Diagnostic Cerebrospinal Fluid Biomarker Discovery and Validation in Patients with Central Nervous System Infections

Background Central nervous system (CNS) infections are common causes of morbidity and mortality worldwide. Rapid, accurate identification of the likely cause is essential for clinical management and the early initiation of antimicrobial therapy, which potentially improves clinical outcome. Methods We applied liquid chromatography tandem mass-spectrometry on 45 cerebrospinal fluid (CSF) samples from a cohort of adults with/without CNS infections to discover potential diagnostic protein biomarkers. We then validated the diagnostic performance of a selected biomarker candidate in an independent cohort of 364 consecutively treated adults with CNS infections admitted to a referral hospital in southern Vietnam. Results In the discovery cohort, we identified lipocalin 2 (LCN2) as a potential biomarker of bacterial meningitis. The analysis of the validation cohort showed that LCN2 could discriminate bacterial meningitis from other CNS infections, including tuberculous meningitis, cryptococcal meningitis and viral/antibody-mediated encephalitis (sensitivity: 0.88 (95% confident interval (CI): 0.77–0.94), specificity: 0.91 (95%CI: 0.88–0.94) and diagnostic odd ratio: 73.8 (95%CI: 31.8–171.4)). LCN2 outperformed other CSF markers (leukocytes, glucose, protein and lactate) commonly used in routine care worldwide. The combination of LCN2 and these four routine CSF markers resulted in the highest diagnostic performance for bacterial meningitis (area under receiver-operating-characteristic-curve 0.96; 95%CI: 0.93–0.99). Conclusions Our results suggest that LCN2 is a sensitive and specific biomarker for discriminating bacterial meningitis from a broad spectrum of CNS infections. A prospective study is needed to further assess the diagnostic utility of LCN2 in the diagnosis and management of CNS infections.

Continuous variables were compared using the Mann-Whitney U test or the Kruskal-Wallis test 1 5 or Wilcoxon signed-rank test. The correlation between continuous variables was assessed using 1 6 Spearman correlation test. All statistical tests were performed two-sided. The area under the 1 7 receiver operating characteristic curve (AUROC) was used to quantify the diagnostic  Table S2.

2
Biomarker discovery 1 Tandem mass-spectrometry analysis of 45 CSF samples of the discovery cohort identified a total 2 of 1,012 proteins. Of these, 891 were included in the analysis based on the number of peptides 3 and sequence coverage. Subsequent analysis identified a total of 729 quantifiable protein 4 signatures that were clinical-entity specific, especially for patients with BM ( Figure 2A). Of 5 these, 60 and 19 were significantly expressed in the CSF of patients with BM and TBM, 6 respectively (Table S3). No diagnostic biomarker candidate was found in patients with viral 7 encephalitis.

8
Of the protein candidates identified in the BM group, lipocalin 2 (LCN2), also known as 9 neutrophil gelatinase-associated lipocalin, had a sensitivity of 1 (95%CI: 0.73-1) and a availability of a quantitative ELISA assay, LCN2 was thus selected for further evaluation. In order to verify the mass-spectrometry findings, we performed quantitative ELISA analysis of   Figure S1). Of the patients with BM, CSF LCN2 levels were higher in those with a confirmed 9 diagnosis than in those without a bacteria identified ( Figure S1), while the duration of illness at 1 0 enrolment was similar between the two groups (data not shown). ng/ml) and a diagnostic odd ratio (DOR) of 44.8 or above ( Figure 3B).

7
Currently, CSF parameters such as leukocytes, protein, lactate and glucose concentrations are infections. We thus compared the diagnostic performance of LCN2 alone and in combination 2 0 with these markers..  to the CSF parameters based model ( Figure 4C). Similar results were obtained when assessing 8 the utility of LCN2 in discriminating between BM and other specific clinical entities (TBM or 9 encephalitis) ( Figure 4D and Figure S2). LCN2 did not, however, help distinguish confirmed 1 0 from suspected BM (Table S4).
1 1 produced in response to the bacterial invasion of the CNS. Here, using a mass-spectrometry-based approach, we initially identified LCN2 as a potential 2 diagnostic marker for BM. Additional validation work on an independent cohort showed that 3 LCN2 could accurately discriminate BM from other CNS infections. LCN2 also outperformed 4 existing BM diagnostic makers (CSF leukocytes, and protein, glucose and lactate concentrations) 5 that are currently used as part of routine care. A diagnostic model consisting of LCN2 and these 6 four CSF parameters gave the best diagnostic performance for BM. Our data thus suggest that 7 LCN2 can act as an independent diagnostic maker of BM alone or in combination with other 8 CSF parameters. the performance of LCN2 against commonly used CSF markers such as leukocytes, glucose, and anti-NMDAR encephalitis). Additionally, we also compared the diagnostic performance of 2 0 LCN2 against that of CSF markers commonly used as part of routine care worldwide. As such, and for the first time provide strong evidence that LCN2 is a highly sensitive biomarker for 1 discriminating BM from a broad-spectrum of CNS infections.

2
The differences in CSF LCN2 levels between laboratory confirmed and clinically suspected BM Beyond viruses: clinical profiles and etiologies associated with encephalitis. Clin Infect Dis. Cerebrospinal fluid markers to distinguish bacterial meningitis from cerebral malaria in children. computational platform for comprehensive analysis of (prote)omics data. Nat Methods.  from label-free quantitative mass-spectrometry analysis of 45 patients of the discovery phase.  Table S1. Diagnostic tests carried out as part of routine care and/or as per the study protocols Note to Table S1: Y: yes. ND: not done Table S2. Baseline characteristics of the discovery and validation cohort Note to Table S2: ♪ outcomes at discharge were recorded as full recovery (n=4) or neurological deficit (n=4)*due to the small sample size, data on two cases with neurotoxoplamosis are not shown, # including 44 laboratory confirmed cases, ♫ including 95 laboratory confirmed cases, $ including 23 laboratory confirmed cases, **Glasgow coma score, ^Modified Rankin Scale (0: full recovery with no symptoms, 1: No significant disability, 2: Slight disability, 3: Moderate disability, 4: Moderately severe disability, 5: Severe disability, and 6: Dead); BM: bacterial meningitis, TBM: tuberculous meningitis; Data are number (%), continuous variables are presented as median (range)