Article Text


Increased serum concentrations of tissue plasminogen activator correlate with an adverse clinical outcome in patients with bacterial meningitis
  1. F Winkler,
  2. S Kastenbauer,
  3. U Koedel,
  4. H W Pfister
  1. Department of Neurology, Klinikum Grosshadern, Ludwig-Maximilians University, Marchioninistr 15, D-81377 Munich, Germany
  1. Correspondence to:
 Dr H-W Pfister;

Statistics from

Bacterial meningitis is the most common serious infection of the central nervous system. It is still characterised by high mortality and morbidity in adults. In this disease extensive perpetuated inflammation with leucocyte invasion into the central nervous system (CNS) results in breakdown of the blood–brain barrier and promotes neuronal damage.1

Tissue type plasminogen activator (tPA) has been shown to have various biological effects that could have an impact on the pathophysiological changes observed in bacterial meningitis. In the CNS, endothelial cells, microglia, astrocytes, and neurones can produce the 70 kDa protein tPA, which normally does not cross the blood–brain barrier.2 Raised tPA levels in the cerebrospinal fluid (CSF) have previously been reported for certain CNS diseases such as multiple sclerosis, leukaemia, and encephalitis,3 and raised serum tPA levels for patients with sepsis.4

tPA converts plasminogen into plasmin, a rate limiting step in the proteolysis of fibrin, but also in the degradation of extracellular matrix, matrix metalloproteinase activation, and the processing of growth factors and cytokines.2 Further, tPA has been shown to increase neuronal cell death during excitotoxicity and cerebral ischaemia.2 Thus tPA may promote blood–brain barrier disruption, proinflammatory signalling, and neuronal damage, and so be involved in the pathophysiology of bacterial meningitis.

We studied the expression of tPA in the CSF and serum of 12 patients with bacterial meningitis (causative pathogens: Str pneumoniae (8); S aureus (3); H influenzae (1)) who had been admitted to our hospital (median age 63 years; range 29 to 78). Clinical outcome was measured according to the Glasgow outcome scale (GOS; 1, death; 2, persistent vegetative state; 3, severe disability; 4, moderate disability; 5, good recovery). Ten patients with non-inflammatory neurological diseases (median age 37 years; range 23 to 81) and 10 patients with Guillain-Barré syndrome, an inflammatory demyelinating polyradiculoneuropathy in which blood–CSF barrier breakdown occurs without CSF pleocytosis, served as controls (median age 59 years; range 34 to 84).

A lumbar puncture was done and venous blood collected for diagnostic purposes after the patient’s informed consent had been obtained. CSF and serum concentrations of tPA were measured by a specific enzyme linked immunosorbent assay (TintElize®, Biopool International, Ventura, California, USA; detection limit 1.5 ng/ml). Immunoreactive tPA concentrations are expressed as ng/ml of biological fluid.

Blood and CSF variables for the three patient groups were compared using the Mann–Whitney U test with α adjustment; a corrected p value of < 0.025 was considered significant. Bivariate correlations between clinical variables and tPA concentrations were analysed according to Spearman ρ (GOS) or Pearson (CSF leucocyte count, CSF/albumin ratio).

In all patients with bacterial meningitis, the CSF leucocyte count was markedly increased (median 1728 cells/μl; range 143 to 23 296). The CSF to serum albumin ratio (1000 × CSF albumin/serum albumin; normal < 7.4), the index used to quantify blood–CSF barrier breakdown, was significantly increased in all patients with bacterial meningitis (median 60.3; range 156 to 1400) and, to a lesser extent, in nine of the 10 patients with Guillain-Barré syndrome (median 12.8; range 4.7 to 39.0).

The tPA protein concentrations in the CSF and serum of patients with bacterial meningitis were increased compared with those of control patients and patients with Guillain-Barré syndrome; in both of the latter groups, tPA concentrations in the CSF were not detectable in nine of 10 patients (fig 1). The serum concentrations of tPA (mean (SD)) in patients with bacterial meningitis were about ninefold higher than the CSF concentrations (22.5 (13.8) v 2.4 (1.6) ng/ml, p < 0.05). CSF and serum concentrations in individual patients were positively correlated (r = 0.733, p < 0.01). Remarkably, high serum tPA concentrations in bacterial meningitis correlated with both an increased CSF to serum albumin ratio (r = 0.818, p < 0.01) and an unfavourable outcome according to the GOS (r = −0.72, p < 0.01). The CSF to serum albumin ratio also showed a high correlation with CSF tPA concentrations (r = 0.942, p < 0.001). For patients with bacterial meningitis no correlations were found between serum tPA and CSF leucocyte count (r = −0.319, p = 0.311), between CSF tPA and CSF leucocyte count (r = −0.070, p = 0.828), or between CSF tPA and the clinical outcome (r = −0.201, p = 0.530).

On the basis of these findings, we hypothesise that increased serum tPA contributes to breaching of the blood–brain/CSF barrier in bacterial meningitis. In turn, the breaching allows the serum tPA, which an intact blood–CSF barrier normally keeps separate from the CNS, to enter the CSF.

Our study shows for the first time that both CSF and serum tPA are increased in bacterial meningitis. Furthermore, upregulation of serum tPA correlated positively with breakdown of the blood–CSF barrier and an adverse clinical outcome of this disease. These findings are of particular importance in the light of earlier studies in rodent models, in which systemic infusion of tPA or plasmin resulted in blood–brain barrier disturbances in healthy control animals or in cerebral ischaemia.5 Disruption of the blood–brain/CSF barrier is an important pathophysiological alteration in bacterial meningitis, which contributes to CNS complications such as cerebral oedema and increased intracranial pressure.1 This may explain the additional correlation we found between high serum tPA levels and an adverse clinical outcome. A similar correlation was seen in patients with severe sepsis, a disease regularly associated with increased vascular permeability, in which serum tPA activity increased and was associated with mortality.4

Figure 1

(A) Concentrations of tissue type plasminogen activator (tPA) in the cerebrospinal fluid (CSF) of control patients (controls, mean (SD): 1.54 (0.15) ng/ml), patients with bacterial meningitis (2.42 (1.59) ng/ml), and patients with Guillain-Barré syndrome (GBS, 1.50 (0.02) ng/ml). (B) Concentrations of tPA in the serum of controls (9.71 (6.92) ng/ml), patients with bacterial meningitis (22.51 (13.84) ng/ml), and patients with GBS (13.28 (8.74) ng/ml). *p < 0.025 v control patients; †p < 0.025 v patients with GBS. Dotted line: detection limit of the assay.


This study was supported by grants from the Förderprogramm Forschung und Lehre of the Ludwig-Maximilians University Munich (to FW) and from the Wilhelm Sander-Stiftung (to HWP). We thank Ms S Walter and B Angele for technical assistance and Ms J Benson for copy editing the manuscript.


View Abstract


  • Competing interests: none declared.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.