Ex vivo NMR metabolomics approach using cerebrospinal fluid for the diagnosis of primary CNS lymphoma: Correlation with MR imaging characteristics

PURPOSE: Primary central nervous system lymphoma (PCNSL) is an uncommon extranodal non‐Hodgkin's lymphoma. Here, the feasibility of nuclear magnetic resonance (NMR) metabolomics for the diagnosis and prognosis prediction of PCNSL, as well as its correlation with magnetic resonance imaging (MRI) characteristics, was assessed. PATIENTS AND METHODS: Cerebrospinal fluid (CSF) samples from PCNSL and normal groups (n = 41 for each) were obtained along with MRI data including pre‐ and postcontrast as well as T1‐, T2‐, and diffusion‐weighted imaging for the treatment‐naïve PCNSL patients (n = 24). The CSF samples were analyzed using nuclear magnetic resonance (NMR). RESULTS: The CSF NMR metabolomic exhibited clear differences with a diagnostic sensitivity of 100% and a specificity of 97.6%. The citrate level of the leptomeningeal enhancement (LE) (+) group was significantly lower than that of the LE (−) group (p = 0.018). In addition, the MRI apparent diffusion coefficient (ADC) value of the tumor was positively correlated with the glucose level (p = 0.025). However, none of the marker metabolites were significant prognosis predictors in univariate analysis. CONCLUSIONS: In conclusion, the NMR metabolomics could be helpful to diagnose PCNSL, but not for the prognosis, and MRI features (LE or ADC) can reflect the metabolic profiles of PCNSL.