- Email: [email protected]
Soluble Triggering Receptor Expressed on Myeloid Cell-1 Is Increased in Patients With Ventilator-Associated Pneumonia
Original Research CRITICAL CARE MEDICINE
Soluble Triggering Receptor Expressed on Myeloid Cell-1 Is Increased in Patients With Ventilator-Associated Pneumonia* A Preliminary Report Grigory Horonenko, DO; Jeffrey C. Hoyt, PhD; Richard A. Robbins, MD, FCCP; Clement U. Singarajah, MD, FCCP; Alp Umar, MD; Jenny Pattengill, BS; and John M. Hayden, PhD
Rationale: The diagnosis of ventilator-associated pneumonia (VAP) can be difficult. Soluble triggering receptor expressed on myeloid cell-1 (sTREM-1) has been reported to be elevated in BAL fluid from patients with VAP. Objectives: To evaluate the utility of sTREM-1 in the diagnosis of VAP in BAL fluid and the fluid collected in the expiratory trap from the ventilator, the exhaled ventilator condensate (EVC). Methods: We prospectively collected BAL fluid and EVC from 23 patients clinically suspected of having VAP. A sensitive enzyme-linked immunosorbent assay was developed to measure sTREM-1. The results derived from this assay were confirmed using an immunoblot technique. The presence of VAP was clinically determined using a modified clinical pulmonary infection score of > 6. Results: VAP was diagnosed in 14 of 23 patients. sTREM-1 was detected in the EVC from 11 of 14 subjects with VAP, but from only 1 of 9 subjects without VAP, and was significantly higher in the pneumonia patients and when expressed as picograms per milliliter or picograms per microgram protein (p ⴝ 0.005, both comparisons). In contrast, sTREM-1 was detected in the BAL fluid of all 14 VAP subjects but also in 8 of 9 subjects with no pneumonia, and did not differ in the VAP subjects compared to the nonpneumonia subjects when expressed as picrograms per milliliter or picograms per microgram protein (p > 0.05 both comparisons). Conclusion: sTREM-1 is detectable in EVC and may be useful in establishing or excluding the diagnosis of VAP. (CHEST 2007; 132:58 – 63) Key words: BAL; exhaled breath condensate; soluble triggering receptor expressed on myeloid cells; ventilator-associated pneumonia Abbreviations: BLOTTO ⫽ bovine lactotransfer technique optimizer; CPIS ⫽ clinical pulmonary infection score; EBC ⫽ exhaled breath condensate; ELISA ⫽ enzyme-linked immunosorbent assay; EVC ⫽ exhaled ventilator condensate; PBS ⫽ phosphate-buffered saline solution; rhTREM-1 ⫽ recombinant human triggering receptor expressed on myeloid cell-1; sTREM-1 ⫽ soluble triggering receptor expressed on myeloid cell-1; VAP ⫽ ventilator-associated pneumonia
of ventilator-associated pneumonia T he(VAP)diagnosis can be a clinical challenge. A clinical 1–3
diagnosis of pneumonia can be made when a new *From the Carl T. Hayden VA Medical Center, Good Samaritan Regional Medical Center and the Arizona Respiratory Center, Phoenix, AZ. The authors have no conflicts of interest to disclose. This work was supported by a grant from the Flight Attendants Medical Research Institute. This study was approved by the Carl T. Hayden VA Medical Center Institutional Review Board as Robbins 0004. Manuscript received November 13, 2006; revision accepted April 3, 2007. 58
radiographic infiltrate develops in a patient with fever, leukocytosis, purulent tracheal secretions, decreasing Pao2, and when microorganisms are isolated from the airways. These clinical parameters Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Richard A. Robbins, MD, Chief, Pulmonary and Critical Care Medicine, Carl T. Hayden VA Medical Center, 650 E Indian School Rd, Phoenix, AZ 85012; e-mail: Richard. [email protected] DOI: 10.1378/chest.06-2731 Original Research
have been used to develop the clinical pulmonary infection score (CPIS), which although imperfect is clinically useful in detecting the onset of VAP.4 Unfortunately, many noninfectious processes may be responsible for fever and new pulmonary infiltrates, and this, combined with the imperfection of the CPIS, have led to the use of other techniques such as BAL quantitative cultures and protected brush microbial cultures to demonstrate VAP.5– 8 However, these require invasive procedure and results can be delayed and often false negative because of the empiric administration of antibiotics. In this context, many biological markers have been studied in an effort to improve the diagnostic accuracy of VAP, most with disappointing results.9 –13 However, detection of soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) in BAL from patients who are receiving mechanical ventilation has been initially reported as a good clinical indicator of VAP with a sensitivity of 98% and a specificity of 90%.14 However, a more recent report15 suggests the utility of this marker as a single measurement in the diagnosis of VAP may be lower with a sensitivity of 75% and specificity of 84%. This is within the range reported in some series for the CPIS score.8 These reports illustrate the need for additional, more sensitive and specific biomarkers of VAP and the need for additional studies regarding the utility of sTREM-1 measurement in diagnosing VAP. Exhaled breath condensate (EBC), the condensate obtained from exhaled breath vapor, has been proposed as a noninvasive method of sampling the lower respiratory tract.16 Most lipids and proteins that can be detected in BAL can also be detected in EBC, albeit in lower concentrations. In intubated subjects, exhaled breath is collected in the collection trap in the expiratory ventilator line.17 We termed this collection exhaled ventilator condensate (EVC) and hypothesized that sTREM-1 is detectable in EVC collected from subjects with VAP and may provide for a useful clinical diagnostic for characterizing individuals with pneumonia. We collected BAL and EVC from subjects suspected of having VAP, and we successfully detected sTREM-1 in most of the EVC samples collected from these subjects but not in subjects who did not have pneumonia with a single exaction. These results suggest the potential utility to EVC sTREM-1 measurement in detecting VAP. Materials and Methods
mechanical ventilation and there was a clinical suspicion of infectious pneumonia in whom bronchoscopy was planned. Our Institutional Review Board approved the study but waived informed consent. Sample Collection BAL was performed through an endotracheal tube using five 20-mL aliquots. An additional aliquot was performed and sent to the clinical laboratory for quantitative culture. The remaining BAL fluid was centrifuged to remove cellular debris (10,000 revolutions per minute; 30 min; 4°C [Eppendorf 5415C; Eppendorf; Hamburg, Germany]),14 and the supernatant fluid and EVC were frozen until studied. One patient was studied twice. EVC was collected from the expiratory line at the time of the BAL. The EVC was removed from the expiratory line trap at the time of the completion of BAL. No effort was made to collect fluid for a standardized time, but the expiratory trap is emptied every 2 h, and therefore the sample was collected over ⬍ 2 h. Five normal, nonintubated subjects were also studied. Their EBC was collected by cold condensation of their exhaled breath for 15 min as previously described.16 We used the CPIS to classify the patients as having VAP for the purposes of this study.18 The CPIS was calculated as previously described using a 0 –2 scoring system using fever, leukocytosis, tracheal aspirates, oxygenation, radiographic infiltrates, and semiquantitative cultures of tracheal aspirates with Gram stain. We used a colony count of 103 organisms per milliliter as significant because of the technique of the BAL.19 A score ⬎ 6 was considered indicative of VAP. The duration of mechanical ventilation and the length and the outcome (death or discharge) of stay in the ICU were also recorded. Two intensivists (G.H. and R.A.R.) reviewed all medical records pertaining to the patient, and independently classified the diagnosis as VAP or no pneumonia based on the CPIS. A consensus concerning the diagnosis was achieved in all cases. The one patient studied twice was classified as VAP once and no pneumonia once. Antibodies Mouse anti-human TREM-1 (cat no. 841555), biotinylated goat anti-human TREM-1 (part 841556), and conjugated streptavidin horseradish-peroxidase (part 890803) were obtained from R&D Systems Inc. (Minneapolis, MN). Chemicals Bovine serum albumin (cat. no. A7030 –10G) was purchased from Sigma-Aldrich (St. Louis, MO). Concentrated buffered surfactant (cat. no. WA126), substrate solutions consisting of hydrogen peroxide and chromogen (color reagents A & B; cat no. DY999), and stop solution consisting of 2N sulfuric acid (cat. no. DY994) were obtained from R&D Systems Inc. Antigen Standard Calibration curves were prepared using recombinant human TREM-1 (rhTREM-1) [cat. no. 841557; R&D Systems Inc.] ranging from 0 to 2,000 pg/mL by serially diluting rhTREM-1 in a solution consisting of phosphate buffered saline solution (PBS) [pH 7.3] supplemented with 1% bovine serum antigen (SigmaAldrich).
Patients
Human sTREM Assay Procedure
Patients ⱖ 18 years old who were hospitalized in our medical ICU were prospectively enrolled in the study if they required
Flat-bottom 96-well immunomicrotiter plates (cat no. 468667; Nalge Nunc International; Rochester, NY) were coated with 100
www.chestjournal.org
CHEST / 132 / 1 / JULY, 2007
59
L of a 1/180 diluted mouse anti-human TREM-1 antibody (4.0 g/mL in PBS) and incubated overnight at room temperature. Each of the following incubation steps were preceded by washing the wells three times with 400 L of wash solution-buffered surfactant with preservatives using an aspirator and multichannel pipette. Subsequently, 300 L of reagent diluent blocking buffer (1% bovine serum antigen in PBS, pH 7.3) was added and incubated at room temperature for 1 h. Then, 100 L of rhTREM -1 (in duplicate) in concentrations between 0 and 4,000 pg/mL (diluted in reagent diluent) and 100 L of nondiluted samples (EVC and BAL) were incubated for 2 h at room temperature. Note that for the few BAL samples that were out of range of the assay, the samples were diluted 1:5 and 1:10 prior to reanalysis. After this incubation, 100 L of a 1/180 diluted biotinylated goat anti-human TREM-1 antibody (400 ng/mL in reagent diluent) were added and incubated for 2 h at room temperature. Subsequently, 100 L of a 1/200 diluted (in reagent diluent) streptavidin conjugated to horseradish-peroxidase was added and incubated in the dark for 20 min at room temperature. At this time, 100 L of a freshly prepared substrate solution of stabilized hydrogen peroxide and chromogen were added and incubated in the dark for 20 min at room temperature. To stop the reaction, 50 L of 2N sulfuric acid was added to each well. Absorbance was then measured at 450 nm with a wavelength correction set to 540 nm using an enzyme-linked immunosorbent assay (ELISA) plate reader (Optimax tunable microplate reader; Molecular Devices; Sunnyvale, CA). Subsequently, corresponding results were corrected for protein concentration (BCA Protein Assay; Pierce Chemical; Rockford, IL). For assay coefficient of variation determination, we pooled EVC and BAL samples from two to four patients who were screened for sTREM-1 concentrations previously. Multiple aliquots of EVC (average sTREM-1 concentration approximately 176 pg/mL) and BAL samples containing either low (approximately 98 pg/mL) or high (approximately 1,352 pg/mL) levels of sTREM-1 were frozen at ⫺ 80°C and used for coefficient of variation determination obtained from at least three separate assays. These control samples were used only once per assay and were not subjected to multiple freeze/thaw cycles. Average intraassay and interassay CV were 11.5% and 16.6%, 9.5% and 14.1%, and 4.0% and 8.6% for EVC control samples, and for low-containing and high-containing BAL control samples, respectively. Dot-Blot Assay Procedure In a subset of samples, EVC was added (1 g of protein) to wells of a dot-blot manifold containing a prewetted 0.45-m polyvinylidene difluoride membranes (BioRad; Hercules, CA). Samples were then incubated at room temperature for 1 to 2 h, and residual liquid was removed by application of a gentle vacuum to the manifold at this time. The membrane was placed in 100 mL of a solution containing 5% nonfat dried milk in Tris-buffered saline solution (20 mmol/L Tris base, 180 mmol/L NaCl) [bovine lactotransfer technique optimizer (BLOTTO)] and
allowed to incubate overnight at 25°C. The next day, the membrane was placed in 5 mL of BLOTTO with 5 L of mouse anti-human TREM-1 and incubated with gentle rocking for 2 h. Membrane was washed with 100 mL of BLOTTO for 12 min. This process was repeated twice for a total of three washings. Five milliliters of BLOTTO were added with 5 L of goat anti-mouse IgG alkaline phosphatase conjugate and gently rocked for 2 h. This solution was washed with 100 mL of BLOTTO for 12 min. This process was repeated twice for a total of three washings. The membrane was rinsed with distilled water, and incubated in 100 mmol/L Tris (pH 9.1) for 1 min. Substrate was prepared by adding a nitroblue tetrazolium/bromochloroindolyl phosphate tablet to 10 mL of distilled H2O and added to the rinsed membranes. When colored dots appeared, the substrate solution was drained and the membrane was rinsed with water and then dried. Quantification of colored dots was determined by scanning densitometry. Statistical Analysis sTREM-1 levels in BAL and EVC were expressed as mean ⫾ SEM. Variables were evaluated for an association with the diagnosis by Pearson 2 test for categorical data and Wilcoxon-Mann-Whitney rank-sum test for numerical data; p ⱕ 0.05 was defined as significant.
Results Characteristics of the Patients The characteristics of the overall study groups are shown in Table 1. All patients were men, with seven smokers (four in the pneumonia group), seven former smokers (four in the pneumonia group with one patient studied twice, once in each group), and eight nonsmokers (six in the pneumonia group). Most of the patients had an associated condition, and four patients (17%) had a history of COPD (two in the pneumonia group, and two in the nonpneumonia group). Other major diagnosis included congestive heart failure in six patients (four in the pneumonia group, and two in the nonpneumonia group); cancer in five patients (three in the pneumonia group, and two in the nonpneumonia group); perforated viscus in three patients (three in the pneumonia group, and none in the nonpneumonia group); ARDS in two patients (none in the pneumonia group, and two in the nonpneumonia group); renal failure in two patients (none in the pneumonia, and two in the nonpneumonia group); peripheral vascular disease in
Table 1—Patient Characteristics* Variables
CPIS Score
Age, yr
BAL Culture Positive, %
Antibiotics, %
Duration of Mechanical Ventilation, d
ICU Stay, d
Deaths, %
No pneumonia Pneumonia Significance
4.4 ⫾ 0.4 7.9 ⫾ 0.2 Yes
69.0 ⫾ 3.3 72.6 ⫾ 2.2 No
0 64 Yes
100 71 No
5.1 ⫾ 0.9 7.7 ⫾ 2.0 No
5.9 ⫾ 0.9 8.4 ⫾ 1.9 No
50 29 No
*Data are presented as mean ⫾ SEM unless otherwise indicated. 60
Original Research
The ELISA described herein allowed for the sensitive detection of sTREM-1 (range, 7 to 4,000 pg/mL) in both human EVC and BAL samples. In a separate series of experiments, we also wished to compare results obtained from the ELISA vs those obtained from dot-blot techniques. In these experiments, a subset of EVC samples were collected from 13 individuals, assayed for sTREM-1 by ELISA, and corrected for protein concentration in the sample. When compared to the image density units obtained from the dot blots, correlation analysis revealed a positive association between both of these types of analyses (r ⫽ 0.59; p ⬍ 0.05). In the EVC samples, the ELISA demonstrated that sTREM-1 was detected in the samples from 10 of 14 subjects with VAP but in only 1 of 9 subjects without VAP (Fig 1; p ⬍ 0.01, 2 analysis). The one nonpneumonia patient who had detectable sTREM-1 was just above the detectable level. The levels of sTREM-1 differed between the pneumonia and nonpneumonia groups when it was assumed that a nondetectable level was 7 pg/mL (the lower limit of detection) [p ⬍ 0.005, both comparisons by Wilcoxon-Mann-Whitney rank-sum test]. The patient studied twice was initially classified as not having pneumonia and had a nondetectable sTREM-1 in the EVC. However, at the time of the second studies, the patient studied twice was classified as having pneumonia and the second EVC collection did have www.chestjournal.org
5
sTREM (pg/µg protein)
sTREM (pg/ml)
A.
300
200
100
0
B.
4
3
2
1
0
No Pneumonia
No Pneumonia
Pneumonia
Pneumonia
Figure 1. sTREM-1 concentration in EVC (left, A) and normalized for total protein concentration (right, B) in the experimental sample.
detectable sTREM-1. In addition, sTREM-1 was not detectable in the EBC obtained from five normal subjects. In contrast, sTREM-1 was detected in the BAL fluid of all 14 VAP subjects, but also in of 9 of 10 subjects with no pneumonia (Fig 2, p ⬎ 0.05). Furthermore, the concentration in sTREM-1 did not differ in the VAP subjects compared to the nonpneumonia subjects (403 ⫾ 140 pg/mL vs 309 ⫾ 146 pg/mL, p ⫽ 0.4767) even when corrected for the variable protein concentration displayed in the BAL fluid (0.7 ⫾ 0.3 pg vs 0.4 ⫾ 0.4 pg of sTREM-1/g total BAL protein, p ⫽ 1.000). Smoking status did not affect sTREM-1 levels in either EVC or BAL fluid, and pack-years of smoking did not correlate with sTREM-1 levels (data not shown). Discussion This study reveals several important findings. First, sTREM-1 is detectable in EVC collected from the expiratory line of patients receiving mechanical
2500
A.
4
2000
sTREM (pg/µg protein)
sTREM-1 Detection
400
sTREM (pg/ml)
two patients (one in each group); one patient with pulmonary embolism in the nonpneumonia group; and one patient with diverticulitis in the pneumonia group. The overall ICU mortality rate of 38% was in agreement with another study.3 Although the mortality rate of 50% was higher in the nonpneumonia group, it did not reach statistical significance. Consistent with the way the groups were defined, the CPIS was higher in patients with VAP than among patients without pneumonia (p ⬍ 0.01). Microbial species grew to a clinically significant concentration (⬎ 103 cfu/mL) in specimens of BAL fluid in 9 of the 14 patients with pneumonia but in none of the 10 patients without pneumonia. Positive culture findings included five with Staphylococcus aureus, two with Enterococcus, one with Haemophilus influenzae, and one with Pseudomonas aeruginosa. Most of the patients were receiving antibiotics at the time of the BAL and EVC collection (83%); and although the percentage of patients without pneumonia receiving antibiotics was higher, this did not reach statistical significance (p ⬎ 0.05).
1500
1000
500
0
B.
3
2
1
0
No Pneumonia
Pneumonia
No Pneumonia
Pneumonia
Figure 2. sTREM-1 concentration in BAL fluid (left, A) and normalized for total protein concentration (right, B) in the experimental sample. CHEST / 132 / 1 / JULY, 2007
61
ventilation. Second, consistent with a previous study16 examining protein levels in EVC, the level of sTREM-1 is lower than those in BAL fluid. Third, the detection of sTREM-1 in EVC may have utility in the detection of VAP. Fourth, sTREM-1 detection in BAL did not appear to be as useful an indicator of VAP, as demonstrated in the small subject population examined in this study. Gibot et al14 reported a rapid detection of sTREM-1 in BAL fluid may be useful in establishing or excluding the diagnosis of bacterial or fungal pneumonia. However, Determann et al15 reported that the initial sTREM-1 levels in BAL were only 75% sensitive and 84% specific for VAP when a cutoff of 200 pg/mL was used, which approximates the CPIS score in some series.8 Furthermore, Determann et al15 used an ELISA, and values in that study approximated the values in our study (mean sTREM-1, 894 pg/mL in VAP vs 403 pg/mL in the present study). However, in Determann et al,15 an increasing BAL sTREM-1 sampled over time, regardless of the initial values, did predict pneumonia. The data reported in this manuscript is more consistent with the study of Determann et al.15 However, serial BAL and EVC collections were not performed in our study. Like other similar cell-surface receptors, TREM-1 has a short intracellular domain and, when bound to its still unidentified ligand, it associates with a signaltransduction molecule called DAP12 and triggers the secretion of inflammatory cytokines, amplifying the host response to bacterial stimuli.20 TREM-1 is greatly up-regulated in the presence of bacteria such as P aeruginosa and S aureus or fungi such as Aspergillus fumigatus, and the soluble form is found both in cell culture media, peritoneal lavage fluid, and tissue samples from patients who have infection with these microorganisms.21 In contrast, sTREM-1 was not up-regulated in samples from patients with noninfectious inflammatory disorders, such as psoriasis, indicating the specific involvement of this receptor occurs only in the case of infection. The results of this study are consistent with these observations. However, patients with culture-positive VAP did not have a significantly higher sTREM-1 level than those with culture-negative VAP (p ⬎ 0.1). One patient in the nonpneumonia group was culture positive for influenza A in BAL fluid. However, it is unclear whether the influenza played a role in the patient’s death or was an incidental finding. A potential limitation of this study is that there is no “gold standard” for the diagnosis of VAP. We used the CPIS score when it became apparent that the level of BAL sTREM-1 was not separating patients with VAP from those without pneumonia. However, clinical criteria have been reported to have approxi62
mately 60 to 80% sensitivity and specificity in the diagnosis of VAP.4,8,22,23 Microbiologic documentation of infection improves these numbers. Therefore, it seems likely that although most of our patients were correctly classified, some of our patients may have been misclassified. Our results demonstrate that detection of sTREM-1 in EVC may improve the ability of clinicians to differentiate patients with VAP from those without pneumonia. This ability should be especially useful in patients for whom the diagnosis is not clinically apparent. Use of EVC compared to BAL may have the additional advantage of ease of collection. The ELISA method used is rapid and relatively inexpensive. The use of this test to detect the presence of sTREM-1 in EVC may lead to more accurate diagnoses of pneumonia in patients who are receiving mechanical ventilation. References 1 Chastre J, Fagon JY. Ventilator-associated pneumonia. Am J Respir Crit Care Med 2002; 165:867–903 2 Fagon JY, Chastre J, Hance AJ, et al. Evaluation of clinical judgment in the identification and treatment of nosocomial pneumonia in ventilated patients. Chest 1993; 103:547–553 3 Wunderink RG. Mortality and the diagnosis of ventilatorassociated pneumonia: a new direction. Am J Respir Crit Care Med 1998; 157:349 –350 4 Fabregas N, Ewig S, Torres A, et al. Clinical diagnosis of ventilator associated pneumonia revisited: comparative validation using immediate post-mortem lung biopsies. Thorax 1999; 54:867– 873 5 Campbell GD Jr. Blinded invasive diagnostic procedures in ventilator-associated pneumonia. Chest 2000; 117:207S–211S 6 Fagon JY, Chastre J, Wolff M, et al. Invasive and noninvasive strategies for management of suspected ventilator-associated pneumonia: a randomized trial. Ann Intern Med 2000; 132: 621– 630 7 Meduri GU, Wunderink RG, Leeper KV, et al. Management of bacterial pneumonia in ventilated patients: protected bronchoalveolar lavage as a diagnostic tool. Chest 1992; 101:500 –508 8 Papazian L, Thomas P, Garbe L, et al. Bronchoscopic or blind sampling techniques for the diagnosis of ventilator-associated pneumonia. Am J Respir Crit Care Med 1995; 152:1982–1991 9 Brunkhorst FM, Al-Nawas B, Krummenauer F, et al. Procalcitonin, C-reactive protein and APACHE II score for risk evaluation in patients with severe pneumonia. Clin Microbiol Infect 2002; 8:93–100 10 Duflo F, Debon R, Monneret G, et al. Alveolar and serum procalcitonin: diagnostic and prognostic value in ventilatorassociated pneumonia. Anesthesiology 2002; 96:74 –79 11 Meduri GU, Headley S, Kohler G, et al. Persistent elevation of inflammatory cytokines predicts a poor outcome in ARDS: plasma IL-1 and IL-6 levels are consistent and efficient predictors of outcome over time. Chest 1995; 107:1062–1073 12 Monton C, Torres A, El-Ebiary M, et al. Cytokine expression in severe pneumonia: a bronchoalveolar lavage study. Crit Care Med 1999; 27:1745–1753 13 Wu CL, Lee YL, Chang KM, et al. Bronchoalveolar interleukin-1 beta: a marker of bacterial burden in mechanically ventilated patients with community-acquired pneumonia. Original Research
Crit Care Med 2003; 31:812– 817 14 Gibot S, Cravoisy A, Levy B, et al. Soluble triggering receptor expressed on myeloid cells and the diagnosis of pneumonia. N Engl J Med 2004; 350:451– 458 15 Determann RM, Millo JL, Gibot S, et al. Serial changes in soluble triggering receptor expressed on myeloid cells in the lung during development of ventilator-associated pneumonia. Intensive Care Med 2005; 31:1495–1500 16 Mutlu GM, Garey KW, Robbins RA, et al. Collection and analysis of exhaled breath condensate in humans. Am J Respir Crit Care Med 2001; 164:731–737 17 Carpenter CT, Price PV, Christman BW. Exhaled breath condensate isoprostanes are elevated in patients with acute lung injury or ARDS. Chest 1998; 114:1653–1659 18 Fartoukh M, Maitre B, Honore S, et al. Diagnosing pneumonia during mechanical ventilation: the clinical pulmonary infection score revisited. Am J Respir Crit Care Med 2003; 168:173–179
www.chestjournal.org
19 Cantral DE, Tape TG, Reed EC, et al. Quantitative culture of bronchoalveolar lavage fluid for the diagnosis of bacterial pneumonia. Am J Med 1993; 95:601– 607 20 Bouchon A, Dietrich J, Colonna M. Cutting edge: inflammatory responses can be triggered by TREM-1, a novel receptor expressed on neutrophils and monocytes. J Immunol 2000; 164:4991– 4995 21 Bouchon A, Facchetti F, Weigand MA, et al. TREM-1 amplifies inflammation and is a crucial mediator of septic shock. Nature 2001; 410:1103–1107 22 Pugin J, Auckenthaler R, Mili N, et al. Diagnosis of ventilatorassociated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic “blind” bronchoalveolar lavage fluid. Am Rev Respir Dis 1991; 143:1121–1129 23 Schurink CA, Van Nieuwenhoven CA, Jacobs JA, et al. Clinical pulmonary infection score for ventilator-associated pneumonia: accuracy and inter-observer variability. Intensive Care Med 2004; 30:217–224
CHEST / 132 / 1 / JULY, 2007
63
Soluble Triggering Receptor Expressed on Myeloid Cell-1 Is Increased in Patients With Ventilator-Associated Pneumonia* A Preliminary Report Grigory Horonenko, DO; Jeffrey C. Hoyt, PhD; Richard A. Robbins, MD, FCCP; Clement U. Singarajah, MD, FCCP; Alp Umar, MD; Jenny Pattengill, BS; and John M. Hayden, PhD
Rationale: The diagnosis of ventilator-associated pneumonia (VAP) can be difficult. Soluble triggering receptor expressed on myeloid cell-1 (sTREM-1) has been reported to be elevated in BAL fluid from patients with VAP. Objectives: To evaluate the utility of sTREM-1 in the diagnosis of VAP in BAL fluid and the fluid collected in the expiratory trap from the ventilator, the exhaled ventilator condensate (EVC). Methods: We prospectively collected BAL fluid and EVC from 23 patients clinically suspected of having VAP. A sensitive enzyme-linked immunosorbent assay was developed to measure sTREM-1. The results derived from this assay were confirmed using an immunoblot technique. The presence of VAP was clinically determined using a modified clinical pulmonary infection score of > 6. Results: VAP was diagnosed in 14 of 23 patients. sTREM-1 was detected in the EVC from 11 of 14 subjects with VAP, but from only 1 of 9 subjects without VAP, and was significantly higher in the pneumonia patients and when expressed as picograms per milliliter or picograms per microgram protein (p ⴝ 0.005, both comparisons). In contrast, sTREM-1 was detected in the BAL fluid of all 14 VAP subjects but also in 8 of 9 subjects with no pneumonia, and did not differ in the VAP subjects compared to the nonpneumonia subjects when expressed as picrograms per milliliter or picograms per microgram protein (p > 0.05 both comparisons). Conclusion: sTREM-1 is detectable in EVC and may be useful in establishing or excluding the diagnosis of VAP. (CHEST 2007; 132:58 – 63) Key words: BAL; exhaled breath condensate; soluble triggering receptor expressed on myeloid cells; ventilator-associated pneumonia Abbreviations: BLOTTO ⫽ bovine lactotransfer technique optimizer; CPIS ⫽ clinical pulmonary infection score; EBC ⫽ exhaled breath condensate; ELISA ⫽ enzyme-linked immunosorbent assay; EVC ⫽ exhaled ventilator condensate; PBS ⫽ phosphate-buffered saline solution; rhTREM-1 ⫽ recombinant human triggering receptor expressed on myeloid cell-1; sTREM-1 ⫽ soluble triggering receptor expressed on myeloid cell-1; VAP ⫽ ventilator-associated pneumonia
of ventilator-associated pneumonia T he(VAP)diagnosis can be a clinical challenge. A clinical 1–3
diagnosis of pneumonia can be made when a new *From the Carl T. Hayden VA Medical Center, Good Samaritan Regional Medical Center and the Arizona Respiratory Center, Phoenix, AZ. The authors have no conflicts of interest to disclose. This work was supported by a grant from the Flight Attendants Medical Research Institute. This study was approved by the Carl T. Hayden VA Medical Center Institutional Review Board as Robbins 0004. Manuscript received November 13, 2006; revision accepted April 3, 2007. 58
radiographic infiltrate develops in a patient with fever, leukocytosis, purulent tracheal secretions, decreasing Pao2, and when microorganisms are isolated from the airways. These clinical parameters Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Richard A. Robbins, MD, Chief, Pulmonary and Critical Care Medicine, Carl T. Hayden VA Medical Center, 650 E Indian School Rd, Phoenix, AZ 85012; e-mail: Richard. [email protected] DOI: 10.1378/chest.06-2731 Original Research
have been used to develop the clinical pulmonary infection score (CPIS), which although imperfect is clinically useful in detecting the onset of VAP.4 Unfortunately, many noninfectious processes may be responsible for fever and new pulmonary infiltrates, and this, combined with the imperfection of the CPIS, have led to the use of other techniques such as BAL quantitative cultures and protected brush microbial cultures to demonstrate VAP.5– 8 However, these require invasive procedure and results can be delayed and often false negative because of the empiric administration of antibiotics. In this context, many biological markers have been studied in an effort to improve the diagnostic accuracy of VAP, most with disappointing results.9 –13 However, detection of soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) in BAL from patients who are receiving mechanical ventilation has been initially reported as a good clinical indicator of VAP with a sensitivity of 98% and a specificity of 90%.14 However, a more recent report15 suggests the utility of this marker as a single measurement in the diagnosis of VAP may be lower with a sensitivity of 75% and specificity of 84%. This is within the range reported in some series for the CPIS score.8 These reports illustrate the need for additional, more sensitive and specific biomarkers of VAP and the need for additional studies regarding the utility of sTREM-1 measurement in diagnosing VAP. Exhaled breath condensate (EBC), the condensate obtained from exhaled breath vapor, has been proposed as a noninvasive method of sampling the lower respiratory tract.16 Most lipids and proteins that can be detected in BAL can also be detected in EBC, albeit in lower concentrations. In intubated subjects, exhaled breath is collected in the collection trap in the expiratory ventilator line.17 We termed this collection exhaled ventilator condensate (EVC) and hypothesized that sTREM-1 is detectable in EVC collected from subjects with VAP and may provide for a useful clinical diagnostic for characterizing individuals with pneumonia. We collected BAL and EVC from subjects suspected of having VAP, and we successfully detected sTREM-1 in most of the EVC samples collected from these subjects but not in subjects who did not have pneumonia with a single exaction. These results suggest the potential utility to EVC sTREM-1 measurement in detecting VAP. Materials and Methods
mechanical ventilation and there was a clinical suspicion of infectious pneumonia in whom bronchoscopy was planned. Our Institutional Review Board approved the study but waived informed consent. Sample Collection BAL was performed through an endotracheal tube using five 20-mL aliquots. An additional aliquot was performed and sent to the clinical laboratory for quantitative culture. The remaining BAL fluid was centrifuged to remove cellular debris (10,000 revolutions per minute; 30 min; 4°C [Eppendorf 5415C; Eppendorf; Hamburg, Germany]),14 and the supernatant fluid and EVC were frozen until studied. One patient was studied twice. EVC was collected from the expiratory line at the time of the BAL. The EVC was removed from the expiratory line trap at the time of the completion of BAL. No effort was made to collect fluid for a standardized time, but the expiratory trap is emptied every 2 h, and therefore the sample was collected over ⬍ 2 h. Five normal, nonintubated subjects were also studied. Their EBC was collected by cold condensation of their exhaled breath for 15 min as previously described.16 We used the CPIS to classify the patients as having VAP for the purposes of this study.18 The CPIS was calculated as previously described using a 0 –2 scoring system using fever, leukocytosis, tracheal aspirates, oxygenation, radiographic infiltrates, and semiquantitative cultures of tracheal aspirates with Gram stain. We used a colony count of 103 organisms per milliliter as significant because of the technique of the BAL.19 A score ⬎ 6 was considered indicative of VAP. The duration of mechanical ventilation and the length and the outcome (death or discharge) of stay in the ICU were also recorded. Two intensivists (G.H. and R.A.R.) reviewed all medical records pertaining to the patient, and independently classified the diagnosis as VAP or no pneumonia based on the CPIS. A consensus concerning the diagnosis was achieved in all cases. The one patient studied twice was classified as VAP once and no pneumonia once. Antibodies Mouse anti-human TREM-1 (cat no. 841555), biotinylated goat anti-human TREM-1 (part 841556), and conjugated streptavidin horseradish-peroxidase (part 890803) were obtained from R&D Systems Inc. (Minneapolis, MN). Chemicals Bovine serum albumin (cat. no. A7030 –10G) was purchased from Sigma-Aldrich (St. Louis, MO). Concentrated buffered surfactant (cat. no. WA126), substrate solutions consisting of hydrogen peroxide and chromogen (color reagents A & B; cat no. DY999), and stop solution consisting of 2N sulfuric acid (cat. no. DY994) were obtained from R&D Systems Inc. Antigen Standard Calibration curves were prepared using recombinant human TREM-1 (rhTREM-1) [cat. no. 841557; R&D Systems Inc.] ranging from 0 to 2,000 pg/mL by serially diluting rhTREM-1 in a solution consisting of phosphate buffered saline solution (PBS) [pH 7.3] supplemented with 1% bovine serum antigen (SigmaAldrich).
Patients
Human sTREM Assay Procedure
Patients ⱖ 18 years old who were hospitalized in our medical ICU were prospectively enrolled in the study if they required
Flat-bottom 96-well immunomicrotiter plates (cat no. 468667; Nalge Nunc International; Rochester, NY) were coated with 100
www.chestjournal.org
CHEST / 132 / 1 / JULY, 2007
59
L of a 1/180 diluted mouse anti-human TREM-1 antibody (4.0 g/mL in PBS) and incubated overnight at room temperature. Each of the following incubation steps were preceded by washing the wells three times with 400 L of wash solution-buffered surfactant with preservatives using an aspirator and multichannel pipette. Subsequently, 300 L of reagent diluent blocking buffer (1% bovine serum antigen in PBS, pH 7.3) was added and incubated at room temperature for 1 h. Then, 100 L of rhTREM -1 (in duplicate) in concentrations between 0 and 4,000 pg/mL (diluted in reagent diluent) and 100 L of nondiluted samples (EVC and BAL) were incubated for 2 h at room temperature. Note that for the few BAL samples that were out of range of the assay, the samples were diluted 1:5 and 1:10 prior to reanalysis. After this incubation, 100 L of a 1/180 diluted biotinylated goat anti-human TREM-1 antibody (400 ng/mL in reagent diluent) were added and incubated for 2 h at room temperature. Subsequently, 100 L of a 1/200 diluted (in reagent diluent) streptavidin conjugated to horseradish-peroxidase was added and incubated in the dark for 20 min at room temperature. At this time, 100 L of a freshly prepared substrate solution of stabilized hydrogen peroxide and chromogen were added and incubated in the dark for 20 min at room temperature. To stop the reaction, 50 L of 2N sulfuric acid was added to each well. Absorbance was then measured at 450 nm with a wavelength correction set to 540 nm using an enzyme-linked immunosorbent assay (ELISA) plate reader (Optimax tunable microplate reader; Molecular Devices; Sunnyvale, CA). Subsequently, corresponding results were corrected for protein concentration (BCA Protein Assay; Pierce Chemical; Rockford, IL). For assay coefficient of variation determination, we pooled EVC and BAL samples from two to four patients who were screened for sTREM-1 concentrations previously. Multiple aliquots of EVC (average sTREM-1 concentration approximately 176 pg/mL) and BAL samples containing either low (approximately 98 pg/mL) or high (approximately 1,352 pg/mL) levels of sTREM-1 were frozen at ⫺ 80°C and used for coefficient of variation determination obtained from at least three separate assays. These control samples were used only once per assay and were not subjected to multiple freeze/thaw cycles. Average intraassay and interassay CV were 11.5% and 16.6%, 9.5% and 14.1%, and 4.0% and 8.6% for EVC control samples, and for low-containing and high-containing BAL control samples, respectively. Dot-Blot Assay Procedure In a subset of samples, EVC was added (1 g of protein) to wells of a dot-blot manifold containing a prewetted 0.45-m polyvinylidene difluoride membranes (BioRad; Hercules, CA). Samples were then incubated at room temperature for 1 to 2 h, and residual liquid was removed by application of a gentle vacuum to the manifold at this time. The membrane was placed in 100 mL of a solution containing 5% nonfat dried milk in Tris-buffered saline solution (20 mmol/L Tris base, 180 mmol/L NaCl) [bovine lactotransfer technique optimizer (BLOTTO)] and
allowed to incubate overnight at 25°C. The next day, the membrane was placed in 5 mL of BLOTTO with 5 L of mouse anti-human TREM-1 and incubated with gentle rocking for 2 h. Membrane was washed with 100 mL of BLOTTO for 12 min. This process was repeated twice for a total of three washings. Five milliliters of BLOTTO were added with 5 L of goat anti-mouse IgG alkaline phosphatase conjugate and gently rocked for 2 h. This solution was washed with 100 mL of BLOTTO for 12 min. This process was repeated twice for a total of three washings. The membrane was rinsed with distilled water, and incubated in 100 mmol/L Tris (pH 9.1) for 1 min. Substrate was prepared by adding a nitroblue tetrazolium/bromochloroindolyl phosphate tablet to 10 mL of distilled H2O and added to the rinsed membranes. When colored dots appeared, the substrate solution was drained and the membrane was rinsed with water and then dried. Quantification of colored dots was determined by scanning densitometry. Statistical Analysis sTREM-1 levels in BAL and EVC were expressed as mean ⫾ SEM. Variables were evaluated for an association with the diagnosis by Pearson 2 test for categorical data and Wilcoxon-Mann-Whitney rank-sum test for numerical data; p ⱕ 0.05 was defined as significant.
Results Characteristics of the Patients The characteristics of the overall study groups are shown in Table 1. All patients were men, with seven smokers (four in the pneumonia group), seven former smokers (four in the pneumonia group with one patient studied twice, once in each group), and eight nonsmokers (six in the pneumonia group). Most of the patients had an associated condition, and four patients (17%) had a history of COPD (two in the pneumonia group, and two in the nonpneumonia group). Other major diagnosis included congestive heart failure in six patients (four in the pneumonia group, and two in the nonpneumonia group); cancer in five patients (three in the pneumonia group, and two in the nonpneumonia group); perforated viscus in three patients (three in the pneumonia group, and none in the nonpneumonia group); ARDS in two patients (none in the pneumonia group, and two in the nonpneumonia group); renal failure in two patients (none in the pneumonia, and two in the nonpneumonia group); peripheral vascular disease in
Table 1—Patient Characteristics* Variables
CPIS Score
Age, yr
BAL Culture Positive, %
Antibiotics, %
Duration of Mechanical Ventilation, d
ICU Stay, d
Deaths, %
No pneumonia Pneumonia Significance
4.4 ⫾ 0.4 7.9 ⫾ 0.2 Yes
69.0 ⫾ 3.3 72.6 ⫾ 2.2 No
0 64 Yes
100 71 No
5.1 ⫾ 0.9 7.7 ⫾ 2.0 No
5.9 ⫾ 0.9 8.4 ⫾ 1.9 No
50 29 No
*Data are presented as mean ⫾ SEM unless otherwise indicated. 60
Original Research
The ELISA described herein allowed for the sensitive detection of sTREM-1 (range, 7 to 4,000 pg/mL) in both human EVC and BAL samples. In a separate series of experiments, we also wished to compare results obtained from the ELISA vs those obtained from dot-blot techniques. In these experiments, a subset of EVC samples were collected from 13 individuals, assayed for sTREM-1 by ELISA, and corrected for protein concentration in the sample. When compared to the image density units obtained from the dot blots, correlation analysis revealed a positive association between both of these types of analyses (r ⫽ 0.59; p ⬍ 0.05). In the EVC samples, the ELISA demonstrated that sTREM-1 was detected in the samples from 10 of 14 subjects with VAP but in only 1 of 9 subjects without VAP (Fig 1; p ⬍ 0.01, 2 analysis). The one nonpneumonia patient who had detectable sTREM-1 was just above the detectable level. The levels of sTREM-1 differed between the pneumonia and nonpneumonia groups when it was assumed that a nondetectable level was 7 pg/mL (the lower limit of detection) [p ⬍ 0.005, both comparisons by Wilcoxon-Mann-Whitney rank-sum test]. The patient studied twice was initially classified as not having pneumonia and had a nondetectable sTREM-1 in the EVC. However, at the time of the second studies, the patient studied twice was classified as having pneumonia and the second EVC collection did have www.chestjournal.org
5
sTREM (pg/µg protein)
sTREM (pg/ml)
A.
300
200
100
0
B.
4
3
2
1
0
No Pneumonia
No Pneumonia
Pneumonia
Pneumonia
Figure 1. sTREM-1 concentration in EVC (left, A) and normalized for total protein concentration (right, B) in the experimental sample.
detectable sTREM-1. In addition, sTREM-1 was not detectable in the EBC obtained from five normal subjects. In contrast, sTREM-1 was detected in the BAL fluid of all 14 VAP subjects, but also in of 9 of 10 subjects with no pneumonia (Fig 2, p ⬎ 0.05). Furthermore, the concentration in sTREM-1 did not differ in the VAP subjects compared to the nonpneumonia subjects (403 ⫾ 140 pg/mL vs 309 ⫾ 146 pg/mL, p ⫽ 0.4767) even when corrected for the variable protein concentration displayed in the BAL fluid (0.7 ⫾ 0.3 pg vs 0.4 ⫾ 0.4 pg of sTREM-1/g total BAL protein, p ⫽ 1.000). Smoking status did not affect sTREM-1 levels in either EVC or BAL fluid, and pack-years of smoking did not correlate with sTREM-1 levels (data not shown). Discussion This study reveals several important findings. First, sTREM-1 is detectable in EVC collected from the expiratory line of patients receiving mechanical
2500
A.
4
2000
sTREM (pg/µg protein)
sTREM-1 Detection
400
sTREM (pg/ml)
two patients (one in each group); one patient with pulmonary embolism in the nonpneumonia group; and one patient with diverticulitis in the pneumonia group. The overall ICU mortality rate of 38% was in agreement with another study.3 Although the mortality rate of 50% was higher in the nonpneumonia group, it did not reach statistical significance. Consistent with the way the groups were defined, the CPIS was higher in patients with VAP than among patients without pneumonia (p ⬍ 0.01). Microbial species grew to a clinically significant concentration (⬎ 103 cfu/mL) in specimens of BAL fluid in 9 of the 14 patients with pneumonia but in none of the 10 patients without pneumonia. Positive culture findings included five with Staphylococcus aureus, two with Enterococcus, one with Haemophilus influenzae, and one with Pseudomonas aeruginosa. Most of the patients were receiving antibiotics at the time of the BAL and EVC collection (83%); and although the percentage of patients without pneumonia receiving antibiotics was higher, this did not reach statistical significance (p ⬎ 0.05).
1500
1000
500
0
B.
3
2
1
0
No Pneumonia
Pneumonia
No Pneumonia
Pneumonia
Figure 2. sTREM-1 concentration in BAL fluid (left, A) and normalized for total protein concentration (right, B) in the experimental sample. CHEST / 132 / 1 / JULY, 2007
61
ventilation. Second, consistent with a previous study16 examining protein levels in EVC, the level of sTREM-1 is lower than those in BAL fluid. Third, the detection of sTREM-1 in EVC may have utility in the detection of VAP. Fourth, sTREM-1 detection in BAL did not appear to be as useful an indicator of VAP, as demonstrated in the small subject population examined in this study. Gibot et al14 reported a rapid detection of sTREM-1 in BAL fluid may be useful in establishing or excluding the diagnosis of bacterial or fungal pneumonia. However, Determann et al15 reported that the initial sTREM-1 levels in BAL were only 75% sensitive and 84% specific for VAP when a cutoff of 200 pg/mL was used, which approximates the CPIS score in some series.8 Furthermore, Determann et al15 used an ELISA, and values in that study approximated the values in our study (mean sTREM-1, 894 pg/mL in VAP vs 403 pg/mL in the present study). However, in Determann et al,15 an increasing BAL sTREM-1 sampled over time, regardless of the initial values, did predict pneumonia. The data reported in this manuscript is more consistent with the study of Determann et al.15 However, serial BAL and EVC collections were not performed in our study. Like other similar cell-surface receptors, TREM-1 has a short intracellular domain and, when bound to its still unidentified ligand, it associates with a signaltransduction molecule called DAP12 and triggers the secretion of inflammatory cytokines, amplifying the host response to bacterial stimuli.20 TREM-1 is greatly up-regulated in the presence of bacteria such as P aeruginosa and S aureus or fungi such as Aspergillus fumigatus, and the soluble form is found both in cell culture media, peritoneal lavage fluid, and tissue samples from patients who have infection with these microorganisms.21 In contrast, sTREM-1 was not up-regulated in samples from patients with noninfectious inflammatory disorders, such as psoriasis, indicating the specific involvement of this receptor occurs only in the case of infection. The results of this study are consistent with these observations. However, patients with culture-positive VAP did not have a significantly higher sTREM-1 level than those with culture-negative VAP (p ⬎ 0.1). One patient in the nonpneumonia group was culture positive for influenza A in BAL fluid. However, it is unclear whether the influenza played a role in the patient’s death or was an incidental finding. A potential limitation of this study is that there is no “gold standard” for the diagnosis of VAP. We used the CPIS score when it became apparent that the level of BAL sTREM-1 was not separating patients with VAP from those without pneumonia. However, clinical criteria have been reported to have approxi62
mately 60 to 80% sensitivity and specificity in the diagnosis of VAP.4,8,22,23 Microbiologic documentation of infection improves these numbers. Therefore, it seems likely that although most of our patients were correctly classified, some of our patients may have been misclassified. Our results demonstrate that detection of sTREM-1 in EVC may improve the ability of clinicians to differentiate patients with VAP from those without pneumonia. This ability should be especially useful in patients for whom the diagnosis is not clinically apparent. Use of EVC compared to BAL may have the additional advantage of ease of collection. The ELISA method used is rapid and relatively inexpensive. The use of this test to detect the presence of sTREM-1 in EVC may lead to more accurate diagnoses of pneumonia in patients who are receiving mechanical ventilation. References 1 Chastre J, Fagon JY. Ventilator-associated pneumonia. Am J Respir Crit Care Med 2002; 165:867–903 2 Fagon JY, Chastre J, Hance AJ, et al. Evaluation of clinical judgment in the identification and treatment of nosocomial pneumonia in ventilated patients. Chest 1993; 103:547–553 3 Wunderink RG. Mortality and the diagnosis of ventilatorassociated pneumonia: a new direction. Am J Respir Crit Care Med 1998; 157:349 –350 4 Fabregas N, Ewig S, Torres A, et al. Clinical diagnosis of ventilator associated pneumonia revisited: comparative validation using immediate post-mortem lung biopsies. Thorax 1999; 54:867– 873 5 Campbell GD Jr. Blinded invasive diagnostic procedures in ventilator-associated pneumonia. Chest 2000; 117:207S–211S 6 Fagon JY, Chastre J, Wolff M, et al. Invasive and noninvasive strategies for management of suspected ventilator-associated pneumonia: a randomized trial. Ann Intern Med 2000; 132: 621– 630 7 Meduri GU, Wunderink RG, Leeper KV, et al. Management of bacterial pneumonia in ventilated patients: protected bronchoalveolar lavage as a diagnostic tool. Chest 1992; 101:500 –508 8 Papazian L, Thomas P, Garbe L, et al. Bronchoscopic or blind sampling techniques for the diagnosis of ventilator-associated pneumonia. Am J Respir Crit Care Med 1995; 152:1982–1991 9 Brunkhorst FM, Al-Nawas B, Krummenauer F, et al. Procalcitonin, C-reactive protein and APACHE II score for risk evaluation in patients with severe pneumonia. Clin Microbiol Infect 2002; 8:93–100 10 Duflo F, Debon R, Monneret G, et al. Alveolar and serum procalcitonin: diagnostic and prognostic value in ventilatorassociated pneumonia. Anesthesiology 2002; 96:74 –79 11 Meduri GU, Headley S, Kohler G, et al. Persistent elevation of inflammatory cytokines predicts a poor outcome in ARDS: plasma IL-1 and IL-6 levels are consistent and efficient predictors of outcome over time. Chest 1995; 107:1062–1073 12 Monton C, Torres A, El-Ebiary M, et al. Cytokine expression in severe pneumonia: a bronchoalveolar lavage study. Crit Care Med 1999; 27:1745–1753 13 Wu CL, Lee YL, Chang KM, et al. Bronchoalveolar interleukin-1 beta: a marker of bacterial burden in mechanically ventilated patients with community-acquired pneumonia. Original Research
Crit Care Med 2003; 31:812– 817 14 Gibot S, Cravoisy A, Levy B, et al. Soluble triggering receptor expressed on myeloid cells and the diagnosis of pneumonia. N Engl J Med 2004; 350:451– 458 15 Determann RM, Millo JL, Gibot S, et al. Serial changes in soluble triggering receptor expressed on myeloid cells in the lung during development of ventilator-associated pneumonia. Intensive Care Med 2005; 31:1495–1500 16 Mutlu GM, Garey KW, Robbins RA, et al. Collection and analysis of exhaled breath condensate in humans. Am J Respir Crit Care Med 2001; 164:731–737 17 Carpenter CT, Price PV, Christman BW. Exhaled breath condensate isoprostanes are elevated in patients with acute lung injury or ARDS. Chest 1998; 114:1653–1659 18 Fartoukh M, Maitre B, Honore S, et al. Diagnosing pneumonia during mechanical ventilation: the clinical pulmonary infection score revisited. Am J Respir Crit Care Med 2003; 168:173–179
www.chestjournal.org
19 Cantral DE, Tape TG, Reed EC, et al. Quantitative culture of bronchoalveolar lavage fluid for the diagnosis of bacterial pneumonia. Am J Med 1993; 95:601– 607 20 Bouchon A, Dietrich J, Colonna M. Cutting edge: inflammatory responses can be triggered by TREM-1, a novel receptor expressed on neutrophils and monocytes. J Immunol 2000; 164:4991– 4995 21 Bouchon A, Facchetti F, Weigand MA, et al. TREM-1 amplifies inflammation and is a crucial mediator of septic shock. Nature 2001; 410:1103–1107 22 Pugin J, Auckenthaler R, Mili N, et al. Diagnosis of ventilatorassociated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic “blind” bronchoalveolar lavage fluid. Am Rev Respir Dis 1991; 143:1121–1129 23 Schurink CA, Van Nieuwenhoven CA, Jacobs JA, et al. Clinical pulmonary infection score for ventilator-associated pneumonia: accuracy and inter-observer variability. Intensive Care Med 2004; 30:217–224
CHEST / 132 / 1 / JULY, 2007
63