CSF neopterin level as a diagnostic marker in primary central nervous system lymphoma

Neuro-Oncology Neuro-Oncology 17(11), 1497 – 1503, 2015 doi:10.1093/neuonc/nov092 Advance Access date 25 May 2015 CSF neopterin level as a diagnostic marker in primary central nervous system lymphoma Neuro-oncology Department, Hospices Civils de Lyon, Hôpital Neurologique, Bron, France (A.V., F.D., L.T.-M., S.C.-C., J.H.); INSERM U1028/CNRS UMR 5292, Lyon Neuroscience Research Center, Lyon, France (A.V., F.D., Y.T., G.K.B., L.T.-M., D.M., I.Q., J.H., A.P.-L.); Université de Lyon, Université Claude-Bernard Lyon 1, Lyon, France (A.V., F.D., L.T.-M., J.H.); Neurochemistry Unit, Biochemistry Department, Hospices Civils de Lyon, Hôpital Neurologique, Bron, France (Y.T., I.Q., A.P.-L.); Department of Anesthesiology, Pharmacology and Intensive Care, Geneva University Hospitals, Geneva, Switzerland (G.K.B.); Hematology Department, Centre Léon Bérard, Lyon, France (H.G.); Neuroradiology Department, Hospices Civils de Lyon, Hôpital Neurologique, Bron, France (G.L.-T.); Neurosurgery Department B, Hospices Civils de Lyon, Hôpital Neurologique, Bron, France (E.J.); Neurosurgery Department D, Hospices Civils de Lyon, Hôpital Neurologique, Bron, France (J.G.) Corresponding Author: Armand Perret-Liaudet, PhD, Neurochemistry Unit, Biochemistry Department, Hospices Civils de Lyon, Hôpital Neurologique, 69677 Bron, France (). Background. The diagnosis of primary central nervous system lymphoma (PCNSL) can be challenging. PCNSL lesions are frequently located deep within the brain, and performing a cerebral biopsy is not always feasible. The aim of this study was to investigate the diagnostic value of CSF neopterin, a marker of neuroinflammation, in immunocompetent patients with suspected PCNSL. Methods. We retrospectively reviewed the characteristics of 124 patients with brain tumor (n ¼ 82) or an inflammatory CNS disorder (n ¼ 42) in whom CSF neopterin levels were assessed. Twenty-eight patients had PCNSL, 54 patients had another type of brain tumor (glioma n ¼ 36, metastasis n ¼ 13, other n ¼ 5), and 13 patients had a pseudotumoral inflammatory brain lesion. Results. CSF neopterin levels were significantly higher in the patients with PCNSL than in those with other brain tumors (41.8 vs 5.1 nmol/L, P , .001), those with pseudotumoral inflammatory brain lesions (41.8 vs 4.3 nmol/L, P , .001), and those with nontumefactive inflammatory CNS disorders (41.8 vs 3.8 nmol/L, P , .001). In the 95 patients with space-occupying brain lesions, at a cutoff of 10 nmol/L, the sensitivity of this approach was 96% and the specificity was 93% for the diagnosis of PCNSL. The positive and negative predictive values were 84% and 98%, respectively. Conclusion. Assessing CSF neopterin levels in patients with a suspected brain tumor might be helpful for the positive and differential diagnosis of PCNSL. A prospective study is warranted to confirm these results. Keywords: biomarker, cerebrospinal fluid, neopterin, primary central nervous system lymphoma, space-occupying brain lesion. Primary central nervous system lymphomas (PCNSLs) are extranodal non-Hodgkin lymphomas affecting the brain, the leptomeninges, the spinal cord, and the eye1 in the absence of systemic lymphoma at the time of diagnosis. These lymphomas account for 3% to 5% of primary brain tumors. Diffuse large B-cell lymphoma (DLBCL) is the most common histopathological PCNSL subtype. PCNSLs are aggressive tumors, and prompt diagnosis is needed to rapidly initiate appropriate treatment. PCNSL MRI characteristics are frequently suggestive of the diagnosis, showing unique or multiple periventricular, homogeneously enhancing lesions. In addition, magnetic resonance spectroscopy and perfusion MRI can show suspicious abnormalities, such as low regional cerebral blood volume ratios and very high lipid resonances.2 Yet, MRI characteristics are not specific, and neuropathological examination is mandatory to establish the diagnosis. In 10% – 30% of cases, lymphoma cells can be identified in the cerebrospinal fluid (CSF) or in a vitreous body biopsy (if a vitreoretinal lymphoma is diagnosed on slit-lamp examination). However, in the majority of cases, a cerebral biopsy needs to be performed. Nevertheless, performing a cerebral biopsy can be technically challenging and potentially Received 25 February 2015; accepted 1 May 2015 # The Author(s) 2015. Published by Oxford University Press on behalf of the Society for Neuro-Oncology. All rights reserved. For permissions, please e-mail: . 1497 Aurélien Viaccoz, François Ducray, Yannick Tholance, Gleicy Keli Barcelos, Laure Thomas-Maisonneuve, Hervé Ghesquières, David Meyronet, Isabelle Quadrio, Stéphanie Cartalat-Carel, Guy Louis-Tisserand, Emmanuel Jouanneau, Jacques Guyotat, Jérôme Honnorat, and Armand Perret-Liaudet Viaccoz et al.: CSF neopterin level as diagnostic marker in PCNSL Materials and Methods Patients Between December 2007 and June 2014, we retrospectively identified patients from our department with brain tumor or who were suspicious for brain tumor or with an inflammatory CNS disorder in whom neopterin levels were assessed in the CSF. In patients with a space-occupying brain lesion, CSF analysis was generally performed to identify potential leptomeningeal dissemination (at diagnosis or at recurrence) or because of the suspicion of a tumefactive, nonneoplastic neurological disorder. The list of patients with available CSF neopterin levels was obtained from the records of the neurochemistry department and matched with the records of the neuro-oncology department. A definite neuropathological diagnosis was available for all patients with a brain tumor. The medical and radiological files of all patients were reviewed. This study has been approved by our institutional ethical committee on human experimentation. CSF Neopterin Analysis After collection, the CSF was immediately transported to the laboratory to be centrifuged at 2000 g at 128+38C for 10 min, after which supernatants were frozen at 2808C until analysis. The maximum delay between CSF sampling and supernatant freezing was 2 h. The maximum delay between freezing and analysis was 15 days. CSF neopterin levels were obtained as previously reported5 using high-performance liquid chromatography coupled with fluorimetric detection (all tests were supervised by A.P.-L.). Using pools of CSF samples prepared each year as an internal quality control, the ranges of 1498 yearly interassay coefficients and variability during the period of patient inclusion were 5.9% and 7% at 7.5 and 1.8 nmol/L of CSF neopterin, respectively. The upper normal reference value was previously determined to be 5 nmol/L (upper limit of the 99% confidence interval).13 Statistical Analysis Differences between the groups of patients were analyzed using the Mann – Whitney U-test and the Kruskal – Wallis test, and P , .05 was considered to be statistically significant. The diagnostic accuracy of CSF neopterin levels was assessed using receiver operati (...truncated)