Metabolic abnormalities underlying symptom clusters in lung cancer patients: A weighted gene coexpression network analysis–based metabolomic analysis
Yanxuan Xiao, Yuanyuan Luo, Ruiyan Bai, Chang Gao, Dongmei Mao, Le Zhang, Zhihui Yang, Yang Zhao, Lili ZhangAbstract
Objectives:
To investigate the metabolic underpinnings of three major symptom clusters (SCs): somato-psychological, respiratory, and gastrointestinal SCs in lung cancer (LC) patients undergoing chemotherapy.
Methods:
Plasma from 100 LC patients receiving chemotherapy was analyzed through untargeted metabolomics. Symptoms were assessed using the M.D. Anderson Symptom Inventory-Chinese version (MDASI-C) and LC module scales. Weighted gene coexpression network analysis (WGCNA) was performed to identify metabolite modules associated with SCs, followed by biological enrichment analysis and core metabolite screening. Patients were categorized into positive and negative SC groups using the 75th percentile. Core metabolites were then used to evaluate their combined effect on SCs.
Results:
Hierarchical clustering of 96 samples identified 12 modules through WGCNA. The black module was negatively correlated with somato-psychological SC, with metabolites such as harmine and dihydroketoprofen linked to neuromodulation and other pathways. The corresponding metabolite panel yielded an AUC of 0.649. The blue module was negatively correlated with respiratory SC, with metabolites such as Ala-Pro and naproxen related to metabolism and synthesis pathways. The corresponding metabolite panel yielded an AUC of 0.706. The brown module was negatively correlated with gastrointestinal SC, whereas the magenta module showed a positive correlation. Key metabolites, including alpha-linolenic acid, acyclovir, diphenylmethylphosphine, DL-glutamic acid, phenylacetyl-L-glutamine, and acetylcysteine, were primarily linked to essential fatty acid metabolism and microbe-host interactions. The corresponding metabolite panel yielded anarea under the curve (AUC) of 0.749.
Conclusions:
This is the first study to apply metabolomics to multiple SCs in LC patients undergoing chemotherapy, identifying key metabolic alterations and discriminatory metabolites. Future studies should validate these findings in larger cohorts and explore mechanistic pathways for biomarker development.