Metabolic Profiling of CNS Lymphoma and Its Microenvironment

Abstract

588 Background: Establishing the pathologic basis of focal brain lesions caused by disease processes such as CNS lymphoma is often difficult. Results of standard MRI-based neuro-imaging are non-specific and cytological testing of cerebrospinal fluid (CSF) is highly insensitive. We tested the hypothesis that the metabolic constituents of the CSF reflect pathologic processes within the brain and thus might lead to the identification of novel diagnostic and prognostic biomarkers and provide insight into mechanisms of CNS lymphoma pathogenesis and its impact on cognitive function. We have pursued two novel approaches to define and model the metabolic microenvironment in CNS lymphoma: (1) Differential metabolite profiling of CSF; (2) Metabolic imaging using hyperpolarized 13C probes using preclinical models involving intracranial xenografts of CNS lymphoma. Methods: We are using GC/MS, in collaboration with Metabolon, to profile a broad spectrum of metabolites (> 140) in CSF from CNS lymphoma patients at diagnosis, relapse as well as from control subjects without brain tumors. In parallel, we are using high resolution MRI (14T) to perform anatomic and contrast enhanced imaging plus hyperpolarized metabolic imaging using 13C-metabolic probes such as [1-13C] pyruvate to monitor the production of intratumoral metabolites such as lactate within highly infiltrative xenograft models of CNS lymphoma. Results: Metabolite concentrations in CSF from 15 subjects with active CNS lymphoma were compared to 15 controls without brain tumors. Seventeen metabolites were significantly upregulated in CSF (> two-fold) compared to controls (p<0.05). Notably, despite similar tumor burdens, the CSF concentrations of lactate, malate and succinate were highest in patients with refractory lymphoma compared to chemotherapy-sensitive tumors. Another TCA intermediate, citrate, was detected at similar concentrations in CSF from controls and lymphoma patients. CSF concentrations of N-acetyl-aspartyl-glutamate, a neurotransmitter, were significantly lower in subjects with CNS lymphoma compared to controls. Metabolic imaging of murine CNS lymphoma xenograft (Raji lymphoma) using hyperpolarized [1-13C] pyruvate demonstrated intratumoral Warburg metabolism with a striking elevation of [1-13C] lactate that was apparent beyond the anatomical margins of the contrast enhancing lymphoma, as demonstrated by T1-weighted post-gadolinium images. Conclusions: These preliminary results suggest that 13C imaging may be able to identify regions of metabolic abnormality in CNS lymphoma outside those identified by 1H methods, and thus potentially provide insight into the metabolic features of the lymphoma as well as its stromal microenvironment. Our hypothesis is that MRI-based, non-invasive metabolic imaging will provide insights into the molecular differences between newly-diagnosed and relapsed cases as well as provide novel, non-invasive diagnostic and prognostic biomarkers. Supported by a grant from CFAR and the Sandler Program for Breakthrough Biomedical Research. Disclosures: No relevant conflicts of interest to declare.