs may reflect a dysregulated systemic immune-metabolic state associated with poor prognosis. While direct mechanistic evidence linking BA metabolism to tumor immunity in NSCLC remains scarce, our results highlight that BA-related immune-metabolic dysregulation also holds critical prognostic implications in lung cancer.

From the perspective of translational medicine, our findings may have potential clinical relevance, although they are still in the exploratory stage. First, fecal detection of gut microbiota-derived metabolites is a minimally invasive and repeatable approach, rendering it a promising approach for dynamic risk monitoring in patients with NSCLC. Second, microbial metabolic indicators, such as butyrate and the SBA/PBA ratio, can be integrated with conventional inflammatory and immunonutritional biomarkers, including NLR and PNI, to refine and optimize OS-related risk stratification in future studies. Third, pending future validation of causal relationships, microbiota-targeted interventions warrant further investigation as a viable adjunct therapeutic strategy for NSCLC management. Potential intervention modalities encompass dietary fiber supplementation, prebiotic or postbiotic administration, targeted microbial modulation, and fecal microbiota transplantation. Emerging evidence demonstrates that gut microbial composition and SCFA production may modulate responses to anticancer immunotherapy, paving the way for ongoing translational studies of microbiome-based therapeutic strategies (40-42).

Of course, several limitations should be acknowledged. First, this study was a single-center retrospective cohort study, which may incur inherent selection bias and restrict the generalizability of our findings. Although OS was analyzed with time-to-event methods, external validation in large-scale, multicenter prospective cohorts is still required to confirm the reliability and universality of our results. Second, fecal metabolite profiling was performed solely at baseline, which precluded the assessment of dynamic microbiome-metabolome changes during disease progression or treatment. Therefore, it remains uncertain whether these metabolic changes drive tumor progression or merely reflect systemic conditions, dietary variation, or microbial dysbiosis. Third, although patients with recent antibiotic use and overt hepatobiliary diseases were excluded in this cohort, fecal metabolite profiles may still be affected by unanalyzed confounding factors, including diet, proton pump inhibitors, corticosteroids, concomitant medications, and treatment heterogeneity. Despite the presence of treatment-stratified analysis and adjustment for treatment regimens in the Cox regression model, subgroup analysis was limited by reduced sample sizes, particularly in radiotherapy-based subgroups. Fourth, this study lacked direct tumor immune profiling, peripheral T-cell functional assessment, and longitudinal microbiome-metabolome monitoring. Therefore, the proposed associations of butyrate with CD8+ T cell-mediated antitumor immunity, ICI responsiveness, and survival remain exploratory and hypothesis-generating, requiring further validation in prospective mechanistic studies. As a consequence, future studies should leverage multicenter prospective cohorts and integrate metagenomic, metabolomic, and immunophenotypic data to rigorously validate the prognostic value of combination of butyrate and the SBA/PBA ratio. In addition, mechanistic studies and clinical trials focusing on microbiota-targeted interventions are merited to clarify the potential therapeutic relevance of these immune-metabolic pathways in NSCLC.

Conclusions

In this retrospective NSCLC cohort, poor OS was markedly associated with a fecal metabolic phenotype characterized by reduced SCFAs, particularly downregulated butyrate, and increased SBA-related indicators. These metabolic perturbations were coupled with promoted systemic inflammation and compromised immunonutritional status, and butyrate and the SBA/PBA ratio remained independently associated with OS following adjustment for selected clinical factors. Prospective multicenter studies are necessary for validating the clinical relevance and potential utility of butyrate and the SBA/PBA ratio for OS-related risk stratification.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0681/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0681/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0681/prf

Funding: This work was supported by Key Technologies and Applications of Gut Microbiota-Based Personalized Monitoring for Lung Cancer in the Elderly (No. K2024008).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-0681/coif). The 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 Declaration of Helsinki and its subsequent amendments. The study was approved by the Human Ethics Committee of Affiliated Jinling Hospital, Medical School of Nanjing University (Ethics Approval Number: DZQH-KYLL-25-08). Given the retrospective nature of the study and the use of de-identified clinical data, the requirement for informed consent was waived.

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. Thai AA, Solomon BJ, Sequist LV, Gainor JF, Heist RS. Lung cancer. Lancet 2021;398:535-54. [Crossref] [PubMed]
2. Stares M, Brown LR, Abhi D, et al. Prognostic Biomarkers of Systemic Inflammation in Non-Small Cell Lung Cancer: A Narrative Review of Challenges and Opportunities. Cancers (Basel) 2024;16:1508. [Crossref] [PubMed]
3. Tang C, Li H, Guo G, et al. Stability evaluation method of gasification coal pillar under thermal coupling condition for prevention of environment secondary pollution. Sci Total Environ 2024;954:176265. [Crossref] [PubMed]
4. Fang Q, Yu J, Li W, et al. Prognostic value of inflammatory and nutritional indexes among advanced NSCLC patients receiving PD-1 inhibitor therapy. Clin Exp Pharmacol Physiol 2023;50:178-90. [Crossref] [PubMed]
5. Fan Y, Pedersen O. Gut microbiota in human metabolic health and disease. Nat Rev Microbiol 2021;19:55-71. [Crossref] [PubMed]
6. Lavelle A, Sokol H. Gut microbiota-derived metabolites as key actors in inflammatory bowel disease. Nat Rev Gastroenterol Hepatol 2020;17:223-37. [Crossref] [PubMed]
7. Yu X, Yu X, Wang Y, et al. Respiratory diseases and the gut microbiota: an updated review. Front Cell Infect Microbiol 2025;15:1629005. [Crossref] [PubMed]
8. Li X, Shang S, Wu M, et al. Gut microbial metabolites in lung cancer development and immunotherapy: Novel insights into gut-lung axis. Cancer Lett 2024;598:217096. [Crossref] [PubMed]
9. Silva YP, Bernardi A, Frozza RL. The Role of Short-Chain Fatty Acids From Gut Microbiota in Gut-Brain Communication. Front Endocrinol (Lausanne) 2020;11:25. [Crossref] [PubMed]
10. Parada Venegas D, De la Fuente MK, Landskron G, et al. Short Chain Fatty Acids (SCFAs)-Mediated Gut Epithelial and Immune Regulation and Its Relevance for Inflammatory Bowel Diseases. Front Immunol 2019;10:277. [Crossref] [PubMed]
11. Kotlyarov S. Role of Short-Chain Fatty Acids Produced by Gut Microbiota in Innate Lung Immunity and Pathogenesis of the Heterogeneous Course of Chronic Obstructive Pulmonary Disease. Int J Mol Sci 2022;23:4768. [Crossref] [PubMed]
12. Liu XF, Shao JH, Liao YT, et al. Regulation of short-chain fatty acids in the immune system. Front Immunol 2023;14:1186892. [Crossref] [PubMed]
13. Dang AT, Marsland BJ. Microbes, metabolites, and the gut-lung axis. Mucosal Immunol 2019;12:843-50. [Crossref] [PubMed]
14. Jia W, Xie G, Jia W. Bile acid-microbiota crosstalk in gastrointestinal inflammation and carcinogenesis. Nat Rev Gastroenterol Hepatol 2018;15:111-28. [Crossref] [PubMed]
15. Deng X, Xiao L, Luo M, et al. Intestinal crosstalk between bile acids and microbiota in irritable bowel syndrome. J Gastroenterol Hepatol 2023;38:1072-82. [Crossref] [PubMed]
16. Cong J, Liu P, Han Z, et al. Bile acids modified by the intestinal microbiota promote colorectal cancer growth by suppressing CD8(+) T cell effector functions. Immunity 2024;57:876-889.e11. [Crossref] [PubMed]
17. Tong Y, Lou X. Interplay between bile acids, gut microbiota, and the tumor immune microenvironment: mechanistic insights and therapeutic strategies. Front Immunol 2025;16:1638352. [Crossref] [PubMed]
18. Jian-Hui C, Iskandar EA, Cai ShI, et al. Significance of Onodera's prognostic nutritional index in patients with colorectal cancer: a large cohort study in a single Chinese institution. Tumour Biol 2016;37:3277-83. [Crossref] [PubMed]
19. Hosseinkhani F, Dubbelman AC, Karu N, et al. Towards Standards for Human Fecal Sample Preparation in Targeted and Untargeted LC-HRMS Studies. Metabolites 2021;11:364. [Crossref] [PubMed]
20. Mirzayi C, Renson A, et al. Reporting guidelines for human microbiome research: the STORMS checklist. Nat Med 2021;27:1885-92. [Crossref] [PubMed]
21. Ren C, Bailey AO, VanderPorten E, et al. Quantitative Determination of Protein-Ligand Affinity by Size Exclusion Chromatography Directly Coupled to High-Resolution Native Mass Spectrometry. Anal Chem 2019;91:903-11. [Crossref] [PubMed]
22. Kakiyama G, Muto A, Takei H, et al. A simple and accurate HPLC method for fecal bile acid profile in healthy and cirrhotic subjects: validation by GC-MS and LC-MS. J Lipid Res 2014;55:978-90. [Crossref] [PubMed]
23. Gui Q, Li H, Wang A, et al. The association between gut butyrate-producing bacteria and non-small-cell lung cancer. J Clin Lab Anal 2020;34:e23318. [Crossref] [PubMed]
24. Sun J, Chen S, Zang D, et al. Butyrate as a promising therapeutic target in cancer: From pathogenesis to clinic Int J Oncol 2024;64:44. (Review). [Crossref] [PubMed]
25. Georgiou K, Marinov B, Farooqi AA, et al. Gut Microbiota in Lung Cancer: Where Do We Stand? Int J Mol Sci 2021;22:10429. [Crossref] [PubMed]
26. Hajjar R, Richard CS, Santos MM. The role of butyrate in surgical and oncological outcomes in colorectal cancer. Am J Physiol Gastrointest Liver Physiol 2021;320:G601-8. [Crossref] [PubMed]
27. Du Y, He C, An Y, et al. The Role of Short Chain Fatty Acids in Inflammation and Body Health. Int J Mol Sci 2024;25:7379. [Crossref] [PubMed]
28. Fleishman JS, Kumar S. Bile acid metabolism and signaling in health and disease: molecular mechanisms and therapeutic targets. Signal Transduct Target Ther 2024;9:97. [Crossref] [PubMed]
29. Luo W, Li R, Pan C, et al. Gut microbiota-derived metabolites in immunomodulation and gastrointestinal cancer immunotherapy. Front Immunol 2025;16:1710880. [Crossref] [PubMed]
30. Perl M, Fante MA, Herfeld K, et al. Microbiota-derived metabolites: Key modulators of cancer immunotherapies. Med 2025;6:100773. [Crossref] [PubMed]
31. Lei W, Zhou K, Lei Y, et al. Gut microbiota shapes cancer immunotherapy responses. NPJ Biofilms Microbiomes 2025;11:143. [Crossref] [PubMed]
32. Zhang M, Liu J, Xia Q. Role of gut microbiome in cancer immunotherapy: from predictive biomarker to therapeutic target. Exp Hematol Oncol 2023;12:84. [Crossref] [PubMed]
33. He Y, Fu L, Li Y, et al. Gut microbial metabolites facilitate anticancer therapy efficacy by modulating cytotoxic CD8(+) T cell immunity. Cell Metab 2021;33:988-1000.e7. [Crossref] [PubMed]
34. Zhu X, Li K, Liu G, et al. Microbial metabolite butyrate promotes anti-PD-1 antitumor efficacy by modulating T cell receptor signaling of cytotoxic CD8 T cell. Gut Microbes 2023;15:2249143. [Crossref] [PubMed]
35. Liu X, Lu B, Tang H, et al. Gut microbiome metabolites, molecular mimicry, and species-level variation drive long-term efficacy and adverse event outcomes in lung cancer survivors. EBioMedicine 2024;109:105427. [Crossref] [PubMed]
36. Bachem A, Clarke M, Kong G, et al. Microbiota-derived butyrate promotes a FOXO1-induced stemness program and preserves CD8(+) T cell immunity against melanoma. Immunity 2025;58:2799-2813.e8. [Crossref] [PubMed]
37. Song Y, Lau HC, Zhang X, et al. Bile acids, gut microbiota, and therapeutic insights in hepatocellular carcinoma. Cancer Biol Med 2023;21:144-62. [Crossref] [PubMed]
38. Zuo S, Ma J, Li X, et al. The Dual Role of Gut Microbiota and Their Metabolites in Hepatocellular Carcinoma: A Context-Dependent Framework. Microorganisms 2025;14:73. [Crossref] [PubMed]
39. Wang L, Qiao W, Zhen X, et al. Targeting the gut-liver axis in cholangiocarcinoma: mechanisms, therapeutic advances, and future directions. Front Oncol 2025;15:1646897. [Crossref] [PubMed]
40. Ren S, Feng L, Liu H, et al. Gut microbiome affects the response to immunotherapy in non-small cell lung cancer. Thorac Cancer 2024;15:1149-63. [Crossref] [PubMed]
41. Spencer CN, McQuade JL, Gopalakrishnan V, et al. Dietary fiber and probiotics influence the gut microbiome and melanoma immunotherapy response. Science 2021;374:1632-40. [Crossref] [PubMed]
42. Nomura M, Nagatomo R, Doi K, et al. Association of Short-Chain Fatty Acids in the Gut Microbiome With Clinical Response to Treatment With Nivolumab or Pembrolizumab in Patients With Solid Cancer Tumors. JAMA Netw Open 2020;3:e202895. [Crossref] [PubMed]