Abstract
BACKGROUND: Leukemia is driven by metabolic reprogramming, yet the specific causal roles of plasma metabolites in distinct leukemia subtypes remain unclear. METHODS: This study employed Mendelian randomization (MR) to explore potential causal links between 690 plasma metabolites (and 143 metabolite ratios) and four leukemia subtypes: ALL, AML, CLL, and CML. Genetic variants from genome-wide association studies served as instrumental variables. Multiple MR approaches, including IVW, MR-Egger, and Weighted Median, along with sensitivity analyses, were applied to ensure robust results. RESULTS: Our findings revealed subtype-specific metabolite associations. In ALL, metabolites such as 3-Hydroxyisobutyrate and γ-Glutamylglutamate showed positive associations, while Phosphocholine and Ceramide showed negative associations. AML was positively linked to GlcNAc/GalNAc and negatively to 1-Methylnicotinamide. CLL showed positive associations with Butyrate/Isobutyrate and Androstenediol Monosulfate, and negative ones with Docosatrienoate and α-Tocopherol to Sulfate ratio. CML exhibited negative associations with Cysteine-Glutathione disulfide and Piperine. CONCLUSION: Our MR study provides a comprehensive evaluation of the metabolomic landscape of leukemia, identifying subtype-specific causal associations involving pathways such as energy metabolism, amino acid metabolism, lipid signaling, and redox homeostasis. These findings offer insights into potential plasma biomarkers and therapeutic targets, revealing distinct metabolic vulnerabilities that warrant further investigation for precision treatment strategies across leukemia subtypes.