Glial fibrillary acidic protein in cerebrospinal fluid in humans is sensitive to various pre-analytical conditions: possible explanation and solution

Abstract

Although glial fibrillary acidic protein (GFAP) has potential as a biomarker in cerebrospinal fluid, it is rarely used in clinical diagnosis due to high variability, low reliability, and poor reproducibility of results. Cerebrospinal fluid (CSF) was collected from patients (n = 167) at two sites at the Department of Neurology. CSF was sampled in various volumes in both 10 mL polypropylene (PP) tubes and small, filled, sealed tubes of ≤2.0 mL (microtubes) for the comparison of GFAP concentrations. The influence of pH, sample volumes during storage and transport of CSF, under different temperatures, was tested to identify the losses and increase the possibilities of replicating data for GFAP. Concentrations of GFAP were measured by a sandwich ELISA. Exposure to air, agitation, and open-close cycles increased pH and lowered CO₂. Compared to corresponding small filled sealed tubes, routine samples stored at −20°C showed 4–30% lower concentrations of GFAP. The loss increased further at lower volumes (< 0.5 mL). A significant difference in GFAP concentrations was seen in samples taken offsite (loss 42%) and onsite (loss 24%) compared to corresponding microtubes. Concentrations of GFAP remained stable in the microtubes, at 2–8°C and at RT for up to 3 weeks. GFAP in CSF is highly sensitive to changes in pH and dependent on adequate volumes for the best results. By avoiding exposure to air and agitation, we were able to stabilize GFAP concentrations in CSF by using small, filled, sealed tubes (microtubes). This handling could have impact on other biomarkers.

Keywords: CSF; GFAP; microtube; pre-analysis; reliability.

Figure 1

GFAP concentrations in samples from on-site and off-site compared to their corresponding concentrations in the small filled sealed tubes (microtube), as a percentage. The graph compares concentrations of GFAP in 500 μL of CSF that was immediately frozen at −20°C and thawed for analysis, with the values of GFAP in microtubes. Samples collected off-site were in transit for up to 48 h, under ambient conditions, from a spinal tap; off-site n = 75, on-site n = 74. Mean SD. **** p ≤ 0.0001.

Figure 2

Concentrations of GFAP in CSF in various tube types and pre-analytical conditions. Off-site samples were in transit, back and forth, up to 48 h under ambient conditions. On-site samples were kept on the bench or at 2–8°C. Each line represents samples from one individual that have been handled in various conditions. Origin = routine sample in a 10 mL PP tube, micro: microtube, or a 2 mL filled PP tube, and RT: room temperature.

Figure 3

GFAP concentrations in CSF: comparison of origin tubes and microtubes over time. CSF was collected simultaneously in 10 mL polypropylene (PP) tubes and microtubes. A 500 μL aliquot was transferred into 3.5 mL PP tubes, frozen at −20°C, and analyzed as a routine sample. Microtubes were stored at 2–8°C and analyzed twice: once within 1 week of lumbar puncture and again after 3 weeks. n = 16. ns = non-significant; ***p ≤ 0.001.