Distinct composition and metabolic functions of human gut microbiota are associated with cachexia in lung cancer patients

Cachexia is a multifactorial syndrome in cancer characterized by weight loss, muscle wasting, adipose tissue changes, physical dysfunction, and anorexia, driven by inflammation and abnormal metabolism. It reduces treatment options, quality of life, and survival, and is common in gastrointestinal and lung cancers. Proposed mechanisms include decreased anabolic signaling (e.g., IGF-1) and increased catabolic, pro-inflammatory cytokines (IL-6, IFN-γ, TNF-α). Serum lipopolysaccharide-binding protein (LBP) has been suggested as a cachexia biomarker. The study aims to investigate the role of the human gut microbiome and its metabolic functions in lung cancer–associated cachexia by integrating shotgun metagenomics with plasma metabolomics.

- Cohort: 31 lung cancer patients (12 women, 19 men) enrolled at the National Koranyi Institute of Pulmonology (Budapest, Hungary) and the County Hospital of Torokbalint (Torokbalint, Hungary). - Cachexia classification: Abridged Patient-Generated Subjective Global Assessment (aPG-SGA). Group A (well-nourished, scores 0–4, n=19) designated non-cachexia; Groups B (scores 5–9, n=8) and C (scores >9, n=4) merged as cachexia (n=12). Groups matched for age, sex, and other potential microbiota confounders; no significant differences in cancer subtype or stage. - Survival and clinical comparisons: BMI, COPD assessment test (CAT), smoking pack-years. Survival analyzed via log-rank test; univariate survival analysis performed. - Plasma metabolomics: Untargeted UHPLC-QTOF-MS on plasma. >5000 features detected; 314 common metabolites identified semi-targetedly. Statistical analyses: Bray–Curtis dissimilarities, ANOSIM for group separation; betadisper for dispersion; Student’s t-test for differential metabolites with FDR correction. - Fecal metagenomics: Shotgun sequencing of bacterial DNA from 31 fecal samples at ~6 Gbp average depth. Diversity analyses: alpha-diversity (Chao1, Shannon, Simpson; Wilcoxon rank-sum), beta-diversity using Bray–Curtis with ANOSIM; dispersion via betadisper; PERMANOVA for association with cancer stage. Taxonomic comparisons at phylum and species levels; multiple testing correction (FDR). Firmicutes/Bacteroidetes ratio and correlation with BMI assessed. NMDS ordination included comparison with a healthy Dutch cohort (n=471) partitioned by BMI. - Integrative analyses: Spearman correlations between differentially abundant species and plasma metabolites to propose mechanistic links (e.g., BCAAs with Prevotella copri; 3-oxocholic acid with Lactobacillus gasseri). Functional pathway inferences included evaluation of lipopolysaccharide biosynthesis capacity. - Machine learning: A high-performance model using gut microbial features was used to observe involvement of the microbiome in cachexia (details not provided in the excerpt).

- Clinical outcomes: - Cachexia group had significantly lower BMI than non-cachexia (p=5.7e−08, Wilcoxon rank-sum). - Cachexia associated with reduced survival probability (p=0.0051, log-rank). Patients with SGA A had increased survival vs B or C (p=0.0019, log-rank). - Plasma metabolomics: - Overall metabolomic profiles differed between cachexia and non-cachexia (ANOSIM p=0.026, r=0.110) with greater dispersion in cachexia (p<0.01, betadisper). - Metabolite classes depleted in cachexia included amino acids, vitamins, and indoles. - 41 individual metabolites differed (p<0.05); isoleucine (and one other) remained significant after FDR correction (FDR p<0.2). - Essential amino acids (isoleucine, leucine, tryptophan) were more abundant in non-cachectic patients. Valine trended lower in cachexia (p=0.103, ns). Serum cholesterol showed no difference (p=0.774). - Pipecolic acid was enriched in cachectic patients (p<0.05). - BCAAs and 3-oxocholic acid were enriched in non-cachectic patients and positively correlated with specific microbes (Prevotella copri and Lactobacillus gasseri, respectively). - Gut microbiome composition: - No phylum-level differential abundance; no significant alpha-diversity differences (Chao1 p=0.21; Shannon p=0.064; Simpson p=0.25). - Significant beta-diversity differences between cachexia and non-cachexia (ANOSIM p=0.001, r≈0.247) and increased compositional dispersion (p<0.001). - No significant association of overall microbiota with cancer stage (PERMANOVA p>0.05). - Compared with a healthy Dutch cohort (n=471), the cachexia group occupied a distinct ordination space (ANOSIM p=0.001, r=0.212), not simply resembling lean microbiota. - Firmicutes/Bacteroidetes ratio did not differ by cachexia status (p=0.1196) or obesity status (p=0.4113) and did not correlate with BMI (rho=0.1044, p=0.5747). - 51 species were differentially abundant (p<0.05); 44 remained significant after FDR (q<0.05). Thirteen prevalent species (≥20% prevalence) were significant, mostly enriched in non-cachexia. Prevotella copri was significantly lower in cachexia (FDR p=0.006). - Functional inference: - Gut microbiota capacity for lipopolysaccharide biosynthesis was enriched in cachectic patients, aligning with inflammation-related mechanisms. - Predictive modeling: - A machine learning model using only gut microbial features indicated the microbiome’s utility in distinguishing cachexia status (performance details not provided).

The study addresses whether and how the gut microbiome relates to cancer-associated cachexia in lung cancer patients. Significant differences in plasma metabolic profiles and gut microbial community structure between cachectic and non-cachectic patients suggest a systemic metabolic-gut axis involvement. Depletion of essential amino acids, particularly BCAAs, and vitamins in cachectic patients aligns with reduced microbial pathways supporting these metabolites, implicating microbiota-linked disruptions in nutrient availability and host anabolism. Positive correlations between non-cachectic enrichment of BCAAs and Prevotella copri, and between 3-oxocholic acid and Lactobacillus gasseri, suggest specific taxa may contribute to favorable metabolic states. Enrichment of LPS biosynthesis capacity in cachexia supports a pro-inflammatory microbial functional signature that could exacerbate catabolic processes and muscle wasting. The distinct separation of cachectic patients from non-cachectic cancer patients and healthy individuals at the species level underscores a cachexia-specific dysbiosis beyond BMI-related shifts. Collectively, these findings support a model where specific microbes and their metabolic functions influence host metabolism and inflammation in cachexia, highlighting potential microbial or metabolite targets for intervention.

This clinical study integrating plasma metabolomics and fecal shotgun metagenomics in lung cancer patients demonstrates that cachexia is associated with distinct gut microbial compositions and functions, coupled with systemic metabolic alterations. Cachectic patients exhibited depleted plasma BCAAs, vitamins, and indoles, reduced abundance of species such as Prevotella copri, and enrichment of microbiome LPS biosynthesis capacity. These host–microbiome links suggest that modulating specific gut taxa or replenishing key metabolites (e.g., BCAAs) could be explored as therapeutic avenues. Future research should validate these findings in larger, longitudinal cohorts; dissect causal mechanisms (e.g., gnotobiotic models or intervention studies); and develop microbiome-based predictive tools and targeted therapies for cachexia management.