Original Contribution

Usefulness of triggering receptor expressed on myeloid cells-1 in differentiating between typical and atypical community-acquired pneumonia

Chorng-Kuang How MD ^a,b,c,1, Sen-Kuang Hou MD ^b,c,d,1, Hsin-Chin Shih MD, PhD ^a,b, David Hung-Tsang Yen MD, PhD ^a,b, Chun-I Huang MD ^a,c,

Chen-Hsen Lee MD ^a,b,c, Gau-Jun Tang MD ^b,d,?

^aEmergency Department, Taipei Veterans General Hospital, Taipei 112, Taiwan

^bInstitute of Emergency and Critical Care Medicine, National Yang-Ming University School of Medicine, Taipei 112, Taiwan

^cDepartment of Emergency Medicine, National Yang-Ming University School of Medicine, Taipei 112, Taiwan

^dEmergency Department, National Yang-Ming University Hospital, I-Lan 260, Taiwan

Received 22 December 2009; revised 31 December 2009; accepted 2 January 2010

Abstract

Objectives: The purpose of this study is to investigate the clinical use of inflammatory marker triggering receptor expressed on myeloid cells (TREM)-1 at admission for differentiating between typical and atypical bacterial community-acquired pneumonia .

Methods: A prospective, noninterventional study of patients with CAP hospitalized through the emergency department was performed. Surface expression of TREM-1 was analyzed using flow cytometry on peripheral blood cells, and soluble TREM-1 (sTREM-1) concentration was determined in plasma.

Results: Eighty-eight patients with clinical suspicion of CAP were eligible. The causative pathogen was identified in 39 patients (44.3%). After excluding 4 mixed pneumonia cases, 21 typical and 14 atypical bacterial infections were enrolled. Patients with typical bacterial CAP demonstrated increased TREM-1 surface expression on monocytes and neutrophils. Median plasma sTREM-1 levels at admission were

65.2 pg/mL (range, 17.6-138.1 pg/mL) in patients with typical CAP and 25.9 pg/mL (range, 11.5- 54.8 pg/mL) in patients with atypical CAP (P b .001). Soluble TREM-1 had good Discriminative value to differentiate typical from atypical pathogens with an area under the receiver operating characteristic curve of

0.87 (95% confidence interval, 0.75-0.98). At a cutoff level of 44.2 pg/mL, sTREM-1 yielded a sensitivity of 81%, a specificity of 79%, a positive likelihood ratio of 3.79, and a negative likelihood ratio of 0.24. Conclusions: In newly admitted patients with CAP, determination of the TREM-1 levels may provide useful additional Diagnostic information on the bacterial etiology.

(C) 2011

* Corresponding author. Emergency Department, National Yang-Ming University Hospital, I-Lan, Taiwan. Tel.: +886 3 9325192; fax: +886 3

9351838.

E-mail address: [email protected] (G.-J. Tang).

¹ Chorng-Kuang How and Sen-Kuang Hou are both first authors of this manuscript because of equal contribution.

Introduction

Community-acquired pneumonia continues to be a worldwide health problem. There are approximately 4 million cases of CAP in the United States each year, resulting in about 1 million hospitalizations [1]. Pneumonia

0735-6757/$ – see front matter (C) 2011 doi:10.1016/j.ajem.2010.01.010

due to atypical organisms (Mycoplasma pneumoniae, Le- gionella species, and Chlamydia pneumoniae) accounts for 20% to 40% of cases. Differential diagnosis between typical and atypical pathogens is difficult and cannot be based solely on the patient’s history, physical examination, or chest radiograph findings [1-3].

Because of the common absence of early etiologic diagnoses, the initial antibiotic regimen should be chosen empirically to cover both typical and atypical pathogens by prescribing combined therapy with a ?-lactam and a macrolide or monotherapy with a respiratory fluoroquino- lone [1,2]. However, overuse of antibiotics and inappropriate antimicrobial selection have been associated with increased drug resistance and cost [2,4-6]. In addition, empirical fluoroquinolone therapy delayed the treatment of tuberculo- sis and was associated with a poor outcome [7].

Different immune responses may underlie various micro- organisms. Infection with extracellular bacteria, such as Streptococcus pneumoniae and Staphylococcus aureus, is particularly proinflammatory, compared with fastidious organisms such as Mycoplasma and Chlamydia [8]. Thus, biomarkers of inflammation may provide useful additional diagnostic information on the etiology of pneumonia. Unfortunately, in practice, there is marked overlap in C-reactive protein (CRP) and procalcitonin levels in patients with CAP caused by the different microorganisms, and another marker is required [3,8].

The triggering receptor expressed on myeloid cells (TREM)-1 is a recently discovered cell-surface molecule that has been identified both on human and murine polymorphonuclear neutrophils and mature monocytes [9,10]. The role of TREM-1 is an amplifier of the inflammatory response [9,10]. In the presence of microbial components, the expression of TREM-1 on the cells’ surface is up-regulated together with the release of a soluble form of this receptor [11,12]. Previous study demonstrated that TREM-1 is selectively expressed in pneumonia caused by extracellular bacteria but not by tuberculosis [13]. Our recent publication showed the inability of intracellular pathogen to up-regulate TREM-1 expression [14]. The different ability of extracellular and intracellular pathogen to induce TREM-1 expression may provide a potential marker for differential diagnosis [13,14]. The aim of this study is to investigate the clinical use of inflammatory marker TREM-1 at admission for differentiating between typical and atypical CAP and to compare with other acute-phase reactants CRP and PCT.

Materials and methods

Setting and study population

This prospective, noninterventional study was conducted in the emergency department (ED) of 2 academic teaching hospitals. The institutional review board approved the study,

and patients or their next of kin provided written informed consent before enrollment. convenience sampling was used in this study. Patients who were 18 years or older with Initial diagnosis of CAP and hospitalized through the ED between September 2007 and March 2008 were eligible for inclusion in the study. Community-acquired pneumonia was defined as an acute illness associated with a new radiographic infiltrate and clinical symptoms consisting of fever or cough or purulent sputum or positive auscultation [1,2]. Exclusion criteria were immunosuppressive treatment and/or steroid (N15 mg/d), leucopenia (b1000/uL), or neutropenia (b500/uL); patients who were admitted within the previous 2 weeks; and those with do-not-resuscitate orders.

Because the purpose of this study was to evaluate the correlation of inflammatory marker TREM-1 with microbi- ologic etiology of CAP, only patients with the single definite causative pathogens after extensive microbiologic workup were enrolled. The attending physician ordered antimicrobial therapy according to the usual practice, without interference by the research team.

Microbiologic diagnostics

At the point of inclusion into the study, extensive microbiologic workup was performed. In brief, it included obtaining sputum samples for Gram stain and culture, 2 blood samples for culture, a Urine sample for detection of Legionella pneumonia and S pneumoniae antigens, and serum samples for detection of antibodies against atypical pathogens.

infectious etiology of pneumonia was classified as definite if one of the following criteria was met: (1) blood cultures isolated a bacterial pathogen in the absence of an apparent extrapulmonary focus; (2) a qualified sputum sample (leucocytes 25 per 10x field) yielded one or more predominant organism for S pneumoniae, Haemophilus influenzae, Moraxella catarrhalis, S aureus, and other gram-negative enterobacteria; (3) C pneumoniae, a 4-fold rise in microimmunofluorescence antibody titer to >=1:128 or the presence of immunoglobulin M antibodies (>=1:20); (4) Legionella, isolation of the organism from respiratory secretions, or detection of urinary antigen, or a 4-fold or greater rise in immunofluorescence antibody titer for L pneumonia; (5) M pneumoniae, significant increases of antibodies in either complement fixation or enzyme immunoassay serology; (6) positive urinary antigen for S pneumoniae. Cases that fulfilled the etiologic diagnostic criteria described previously for more than one pathogen were considered to be mixed pneumonia [2].

Data collection

Data were collected at admission including age, sex, comorbid diseases, and routine blood test values. The confusion, urea nitrogen, respiratory rate, blood pressure,

Patient group	No. of patients
Typical Bacterial pneumonia	21
S pneumoniae	8
H influenzae	5
S aureus	4
Gram-negative bacilli	3
M catarrhalis	1
Pneumonia caused by atypical pathogens	14
M pneumoniae	7
Legionella species	5
C pneumoniae	2

65 years or older (CURB-65) score was calculated for each patient based on the following conditions: the presence at admission of confusion, urea concentration greater than 7 mmol/L, respiratory rate of 30 breaths per minute or more, systolic blood pressure less than 90 mm Hg, or diastolic blood pressure less than 60 mm Hg [15]. CURB-65 was stratified as 0 to 1, 2, and 3 to 5. Plasma PCT concentrations were determined by an immunofluorescent assay (BRAHMS PCT-sensitive KRYPTOR; BRAHMS AG, Hennigsdorf, Germany) with a functional assay sensitivity of 0.06 ng/mL.

Table 1 Microorganisms identified in patients (n = 35) with CAP

Determination of cell-surface TREM-1 expression

Within 12 hours of admission, EDTA anticoagulated blood was drawn and processed in less than 2 hours. Erythrocytes were lysed and leucocytes incubated with

Table 2 Clinical and biologic data at admission

fluorescein isothiocyanate-conjugated antihuman CD14 (BioLegend, San Diego, CA) and phycoerythrin-conjugated antihuman TREM-1 antibody (BioLegend) in accordance with the manufacturers’ recommendations. Samples were analyzed by a flow cytometer (FACScan, BD Biosciences, San Jose, CA). Granulocytes were defined according to their scatter pattern and monocytes according to their scatter pattern and CD14 positivity. Expression levels were measured as mean fluorescence intensity (MFI) compared with the isotype control. Seven healthy volunteers were included as healthy controls for flow cytometry analysis.

Measurement of plasma levels of soluble TREM-1

The plasma soluble TREM-1 (sTREM-1) was measured on admission by enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, Minn). The sTREM-1 levels were established in a sandwich assay by comparison with standard curves according to the recommendations of the manufac- turer. All measurements were performed in duplicate. The level was expressed as pg/mL.

Statistical analyses

Descriptive results were reported as mean +- SD or median (range) when appropriate. Variables were evaluated for an association with the diagnosis with the use of Pearson ?2 test (or the Fisher exact test when appropriate) for categorical data and Mann-Whitney U test for numerical data. The groups were compared with the use of the Mann-Whitney U test for numerical data and the Pearson ?2 test for

All patients (N = 35)		Patients with typical CAP (n = 21)	Patients with atypical CAP (n = 14)	P a
Age, y	50.8 +- 21.3	53.3 +- 23.5	47.0 +- 17.7	.429
Sex, n, M/F	24/11	15/6	9/5	.721
Comorbidity, n (%)
Diabetes	9 (25.7)	6 (28.6)	3 (21.4)	.636
COPD	6 (17.1)	4 (19.0)	2 (14.3)	.714
CURB-65, n (%)				.882
0-1	9 (25.7)	5 (23.8)	4 (28.6)
2	17 (48.6)	10 (47.6)	7 (50)
3-5	9 (25.7)	6 (28.6)	3 (21.4)
Leukocyte count, x109 cells/L	12.9 +- 3.8	13.7 +- 3.9	11.6 +- 3.5	.098
AST, U/L	31.9 +- 13.0	29.7 +- 14.5	35.4 +- 9.7	.138
Median CRP (range), mg/dL	7.5 (0.13-26.9)	9.45 (0.13-26.9)	6.71 (1.0-22.86)	.459
Median PCT (range), ng/mL	0.85 (0.24-3.28)	0.94 (0.39-3.28)	0.74 (0.24-1.35)	.092
Monocytic TREM-1 expression (MFI)	1.56 +- 0.66	1.82 +- 0.74	1.18 +- 0.22	b.001
Neutrophilic TREM-1 expression (MFI)	1.8 +- 0.53	2.0 +- 0.56	1.51 +- 0.33	.006
Median sTREM-1 (range), pg/mL	49.9 (11.5-138.1)	65.2 (17.6-138.1)	25.9 (11.5-54.8)	b.001
Values presented with mean +- SD. COPD indicates chronic obstructive pulmonary disease; AST, aspartate aminotransferase. a For comparison of patients with typical CAP and those with atypical CAP.

categorical data. Receiver operating characteristic (ROC) curve was applied to determine the most suitable diagnostic value for sTREM-1 levels. All statistical analyses were completed with SPSS 13.0 version software (SPSS, Chicago, Ill), and a 2-tailed P b .05 was considered significant.

Results

Patient populations

During the study period, 88 patients with clinical suspicion of CAP were eligible. The causative pathogen

Fig. 1 Cell-surface expression of TREM-1. Monocytes were defined as CD14⁺ cells (upper panel), and neutrophils were gated according to their scatter pattern (lower panel). Results were expressed as MFI.

Fig. 2 Plasma levels at admission of sTREM-1, PCT, and CRP according to diagnosis.

was identified in 39 patients (44.3%). After excluding 4 mixed pneumonia cases, 21 typical and 14 atypical bacterial infections were enrolled. The etiologic distribution of the causative organisms is summarized in Table 1. The main demographic characteristics, comorbidity, and initial sever-

ity measured by CURB-65 were not significantly different between patients with typical and atypical CAP at inclusion (Table 2).

Baseline cell-surface TREM-1 expression on peripheral blood cells

Monocytic TREM-1 expression was significantly higher in patients with typical CAP (MFI, 1.82 +- 0.74) than that in patients with atypical CAP (1.18 +- 0.22; P b .001) and that in healthy controls (1.22 +- 0.21, P = .015) (Fig. 1A). Neutrophilic TREM-1 expression was also significantly increased in patients with typical CAP (2.0 +- 0.56) compared with patients with atypical CAP (1.51 +- 0.33, P = .006) and in healthy controls (1.47 +- 0.2, P = .007) (Fig. 1B).

Plasma levels of sTREM-1, PCT, and CRP

There were no missing or indeterminate values for plasma levels of sTREM-1, PCT, and CRP. Median plasma sTREM- 1 levels at admission were 65.2 pg/mL (range, 17.6-138.1 pg/mL) in patients with typical CAP and 25.9 pg/mL (range, 11.5-54.8 pg/mL) in patients with atypical CAP (P b .001). There was no difference in PCT and CRP between typical and atypical pneumonia groups (Fig. 2). Plasma sTREM-1

Fig. 3 Receiver operating characteristic curves for various cutoff levels of plasma sTREM-1, PCT, and CRP in differentiating between typical and atypical CAP. Values of the area under the ROC curves were 0.87 (95% CI, 0.75-0.98) for sTREM-1, 0.67 (95% CI, 0.49-0.85) for PCT, and 0.58 (95% CI, 0.38-0.77) for CRP.

appeared to be helpful on the etiology of pneumonia. The capacities of plasma sTREM-1, PCT, and CRP to differen- tiate typical bacteria CAP from CAP caused by atypical pathogens were assessed with the ROC curves analysis. The values of the area under the ROC curves were 0.87 (95% confidence interval [CI], 0.75-0.98) for sTREM-1, 0.67 (95% CI, 0.49-0.85) for PCT, and 0.58 (95% CI, 0.38-0.77)

for CRP (Fig. 3). At a cutoff level of 44.2 pg/mL, sTREM-1 yielded a sensitivity of 81%, a specificity of 79%, a positive likelihood ratio of 3.79, and a negative likelihood ratio of 0.24.

Discussion

Community-acquired pneumonia is the leading cause of death from infectious disease in the United States and is associated with a substantial Economic burden to the health case system [1]. Community-acquired pneumonia can have various etiologies. The distribution of pathogens in our study was comparable with previous studies [2,16,17]. In the study, we found significantly higher TREM-1 expression in patients with typical compared with atypical bacterial CAP. Plasma CRP and PCT did not discriminate between typical and atypical etiology. Determination of the TREM-1 level may provide useful additional diagnostic information on the etiology of pneumonia and could have a crucial influence on the initial antimicrobial therapy.

The discrimination between typical and atypical bacterial pneumonia is important in clinical practice. Mycoplasma pneumoniae is devoid of cell wall. Pneumonia due to atypical organisms is usually intracellular pathogens. Most intracellular bacterial infection may need particular anti- biotics that can produce adequate intracellular concentrations for bactericidal effect, for example, fluoroquinolones, macrolides, tetracycline, or rifampin. Clinicians may select the empirical antibiotic regimen after classifying patients according to likely pathogens. The addition of macrolides or treatment with fluoroquinolones may lead to enhanced antibiotic resistance, increased adverse effects, and health care-related costs [5].

The host immune response to pathogen relies on both innate and adaptive immunity. Triggering receptor expressed on myeloid cell-1 is mostly implicated in innate immunity that plays an important role in myeloid cell-activated inflamma- tory response [10]. Triggering receptor expressed on myeloid cell-1 is an amplifier of toll-like receptor-mediated inflam- matory response to invading microbes [9,10]. The data reported here show that cell-surface TREM-1 expression is modulated and sTREM-1 is released during typical bacterial pneumonia. We are in line with previous reports that both cell- surface and sTREM-1 were strongly up-regulated during infection [12,13,18]. The origin of sTREM-1 is controversial. Some reports suggest that sTREM-1 results from alternative splicing of mRNAs producing secreted receptor isoforms,

whereas others propose that extracellular domains are cleaved off the cell surface [19]. The present study in CAP is also consistent with the studies showing the different ability of extracellular and intracellular pathogens to induce TREM-1 expression in vivo [13,14].

The area under the ROC curves of sTREM-1 reported in this study is good. However, there is still a small overlap in sTREM-1 levels in our patients with CAP caused by the different microorganism. In addition, drug-resistant bacteria are prevalent and growing [4,6]. Streptococcus pneumoniae, the most common cause of CAP, is steadily acquiring resistance to most of the currently available antibiotics [4,6]. These may limit the clinical application of TREM-1 to optimize initial empirical therapy for CAP. It is unlikely that a single biomarker of inflammation will be discriminatory to target antibiotic treatment to specific organisms [3,8,17].

There were several limitations in this study. First, the sample size is limited. Some small sample bias may be present. Still, a statistically significant difference was found, which implied a large differential power [20]. Despite the use of many diagnostic techniques to make an etiologic diagnosis, nearly half of the cases remain undiagnosed [16]. Second, no case of viral infection was included in this study. The TREM-

1 expression is unknown in such cases. Third, patients meeting the inclusion criteria were enrolled only when a study author or research associate was available. The convenience sample in this study may present the possibility of selection bias. Finally, the cytofluorimetric technology used in our study is a more expensive and labor-intensive assay not readily available in every ED. Although the enzyme-linked immunosorbent assay technique is less expensive, it still needs 3 to 4 hours to perform and is usually applied to batches of specimens. The practical difficulties, timeliness, and costs may limit clinical use of TREM-1 in acute setting.

Conclusions

Our results demonstrated that patients with typical bacterial CAP had increased TREM-1 surface expression on monocytes and neutrophils and augmented sTREM-1 release. Determination of the TREM-1 levels may provide useful additional diagnostic information on the bacterial etiology for the newly admitted patients with CAP and help clinicians to optimize initial empirical therapy for CAP.

Acknowledgments

This work was supported, in part, by grant RD 2008-013 from National Yang-Ming University Hospital, I-Lan, Taiwan.

References

1. Halm EA, Teirstein AS. Management of community-acquired pneumonia. N Engl J Med 2002;347:2039-45.
2. Masia M, Gutierrez F, Padilla S, et al. Clinical characterisation of pneumonia caused by atypical pathogens combining classic and novel predictors. Clin Microbiol Infect 2007;13:153-61.
3. Korppi M, Don M, Valent F, et al. The value of clinical features in differentiating between viral, pneumococcal and atypical bacterial pneumonia in children. Acta Paediatr 2008;97:943-7.
4. Kuti JL, Capitano B, Nicolau DP. Cost-effective approaches to the treatment of community-acquired pneumonia in the era of resistance. Pharmacoeconomics 2002;20:513-28.
5. Oosterheert JJ, Bonten MJ, Hak E, et al. How good is the evidence for the recommended empirical Antimicrobial treatment of patients hospitalized because of community-acquired pneumonia? A systemic review. J Antimicrob Chemother 2003;52:555-63.
6. Doern GV, Brown SD. Antimicrobial susceptibility among commu- nity-acquired respiratory tract pathogens in the USA: data from PROTEKT US 2000-01. J Infect 2004;48:56-65.
7. Wang JY, Hsueh PR, Jan IS, et al. Empirical treatment with a fluoroquinolone delays the treatment for tuberculosis and is associated with a poor prognosis in endemic areas. Thorax 2006;61:903-8.
8. Brown JS. Biomarkers and community-acquired pneumonia. Thorax 2009;64:556-8.
9. Bouchon A, Facchetti F, Weigand MA, et al. TREM-1 amplifies inflammation and is a crucial mediator of septic shock. Nature 2001; 410:1103-7.
10. Colonna M. TREMs in the immune system and beyond. Nat Rev Immunol 2003;6:445-53.
11. Gibot S, Kolopp-Sarda MN, Bene MC, et al. Plasma level of a triggering receptor expressed on myeloid cells-1: its diagnostic accuracy in patients with suspected sepsis. Ann Intern Med 2004; 141:9-15.
12. Gibot S, Le Renard PE, Bollaert PE, et al. Surface triggering receptor expressed on myeloid cells 1 expressed patterns in septic shock. Intensive Care Med 2005;31:594-7.
13. Richeldi L, Mariani M, Losi M, et al. Triggering receptor expressed on myeloid cells: role in the diagnosis of lung infections. Eur Respir J 2004;24:247-50.
14. How CK, Chern CH, Wu MF, et al. Expression of the triggering receptor expressed on myeloid cells-1 mRNA in a heterogeneous infected population. Int J Clin Pract 2009;63:126-33.
15. Lim WS, van der Eerden MM, Laing R, et al. Defining community acquired pneumonia severity on presentation to hospital: an interna- tional derivation and validation study. Thorax 2003;58:377-82.
16. Garau J, Calbo E. Community-acquired pneumonia. Lancet 2008;371:

455-8.

1. Kruger S, Ewig S, Papassotiriou J, et al. Inflammatory parameters predict etiologic patterns but do not allow for individual prediction of etiology in patients with CAP: results from the German competence network CAPNETZ. Respir Res 2009;10:65.
2. Tejera A, Santolaria F, Diez ML, et al. Prognosis of Community acquired pneumonia : value of triggering receptor expressed on myeloid cells-1 (TREM-1) and other mediators of the inflammatory response. Cytokine 2007;38:117-23.
3. Ford JW, McVicar DW. TREM and TREM-like receptors in inflammation and disease. Curr Opin Immunol 2009;21:38-46.
4. Sterne JA, Gavaghan D, Egger M. Publication and related bias in meta- analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol 2000;53:1119-29.