Plasma Metabolites Forecast Occurrence and Prognosis for Patients With Diffuse Large B-Cell Lymphoma

Abstract

BackgroundDiffuse large B-cell lymphoma (DLBCL) is the most common non-Hodgkin’s lymphoma with considerable heterogeneity and different clinical prognosis. However, plasma metabomics used to forecast occurrence and prognosis of DLBCL are rarely addressed.MethodA total of 65 volunteers including 22 healthy controls (Ctrl), 25 DLBCL patients newly diagnosed (ND), and 18 DLBCL patients achieving complete remission (CR) were enrolled. A gas chromatography mass spectrometry-based untargeted plasma metabolomics analysis was performed.ResultsMultivariate statistical analysis displayed distinct metabolic features among Crtl, ND, and CR groups. Surprisingly, metabolic profiles of newly diagnosed DLBCL patients undergoing different prognosis showed clear and distinctive clustering. Based on the candidate metabolic biomarkers (glucose and aspartate) and clinical indicators (lymphocyte, red blood count, and hemoglobin), a distinct diagnostic equation was established showing improved diagnostic performance with an area under curve of 0.936. The enrichment of citric acid cycle, deficiency of branched chain amino acid, methionine, and cysteine in newly diagnosed DLBCL patients was closely associated with poor prognosis. In addition, we found that malate and 2-hydroxy-2-methylbutyric acid were positively correlated with the baseline tumor metabolic parameters (metabolically active tumor volume and total lesion glycolysis), and the higher abundance of plasma malate, the poorer survival.ConclusionOur preliminary data suggested plasma metabolomics study was informative to characterize the metabolic phenotypes and forecast occurrence and prognosis of DLBCL. Malate was identified as an unfavorable metabolic biomarker for prognosis-prediction of DLBCL, which provided a new insight on risk-stratification and therapeutic targets of DLBCL. More studies to confirm these associations and investigate potential mechanisms are in the process.