| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Technical Briefs |
1 Division Clinical Biochemistry & Immunology, Department of Laboratory Medicine & Pathology, Mayo Clinic College of Medicine, Rochester, MN
aaddress correspondence to this author at: 210 Hilton Bldg., Mayo Clinic, Rochester, MN 55905; fax 507-266-4088, e-mail katzmann{at}mayo.edu
Cerebrospinal fluid (CSF) leakage may occur with head trauma, surgery, inflammation, tumors, or congenital malformations. Determining the presence of CSF leakage is important so the leakage can be repaired and treatment can be initiated to prevent infection in the central nervous system. In this study, electrophoretic methods using three different detection systems for ß2-transferrin (ß2Trf) were compared with a nephelometric method for ß-trace protein (ßTP) for ease of use and sensitivity in the clinical laboratory setting to diagnose CSF leakage. Results were compared for 30 samples that had been submitted to the clinical laboratory for assessment of CSF leakage.
Detection of ß2Trf has been used as an endogenous marker to identify CSF in fluids to diagnose CSF leaks (1)(2)(3)(4). Transferrin that migrates in the ß-1 electrophoretic fraction (ß1Trf) is found in most body fluids. ß2Trf is used as a CSF-specific variant of transferrin, but it is also found in the aqueous humor and serum of patients with rare metabolic glycoprotein disorders or genetic variants of transferrin. If ß1Trf and ß2Trf are detected in drainage fluids by electrophoresis and subsequent immunodetection, the specimen is presumed to be contaminated with CSF.
The first electrophoresis method we evaluated used Titan Gel Silver Stain reagents (Helena) and anti-transferrin immunofixation. Samples, a negative control (serum), and a known positive control (CSF) were electrophoresed on the Titan Gel and immunofixed with 600 µL of anti-human transferrin that was diluted with 1200 µL of saline (1:3 dilution; DiaSorin Inc.); antigen-antibody complexes were then stained with Titan Gel Silver Stain reagents, according to the manufacturer’s procedure.
The second electrophoresis method evaluated was the Hydrogel 6 CSF assay (Sebia). Samples (10 µL of sample diluted with 40 µL of sample diluent), a negative control, and a positive control were electrophoresed on an agarose gel (Sebia) and immunofixed with 20 µL of anti-human transferrin horseradish peroxidase conjugate diluted with 200 µL of antiserum diluent. Antigen–antibody complexes were visualized with TTF3 solution (acidic dimethylformamide) provided with the Sebia assay.
The third electrophoresis method used the Titan High Resolution Gel and immunoblotting with anti-transferrin antiserum. The electrophoresis conditions were the same as in the first method. Samples, negative control, and a known positive control were electrophoresed and then immunoblotted according to the procedure for the SPIFE IgG IEF method (Helena), using 250 µL of anti-human transferrin horseradish peroxidase conjugate (Bethyl Laboratories, Inc.) diluted in 15 mL of blocking solution B. The enzyme-linked antibody was visualized with 3-amino-9-ethyl-carbozole (5 mg/5 mL of methanol) diluted in 25 mL of 0.02 mol/L sodium acetate buffer (pH 5.1) with 25 µL of 300 g/L hydrogen peroxide.
The nephelometric method quantified ßTP. ßTP is the second most abundant protein in CSF behind albumin (5) and has been identified as prostaglandin D2 synthase. ßTP is a brain-specific protein found in CSF, but it is also found at low concentrations in serum and other body fluids (6). We quantified ßTP with the Dade Behring BNII nephelometer and N Latex ßTP reagents and procedure. This Dade Behring reagent set is currently available in the United States as a "for research use only" product.
A summary of the assay validation and method comparison is given in Table 1
. The serum values in Table 1
are the range of results obtained from 39 donors; we excluded 1 patient with reduced renal function and serum ßTP of 2.0 mg/L as decreased glomerular filtration is associated with increased ßTP in the range of 1.2–6.6 mg/L (7). The expected results for the three electrophoretic methods are a single transferrin band (ß1) in serum and ß1 and ß2 bands in the CSF. The ranges of serum and CSF results are consistent with published values (6)(8)(9). The Dade Behring nephelometric assay has a CV of 3.1% and has the advantage of being sufficiently rapid to provide results in time to inform decisions in the emergency room and operating room settings.
|
After dilutions of CSF with saline, the ß2Trf band remained detectable to 2.5% CSF by the Sebia method, lower than the other methods. Comparison of results for 30 fluid specimens that were submitted for CSF leak detection showed good agreement among the four procedures. The 16 concordant positive samples all had ß2Trf bands and had ßTP at concentrations that ranged from 19 to 28.9 mg/L. The 12 concordant negative samples all contained the single ß1Trf band but no ß2Trf band and had ßTP <1.0 mg/L. The electrophoretic procedures were easy to interpret, and the nephelometric assay gave two clearly separate groups of positive and negative samples with no overlap of results.
Although we did not confirm CSF leakage with invasive procedures, we interpreted the two discordant samples as illustrating the lack of sensitivity of the immunofixation/silver stain method (method 1) compared with the other methods as well as the difficulty of the nephelometric procedure for classifying borderline abnormal results. The first of the two discordant samples had a negative result by method 1, a positive result by methods 2 and 3, and a ßTP result of 2.4 mg/L. We interpret these results as positive for CSF leakage, but the ßTP result was only modestly increased above the serum range and should be interpreted cautiously. The second discordant sample is clearly negative by methods 2, 3, and 4, but its low protein content and the insensitivity of our assay make the result of method 1 uninterpretable.
Among the electrophoretic methods, the Sebia Hydrogel 6 CSF test (method 2) had the additional advantage of a shorter procedural time of 2 h 20 min compared with 4 h 16 min and 3 h 40 min for methods 1 and 3, respectively. The Dade Behring N Latex ßTP assay was the fastest of all methods, requiring 12 min in the "stat" mode and little hands-on time. All of the electrophoretic methods could analyze the viscous samples that are occasionally encountered in these fluids, and they also used small sample volumes for testing (4 or 10 µL). If samples are manually prediluted 1:100, the N Latex ßTP assay also requires <10 µL. The N Latex ßTP assay measured a minimum concentration of 0.2 mg/L. However, because serum has a range of 0.5–1.2 mg/L and CSF a range of 4.5–22.5 mg/L, there is a theoretical maximum limit of detection of 
5% CSF in serum.
ßTP in CSF was stable for 14 days at room temperature, refrigerated, and frozen with and without three freeze-thaw cycles. ß1- and ß2Trf in CSF were also stable for 14 days at room temperature, refrigerated, or frozen with or without three freeze-thaw cycles unless saliva was mixed in with the CSF. A CSF/saliva mixture was unstable at room temperature or refrigerated, but was stable frozen with or without three freeze-thaw cycles.
In summary, it is our conclusion that the Sebia Hydrogel 6 CSF is the preferred electrophoresis method, and our calculations suggest that it will be positive at a lower percentage of CSF than will ßTP. The Dade Behring N Latex ßTP assay, however, assesses a ligand that is stable and can assay samples in 12 min. These characteristics suggest that the Dade Behring ßTP will be a useful screening assay to identify CSF leaks in most fluids submitted to the clinical laboratory. If, however, the ßTP result is equivocal, then the Sebia CSF reagent set will be needed to identify ß2Trf. Because the nephelometric assay is automated and rapid, this approach could allow many laboratories to provide "real-time" results for most of the patient samples submitted for CSF leak detection.
Acknowledgments
Sebia (Norcross, GA) and Dade Behring (Newark, DE) provided the reagent sets for this study at no cost.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
