| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Laboratory Medicine,
2
Ear, Nose, and Throat Department, and
3
Department of Neurosurgery, Landeskliniken Salzburg, Müllner Hauptstrasse 48, A-5020 Salzburg, Austria
aauthor for correspondence: fax 43-662-4482-885, e-mail e.arrer{at}lks.at
Cerebrospinal fluid (CSF) leakage occurs mainly as a complication of head injuries or skull-base surgeries, but may also occur spontaneously or as a result of nontraumatic processes such as inflammatory disorders or tumors (1). Detection and management of CSF leakage is essential to prevent possible life-threatening infections of the central nervous system (2). Radiologic and invasive procedures may be used for the diagnosis of CSF leaks, but these procedures are laborious, expensive, and present potential risk to the patient. Therefore, noninvasive laboratory methods serve as screening procedures before definitive procedures are used to localize the site of the defect.
Laboratory assessment of CSF leakage relies on compositional differences between CSF and other body secretions. The ß2-transferrin (ß2Tr) 
-fraction, or asialotransferrin, is a brain-specific variant of transferrin that lacks neuraminic acid. It therefore can be distinguished from serum transferrin by electrophoretic procedures and used to detect CSF rhinorrhea (3)(4). However, ß2Tr is present in aqueous humor and in perilymph fluid and can be detected in serum, especially in chronic alcohol abusers and in patients with inborn errors of glycoprotein metabolism or genetic variants of transferrin (5)(6).
ß-Trace protein (ßTP), recently identified as prostaglandin D2 synthase (7), is another brain-specific protein that is produced mainly in the leptomeninges and the choroid plexus and is secreted into the CSF. ßTP is the second most abundant protein in CSF after albumin. However, it is also present in other body fluids, including serum, albeit at much lower concentrations than in CSF (8). Immunoelectrophoretic methods for ßTP have been applied as a screening procedure for CSF leaks (9)(10). Recently, a nephelometric assay for the quantification of ßTP was introduced that promised several advantages over current CSF detection methods, including enhanced sensitivity and reduced turnaround time (11). We therefore evaluated this assay with respect to its clinical utility in patients with suspected CSF rhinorrhea.
ßTP was measured using the Prospec® nephelometer and the N Latex ßTP® test (Dade-Behring) according to the manufacturer’s instructions. For standardization, highly purified ßTP from human CSF (N Protein Standard UY®) was used because no international standard is available. Precision was monitored using N/T protein Control LC® (Dade-Behring). Nasal secretions were collected using intranasal pledgets (12) and were centrifuged (2000g for 10 min). Viscous secretions were prediluted 1:100 (1 mL + 99 mL) with N dilution buffer®, and the usual 1:100 predilution step included in the automated procedure was omitted for such samples. The range for reliable measurements in serum, nasal secretions, and CSF was 0.1–19.6 mg/L, and the intraassay and interassay CVs were 1.5–5.7% and 1.4–7.2%, respectively (11). Intra- and interassay CVs, determined in nasal secretions with a ßTP concentration of 1.27 mg/L, were both <4% as determined in our laboratory. Analysis time was <15 min. ß2Tr testing was performed as described previously (12)(13). In brief, nasal secretions were subjected to electrophoresis, immunofixation, and silver staining for visual inspection.
The sensitivity of the ßTP assay for CSF leakage detection and possible matrix effects in nasal secretions was ascertained in mixtures of nasal secretions and CSF. Various dilutions from a control with ßTP concentrations of 15.4 mg/L in the CSF and 0.117 mg/L in the nasal secretions were analyzed in duplicate. Measured and calculated ßTP concentrations were nearly identical, suggesting no interference by nasal secretion matrix components. CSF dilutions of 1:80 (1 mL + 79 mL), corresponding to 1–2% CSF in nasal secretions, were reliably detected, indicating excellent analytical sensitivity. We then measured ßTP in nasal secretions and sera of healthy adults (control group), in patients with suspected CSF rhinorrhea, and in patients on hemodialysis or with a reduced glomerular filtration rate. In addition, ßTP was measured in the CSF of patients who underwent a detailed neurologic work-up for various disorders but had a normal CSF as judged by protein, electrolyte, and cellular content (Table 1
). The mean (SD) ßTP in the sera of the 116 controls was 0.59 (0.23) mg/L. The mean ßTP concentration in normal CSF was 19.6 (5.8) mg/L (n = 19), and the mean ratio of CSF to serum ßTP was 33.2. The mean ßTP concentration in nasal secretions in the control group was 0.39 (0.29) mg/L (n = 160). There was no difference between nasal secretions collected from the right and left nostrils (n = 104; P = 0.92, t-test). Although the mean ßTP concentration was significantly lower in nasal secretions than in serum (P = 0.0005, t-test), in 14 of 101 individuals the ßTP concentration was higher in the nasal secretions than in the serum. We have no clear explanation for this result. It is possible that evaporation of fluid during collection caused an increase in the apparent ßTP concentration.
|
Serum ßTP may influence its concentration in nasal secretions, affecting the cutoff value for CSF rhinorrhea screening. We therefore measured ßTP in the sera of patients with conditions known or suspected to affect serum and nasal secretion concentrations. Confirming previous findings (14), the mean serum ßTP in hemodialysis patients was higher [11.15 (2.95) mg/L; n = 14] than in the controls (P <0.001). We also observed increased serum ßTP in patients with increased serum creatinine (P <0.001; data not shown).
To develop criteria for the presence of CSF in nasal secretions, we calculated the 97.5 percentile of the ßTP concentration in nasal secretions from the control group and obtained a value of 1.31 mg/L (Table 1
and Fig. 1
). Additional approaches that considered differences in ßTP between serum and nasal secretions arrived at comparable cutoff values. Of note was that preliminary cutoff values were obtained in individuals with creatinine values <11 mg/L in males and <9 mg/L in females, respectively, the upper limits of the sex-specific reference intervals.
|
We next evaluated these criteria in 187 patients who harbored suspicious lesions on the skull base as detected by computed tomography and/or who displayed clinical or subclinical signs of CSF rhinorrhea. Laboratory evaluation was done without knowledge of the clinical diagnosis. In the majority of patients (n = 157), the ßTP concentration in the nasal secretions was below the cutoff (1.31 mg/L), but 30 patients had concentrations above the cutoff (Fig. 1
). All but 2 of the 30 values were far above the calculated cutoff and were therefore highly suspicious for CSF rhinorrhea. In all 30 patients, surgical procedures established CSF leakage, whereas in the patients with values below the cutoff, no CSF leakage was established by any other procedure. Depending on the individual clinical situation, these procedures included high-resolution computed tomography, magnetic resonance cysternography, endoscopic detection of sodium-fluorescein, surgery, and follow-up. Results of ß2Tr testing, although known to the clinician, were not used as main decision criteria for performing the above procedures.
We also compared ßTP and ß2Tr testing for the detection of CSF rhinorrhea in this population. In the 30 cases with ßTP >1.31 mg/L and established CSF rhinorrhea, ß2Tr testing was unambiguously positive in 28 cases and negative in 2 cases. In four cases, in whom CSF rhinorrhea was not verified by other means, including follow-up of patients, qualitative ß2Tr testing was weakly positive, whereas ßTP was <1.31 mg/L. Thus, in our patients with clinically significant CSF rhinorrhea, the sensitivities of ßTP and ß2Tr were 100% (95% confidence interval, 88–100%) and 93% (78–99%), respectively. The corresponding specificities were 100% (98–100%) and 97% (94–99%).
CSF leakage is associated with a 10% risk of life-threatening meningitis. Hence, identification and subsequent repair of CSF leakage is mandatory. Although CSF rhinorrhea is not a common condition, our hospital, which serves as a tertiary care center for a population of more than 800 000, receives 
180 laboratory referrals per year (15% with positive test results) for suspected CSF rhinorrhea. Screening for CSF-specific components in nasal secretions has become a valuable component in our diagnostic work-up. Our studies suggest that measurement of ßTP in serum and nasal secretions fulfills the criteria for such a screening procedure. The rationale for using ßTP to detect CSF in nasal secretions is based on observations showing a 30- to 40-fold higher concentration in CSF than in serum (8). Compared with ß2Tr testing, isoform separation is not required. Hence, ßTP testing can be automated, has a higher analytical sensitivity, and is less time-consuming. Contaminating blood or wound secretions that affect the analytical performance of ß2Tr testing are of little concern for ßTP testing because of the predilution step.
The high analytical performance of ßTP measurements coupled with a large difference in ßTP between CSF and serum/nasal secretions would allow detection of 1–2% CSF admixtures to nasal secretions as shown by our dilution experiments. However, interindividual variability of ßTP in CSF, serum, and nasal secretions (Table 1
) may affect the detection limit. Preanalytical variability in nasal secretion collection, as suggested by higher ßTP values in nasal secretions in comparison with serum in some healthy individuals, also broadens the range of reference values in nasal secretions. Moreover, the decreased ßTP concentration in CSF found in acute bacterial meningitis [our unpublished observations and Ref. (15)] must be taken in account. Thus, the detection limit for ßTP testing in the clinical setting may increase from 1–2% CSF in nasal secretions to 5% or even higher in the case of acute bacterial meningitis.
The development of cutoff values is essential for the clinical utility of ßTP as a screening procedure for CSF rhinorrhea. Clinicians expect yes-or-no answers for the presence of CSF in secretions. On the basis of our own experience in more than 200 patients with suspected CSF rhinorrhea, we propose a cutoff value of 1.31 mg/L in nasal secretions. Results between 1.31 mg/L (97.5 percentile) and 1.7 mg/L (the highest value measured in controls) should be confirmed by testing of another sample because cutoff values are affected by preanalytical variability. Importantly, these criteria are valid only when ßTP in serum is 
1.27 mg/L (97.5 percentile). Thus, suspected CSF rhinorrhea in patients with impaired glomerular function requires additional considerations.
Acknowledgments
We thank Henny Brunnbauer, Barbara Didl, and Hermine Ploner for excellent technical assistance and Dade-Behring for providing us with N Latex ßTP nephelometric test reagent sets.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  K. M. Lloyd, J. M. DelGaudio, and P. A. Hudgins Imaging of Skull Base Cerebrospinal Fluid Leaks in Adults Radiology, September 1, 2008; 248(3): 725 - 736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Gorogh, P. Rudolph, J. E. Meyer, J. A. Werner, B. M. Lippert, and S. Maune Separation of {beta}2-Transferrin by Denaturing Gel Electrophoresis to Detect Cerebrospinal Fluid in Ear and Nasal Fluids Clin. Chem., September 1, 2005; 51(9): 1704 - 1710. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. L. Sanders, R. J. Clark, and J. A. Katzmann Cerebrospinal Fluid Leakage: Agarose Gel Electrophoresis Detection of {beta}2-Transferrin and Nephelometric Quantification of {beta}-Trace Protein Clin. Chem., December 1, 2004; 50(12): 2401 - 2403. [Full Text] [PDF]
|
|
|
|
|
|
C. Schnabel, E. Di Martino, J. M. Gilsbach, D. Riediger, A. M. Gressner, and D. Kunz Comparison of {beta}2-Transferrin and {beta}-Trace Protein for Detection of Cerebrospinal Fluid in Nasal and Ear Fluids Clin. Chem., March 1, 2004; 50(3): 661 - 663. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
