See the article by Geng et al. pp. 1668–1679
About half of the patients with primary CNS lymphoma (PCNSL) suffer from cognitive disturbances, behavioral changes, and disorientation at initial presentation.1 These symptoms may disappear completely under effective therapy resulting in complete tumor remission.2 Neurocognitive dysfunction in PCNSL may in part result from the mere mass effect of tumor. However, PCNSL is a diffusely infiltrating disease affecting widespread regions of the brain not necessarily identifiable as harboring tumor on MRI.3 Rapidly proliferating lymphoma cells with a high metabolic turnover may cause disturbances of neurotransmission, metabolic changes, and inflammatory response in affected and possibly even in remote brain regions and therefore alter neuronal and glial function.
In this issue, Geng and coworkers4 present results on patients with recurrent PCNSL, studied by serial cerebrospinal fluid (CSF) analyses in order to elucidate some of the mechanisms responsible for cognitive dysfunction in PCNSL or in brain tumors in general. Metabolites involved in neurotransmission and bioenergetics were measured in ventricular CSF and serial Mini-Mental State Examination (MMSE) as well as tumor volumetric analyses on MRI were carried out in 14 patients enrolled in a phase I trial of lenalidomide plus rituximab for recurrent PCNSL. Via an Ommaya reservoir, a prerequisite for the clinical trial, investigators had virtually unlimited access to ventricular CSF. Therefore, patients could undergo serial CSF analyses, repeated neuropsychological testing with MMSE, and serial MRI including volumetric analysis using semi-automated software to assess contrast-enhancing lesion volumes on T1-weighted post-contrast images.
“Candidate metabolomic biomarkers” of neurocognition were analyzed in ventricular CSF samples using a demanding methodology of liquid mass spectrometry. Among 18 different biomarkers, gamma-aminobutyric acid (GABA), dopamine, glutamate, acetylcholine, choline, epinephrine, serotonin, and lactate were examined. All 14 patients within this cohort (median age 66 years; range 47-79) experienced disease progression while under study therapy after 2-7 cycles and therefore could be monitored for the effects of tumor progression. A total of 51 CSF samples were analyzed, 20 at baseline and 31 after initiation of lenalidomide, and data were subjected to complex biostatistics. All biological data were correlated with MMSE results from tests performed at the same time CSF was collected. Mean CSF concentration of choline in CNS lymphomas was markedly elevated as expected. Not surprisingly MMSE scores were lower in 5 patients having been irradiated as compared to those 9 who had not. While the median lactate CSF concentration was within the normal range, concentrations as high as 3-fold the median were detected in some samples. Pretreatment baseline concentration of the CSF metabolites investigated did not significantly change under treatment. However, the authors found a “striking correlation” between high CSF lactate concentrations and low MMSE scores. An elevated ratio of glutamate to GABA, a CSF measure associated with some neuropsychiatric disorders as autism, schizophrenia, and alcohol withdrawal, also correlated with a lower MMSE score.
Tumor cell metabolism often is characterized by an acquired dependence on non-oxidative glycolysis for energy regeneration with increased lactate production as a consequence (Warburg effect). High CSF lactate in the present study correlated with lower CSF GABA concentrations. Hence, the authors speculated that elevated CSF lactate may impact the enzymatic conversion of glutamate to GABA via glutamic acid decarboxylase. High lactate and low GABA concentrations correlated with contrast-enhancing tumor volume and lower MMSE scores. Therefore, the authors concluded that an increased lactate concentration reflecting this Warburg effect of tumor metabolism has an impact on neurotransmitter concentration and may thus have a detrimental effect on neurocognition.
The authors provide first data on neurotransmitter imbalance and altered concentrations of metabolites involved in bioenergetics in the CSF of PCNSL patients. They took advantage of the fact, that via an Ommaya reservoir, they had virtually unlimited access to ventricular CSF. The present data provide us with hypotheses on mechanisms that may lead to cognitive disturbance in PCNSL. When interpreting these highly interesting findings, we nevertheless should be aware of some inherent limitations of this study:
Findings presented here in patients with recurrent PCNSL cannot be generalized for their metabolic alterations to other brain tumors: PCNSL show a unique angiocentric growth pattern5; the tumor cells are densely packed, which is visible on MRI by a high diffusion restriction6 and they show a very high proliferation rate. Furthermore, it has been shown that the proteome within the CSF detected by liquid chromatography coupled mass spectrometry is profoundly different in PCNSL in comparison to other brain tumors.7
CSF is a compartment that may reflect metabolomics and neurotransmitter alterations to a certain extent, however, metabolic and neurotransmitter changes may be quite different in CSF and in brain parenchyma infiltrated by lymphoma, particularly in patients, who had been heavily pretreated.
Pretreatment is a strong “confounder” of neurocognitive dysfunction in PCNSL8 and may itself have a profound influence on all the metabolites that have been investigated within this trial on a necessarily heterogeneous cohort of patients within a phase I trial on experimental therapy for recurrent PCNSL.
An additional note of caution is necessary regarding the outcome measure of cognitive dysfunction used in the study: MMSE is a screening test for dementing diseases and is not validated nor established in patients with brain tumors. Appropriate neuropsychometric testing for PCNSL patients has been proposed and published in a consensus by experts in the field.9 While it is more time-consuming than MMSE, PCNSL patients within clinical trials often are highly cooperative and willing to undergo extensive neuropsychometric testing.2 Taking into consideration the laborious, demanding, and advanced methodology applied in Dr Geng’s study to measure neurotransmitters and metabolites in CSF of PCNSL patients, one would have wished for a comparably sensitive and validated “assay” when evaluating cognitive dysfunction.
Irrespective of these considerations, Geng and colleagues ought to be congratulated for the novelty of their work exploiting advanced methodology to elucidate possible mechanisms of neurocognitive dysfunction in PCNSL, a matter of high clinical importance.
Funding
None.
Acknowledgments
The text is the sole product of the authors and no third party had input or gave support to its writing.
Conflict of interest statement. None.
Authorship statement. M.P. and U.S. have been involved in the writing of the manuscript at draft and any revision stages and have read and approved the final version.
References
- 1.Grommes C, Rubenstein JL, DeAngelis LM, Ferreri AJM, Batchelor TT. Comprehensive approach to diagnosis and treatment of newly diagnosed primary CNS lymphoma. Neuro Oncol. 2019;21(3):296–305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fliessbach K, Helmstaedter C, Urbach H, et al. Neuropsychological outcome after chemotherapy for primary CNS lymphoma: a prospective study. Neurology. 2005;64(7):1184–1188. [DOI] [PubMed] [Google Scholar]
- 3.Lai R, Rosenblum MK, DeAngelis LM. Primary CNS lymphoma: a whole-brain disease? Neurology. 2002;59(10):1557–1562. [DOI] [PubMed] [Google Scholar]
- 4.Geng H, Tsang M, Subbaraj L, et al. Tumor metabolism and neurocognition in CNS lymphoma. Neuro Oncol. 2021;23(10):1668–1679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Deckert M, Brunn A, Montesinos-Rongen M, Terreni MR, Ponzoni M. Primary lymphoma of the central nervous system – a diagnostic challenge. Hematol Oncol. 2014;32(2):57–67. [DOI] [PubMed] [Google Scholar]
- 6.Kickingereder P, Wiestler B, Sahm F, et al. Primary central nervous system lymphoma and atypical glioblastoma: multiparametric differentiation by using diffusion-, perfusion-, and susceptibility-weighted MR imaging. Radiology. 2014;272(3):843–850. [DOI] [PubMed] [Google Scholar]
- 7.Waldera-Lupa DM, Poschmann G, Kirchgaessler N, et al. A multiplex assay for the stratification of patients with primary central nervous system lymphoma using targeted mass spectrometry. Cancers. 2020;12(7):1–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Doolittle ND, Korfel A, Lubow MA, et al. Long-term cognitive function, neuroimaging, and quality of life in primary CNS lymphoma. Neurology. 2013;81(1):84–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Correa DD, Maron L, Harder H, et al. Cognitive functions in primary central nervous system lymphoma: literature review and assessment guidelines. Ann Oncol. 2007;18(7):1145–1151. [DOI] [PubMed] [Google Scholar]