Clinical utility of cerebrospinal fluid vitamin D-binding protein as a novel biomarker for the diagnosis of viral and bacterial CNS infections

Abstract

Background: Rapid and accurate diagnosis of central nervous system (CNS) infections is important, and laboratory tests help diagnose CNS infections. Even when the patient has symptoms, laboratory tests often do not reveal any specific findings. The potential of vitamin D-binding protein (VDBP) to be used as a biomarker for viral and bacterial CNS infections was studied.

Methods: A total of 302 subjects with suspected CNS infection who underwent lumbar puncture were included. Clinical and laboratory data were collected retrospectively. VDBP levels were measured in the cerebrospinal fluid (CSF) samples. Genotyping for the GC gene encoding VDBP was also performed. VDBP levels were analyzed and compared by CNS infection, pathogen, CSF opening pressure, and GC genotype.

Results: A CNS infection group (n = 90) and a non-CNS infection group (n = 212) were studied. In terms of its receiver operating characteristic, CSF VDBP showed an area under the curve of 0.726 for the diagnosis of CNS infection. CSF VDBP levels were significantly different between the CNS infection and non-infection groups. The CNS infection group with enterovirus showed a statistically lower distribution of CSF VDBP levels than the other virus groups. The group with CSF opening pressure > 25 cmH₂O showed higher CSF VDBP levels than the other groups. There was no significant difference in GC gene allele distribution between the CNS infection and non-infection groups.

Conclusions: CSF VDBP levels were increased in patients with CNS infection. The CSF VDBP showed potential as a new biomarker for viral and bacterial CNS infections.

Keywords: Biomarker; Cerebrospinal fluid; Diagnosis; GC genotype; Infection; Vitamin D-binding protein.

Fig. 1

Comparison of median CSF VDBP levels and interquartile ranges (a) The difference in CSF VDBP levels between the CNS infection (3.20 μg/mL, 1.70–5.10) and non-CNS infection groups (1.10 μg/mL, 0.80–1.80)(P < 0.001). (b) The difference in CSF VDBP levels between the pathogen-identified group (2.34 μg/mL, 1.46–4.53) and non-identified group (3.76 μg/mL, 2.25–5.21)(P = 0.075) within the CNS infection group. Outliers of 1.5 to 3 IQR are marked with circles, and outliers > 3 IQR are marked with squares. Abbreviations: VDBP, vitamin D-binding protein; CNS, central nervous system; IQR, interquartile range

Fig. 2

ROC curve for VDBP versus CNS infections showed an AUC of 0.726. CNS infections included both clinically diagnosed cases and laboratory-confirmed cases

Fig. 3

CSF VDBP level comparisons among the EBV (n = 5), enterovirus (n = 29), and VZV (n = 4) groups. The Enterovirus group (median, IQR; 1.89 μg/mL, 1.44–2.54) showed significantly lower distribution than the EBV (7.58 μg/mL, 2.93–14.79) and VZV (6.62 μg/mL, 2.72–15.10) (P = 0.022) groups. Outliers of 1.5 to 3 IQR are marked with circles, and outliers > 3 IQR are marked with squares. * Statistically significant difference (P < 0.05) according to the Kruskal-Wallis test; n.s.: not significant. Abbreviations: VDBP, vitamin D-binding protein; CSF, cerebrospinal fluid; EBV, Epstein-Barr virus; ZVZ, varicella zoster virus

Fig. 4

The CSF VDBP level distribution according to CSF opening pressure. The VDBP level shows significant differences between patients with ≤25 cmH₂O pressure (median, IQR; 2.49 μg/mL, 1.64–4.28) and those with pressures > 25 cmH₂O (12.29 μg/mL, 10.53–14.74) within the CNS infection group (P = 0.002). Outliers of 1.5 to 3 IQR are marked with circles and outliers > 3 IQR are marked with squares. Abbreviations: VDBP, vitamin D-binding protein; CSF, cerebrospinal fluid; CNS, central nervous system; IQR, interquartile range