Intensive Care Medicine

, Volume 32, Issue 6, pp 946–947

Bronchoalveolar levels of plasminogen activator inhibitor-1 and soluble tissue factor are sensitive and specific markers of pulmonary inflammation

Authors

    • Departments of Intensive Care MedicineAcademic Medical Center, University of Amsterdam
    • Laboratory of Experimental Intensive Care and AnaesthesiologyAcademic Medical Center, University of Amsterdam
  • Julian L. Millo
    • Intensive Care UnitJohn Radcliffe Hospital
  • Chris S. Garrard
    • Intensive Care UnitJohn Radcliffe Hospital
  • Marcus J. Schultz
    • Departments of Intensive Care MedicineAcademic Medical Center, University of Amsterdam
    • Laboratory of Experimental Intensive Care and AnaesthesiologyAcademic Medical Center, University of Amsterdam
Correspondence

DOI: 10.1007/s00134-006-0167-9

Cite this article as:
Determann, R.M., Millo, J.L., Garrard, C.S. et al. Intensive Care Med (2006) 32: 946. doi:10.1007/s00134-006-0167-9

Sir: Activation of coagulation and inhibition of fibrinolysis are hallmarks of pulmonary inflammation. Changes in alveolar fibrin turnover have been reported in pneumonia [1, 2, 3] and acute respiratory distress syndrome (ARDS) [1]. Recently in Intensive Care Medicine El Sohl et al. [4] proposed plasminogen activator inhibitor (PAI)-1 as a biomarker of ARDS in patients with aspiration pneumonitis. They reported bronchoalveolar lavage fluid PAI-1 levels in 51 patients with witnessed aspiration who had a PaO2/FIO2 ratio lower than 300 mmHg for a period no less than 4 h from admission. PAI-1 antigen levels were more than five times higher in those who progressed to ARDS than in those who did not.

We recently reported procoagulant changes in the lungs during development of ventilator-associated pneumonia (VAP) [2, 5]. Non-directed lavage fluid was collected each alternate day beginning at the start of ventilation in patients expected to be ventilated longer than 3 days. All lavages were performed by the same investigator. Diagnosis of VAP required a clinical pulmonary infection score higher than 6 plus microbiological confirmation. The day on which antibiotic treatment was started was retrospectively assigned as the day of VAP diagnosis. Nine patients developed VAP and 19 served as controls. PAI-1, soluble tissue factor (sTF) and soluble thrombomodulin (sTM) were determined by enzyme-linked immunosorbent assay [2, 5]. PAI-1 and sTF increased before VAP was diagnosed, while they remained unchanged in controls. Levels of soluble thrombomodulin (sTM) increased from the diagnosis of VAP onwards.

We calculated sensitivity and specificity to determine whether PAI-1, sTF and sTM can be used as biomarkers that predict VAP, using levels 2 days before diagnosis of VAP. The highest level of PAI-1, sTF and sTM recorded in controls during the study period was used for comparison with VAP patients. A receiver-operating characteristic curve was constructed for all three parameters, and the inflection point was used to identify the optimal cut-off values (Fig. 1). A PAI-1 level of 8.5 ng/ml was detected in all but one VAP patient (sensitivity of 0.89, 95% CI 0.62–0.98) and in only one control patient (specificity of 0.95, 95% CI 0.79–0.99). sTF yielded comparable results: a level of at least 19.0 pg/ml was detected in all but two VAP patients (sensitivity of 0.78, 95% CI 0.50–0.92) and in five control patients (specificity of 0.74, 95% CI 0.55–0.87). sTM was not predictive of pneumonia.
https://static-content.springer.com/image/art%3A10.1007%2Fs00134-006-0167-9/MediaObjects/134_2006_167_Fig1_HTML.gif
Fig. 1

Receiver-operating characteristic curves of plasminogen activator 1 (PAI-1), soluble tissue factor (sTF) and soluble thrombomodulin (sTM) for the diagnosis of ventilator-associated pneumonia (VAP). AUC Area under the curve

Our data suggest that bronchoalveolar PAI-1 and sTF levels may be useful in identifying mechanically ventilated patients who develop VAP. However, it is important to mention that our control patients did not suffer from any pulmonary disease. Our data do not address the differentiation of VAP from other lung disease. In addition, intra- and inter-variability of the sampling method used in our study is largely unknown. It is possible that the levels of these biomarkers differ more when the sampling technique is performed by different investigators. Further studies are warranted to further analyse the value of these biomarkers in practice.

Copyright information

© Springer-Verlag 2006