Causal Effects of Plasma Metabolites on Leukemia: A Mendelian Randomization Study

Background: Leukemia comprises heterogeneous hematologic malignancies, and whether circulating metabolites contribute causally to subtype-specific risk remains uncertain. Objectives: The aim of this study was to assess the causal effects of plasma metabolites for acute myeloid leukemia (AML), chronic myeloid leukemia (CML), acute lymphoblastic leukemia (ALL), and chronic lymphocytic leukemia (CLL). Methods: A two-sample Mendelian randomization (MR) using summary-level genome-wide association study statistics was conducted. For each metabolite, a single variant showing the strongest association with the metabolite that had the largest variance explained (R²) among the independent genome-wide significant (p < 5 × 10^-8) SNPs assigned to effector genes was selected as sentinel. Multiple comparisons using Bonferroni correction (0.05/83 = 6.02 × 10^-4) were applied to minimize the risk of obtaining false positive results. Results: Totally, 83 metabolites and metabolite ratios were analyzed for AML, CML, ALL, and CLL. Lithocholate sulfate (1), instrumented by the rs10425975 variant in the SULT2A1 gene, was significantly associated with an increased risk of CLL (OR = 2.19; 95% CI: 1.45-3.31; p = 2 × 10^-4). An additional 17 metabolite-leukemia associations showed suggestive evidence of significance. Approximately 300 drug entries linked to candidate metabolites were curated to provide a basis for mechanistic follow-up. Conclusions: Our MR result supports a causal link between higher genetically proxied lithocholate sulfate (1) and increased CLL risk. The discovery of these "metabolite-gene-drug" relationships suggests a central role in leukemia pathogenesis and warrants further functional investigation for their therapeutic potential.