Metabolism. 2024 Sep 17. pii: S0026-0495(24)00262-2. [Epub ahead of print] 156034
Na Sun,
Tanja Krauss,
Claudine Seeliger,
Thomas Kunzke,
Barbara Stöckl,
Annette Feuchtinger,
Chaoyang Zhang,
Andreas Voss,
Simone Heisz,
Olga Prokopchuk,
Marc E Martignoni,
Klaus-Peter Janssen,
Melina Claussnitzer,
Hans Hauner,
Axel Walch.
BACKGROUND: Cancer cachexia (CCx) presents a multifaceted challenge characterized by negative protein and energy balance and systemic inflammatory response activation. While previous CCx studies predominantly focused on mouse models or human body fluids, there's an unmet need to elucidate the molecular inter-organ cross-talk underlying the pathophysiology of human CCx.METHODS: Spatial metabolomics were conducted on liver, skeletal muscle, subcutaneous and visceral adipose tissue, and serum from cachectic and control cancer patients. Organ-wise comparisons were performed using component, pathway enrichment and correlation network analyses. Inter-organ correlations in CCx altered pathways were assessed using Circos. Machine learning on tissues and serum established classifiers as potential diagnostic biomarkers for CCx.
RESULTS: Distinct metabolic pathway alteration was detected in CCx, with adipose tissues and liver displaying the most significant (P ≤ 0.05) metabolic disturbances. CCx patients exhibited increased metabolic activity in visceral and subcutaneous adipose tissues and liver, contrasting with decreased activity in muscle and serum compared to control patients. Carbohydrate, lipid, amino acid, and vitamin metabolism emerged as highly interacting pathways across different organ systems in CCx. Muscle tissue showed decreased (P ≤ 0.001) energy charge in CCx patients, while liver and adipose tissues displayed increased energy charge (P ≤ 0.001). We stratified CCx patients by severity and metabolic changes, finding that visceral adipose tissue is most affected, especially in cases of severe cachexia. Morphometric analysis showed smaller (P ≤ 0.05) adipocyte size in visceral adipose tissue, indicating catabolic processes. We developed tissue-based classifiers for cancer cachexia specific to individual organs, facilitating the transfer of patient serum as minimally invasive diagnostic markers of CCx in the constitution of the organs.
CONCLUSIONS: These findings support the concept of CCx as a multi-organ syndrome with diverse metabolic alterations, providing insights into the pathophysiology and organ cross-talk of human CCx. This study pioneers spatial metabolomics for CCx, demonstrating the feasibility of distinguishing cachexia status at the organ level using serum.
Keywords: Cancer cachexia patients; Diagnostics of cachexia; Inter-organ cross-talk; Machine learning; Metabolomics classifier; Spatial metabolomics