Soluble TREM-1 and the Diagnosis of Pneumonia
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Gibot et al. (Jan. 29 issue)1 state that in mechanically ventilated patients, detection of soluble triggering receptor expressed on myeloid cells (sTREM-1) may be useful for diagnosing or ruling out bacterial pneumonia. The 148 patients enrolled in the study had radiographic evidence of a new and persistent infiltrate, with purulent tracheal secretions, fever, leukocytosis or leukopenia, or some combination of these findings. The 84 patients with microbiologic evidence of infection had a higher level of sTREM-1 in bronchoalveolar-lavage fluid than did the 64 patients without confirmed infection.
The fact that these latter patients were considered not to have pneumonia even though they met the usual criteria is disturbing. Causative pathogens are identified in only 50 percent of patients with community-acquired pneumonia.2 Moreover, the sensitivity of bronchoalveolar-lavage findings for the diagnosis of ventilator-associated pneumonia is around 70 percent.3 Consequently, we wonder whether all the patients with negative bronchoalveolar-lavage specimens could be considered not to have pneumonia. And it is surprising that 11 patients with chronic obstructive pulmonary disease were considered to have an acute exacerbation without pneumonia, despite having new pulmonary infiltrates. Consequently, we wonder whether the no-pneumonia group used to validate the usefulness of sTREM-1 detection can really be considered a control group.
Georges Hugues, M.D.
Leroy Olivier, M.D.
Guery Benoit, M.D., Ph.D.
H?pital Chatilliez
59208 Tourcoing, France
References
Gibot S, Cravoisy A, Levy B, Bene MC, Faure G, Bollaert P-E. Soluble triggering receptor expressed on myeloid cells and the diagnosis of pneumonia. N Engl J Med 2004;350:451-458.
Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia: diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001;163:1730-1754.
Torres A, El-Ebiary M. Bronchoscopic BAL in the diagnosis of ventilator-associated pneumonia. Chest 2000;117:Suppl 2:198S-202S.
To the Editor: With regard to the report by Gibot et al. concerning the levels of sTREM-1, sensitivity and specificity should be expressed with confidence intervals, to show the range within which true values are likely to lie.1 The authors used nonparametric tests; hence, they should have expressed their results as medians (with interquartile ranges) instead of means (with standard deviations).
With pneumonia defined by the association of histologic signs and positive lung-tissue cultures, the accuracy of mini–bronchoalveolar lavage in the diagnosis is somewhat poor (sensitivity, 78 percent; specificity, 86 percent).2 Bronchoscopy with direct examination of bronchoalveolar-lavage fluid with the use of a threshold of 1 percent of infected cells is easier to perform and less expensive than sTREM-1 detection with similar accuracy (sensitivity, 93.6 percent; specificity, 91.5 percent).3 The usefulness of sTREM-1 detection in the diagnosis of pneumonia is therefore unclear. Finally, measurements of sTREM-1 in mini–bronchoalveolar-lavage fluid should have been corrected by the dilution factor, since the observed concentrations of sTREM-1 are closely correlated with the volumes of the aspirates.4
Bernard Allaouchiche, M.D., Ph.D.
Edouard Herriot Hospital
69433 Lyons, France
Emmanuel Boselli, M.D.
H?tel Dieu Hospital
69002 Lyons, France
References
Harper R, Reeves B. Reporting of precision of estimates for diagnostic accuracy: a review. BMJ 1999;318:1322-1323.
Bregeon F, Papazian L, Thomas P, et al. Diagnostic accuracy of protected catheter sampling in ventilator-associated bacterial pneumonia. Eur Respir J 2000;16:969-975.
Timsit JF, Cheval C, Gachot B, et al. Usefulness of a strategy based on bronchoscopy with direct examination of bronchoalveolar lavage fluid in the initial antibiotic therapy of suspected ventilator-associated pneumonia. Intensive Care Med 2001;27:640-647.
Yamazaki K, Ogura S, Ishizaka A, Oh-hara T, Nishimura M. Bronchoscopic microsampling method for measuring drug concentration in epithelial lining fluid. Am J Respir Crit Care Med 2003;168:1304-1307.
The authors reply: We thank Hugues et al. and Allaouchiche and Boselli for their comments and want to take the opportunity to clarify our statement about the diagnostic performance of sTREM-1 assessment in patients with suspected pneumonia. We agree that the lack of a gold-standard definition for ventilator-associated pneumonia may have hampered the classification of some of our patients. Nevertheless, we used the most widely used definition1 along with a mini–bronchoalveolar-lavage analysis, which has been reported to compare favorably with traditional bronchoscopic bronchoalveolar-lavage examination.2
The assertion that bronchoscopic bronchoalveolar-lavage direct examination is similar in accuracy to sTREM-1 determination, as well as easier and less expensive, is not correct. First, bronchoscopic bronchoalveolar lavage is an invasive technique, which may cause a transient worsening of the patient's condition, and it must be performed by an experienced physician. Second, no cutoff level for bronchoalveolar-lavage cells containing intracellular bacteria has been clearly validated — should it be 1 percent? 5 percent? Third, Allaouchiche and Boselli showed that the agreement between Gram's staining of bronchoalveolar-lavage fluid and protected-specimen-brushing quantitative cultures was complete in only 51 percent of cases and was partial in 39 percent and absent in 10 percent.3
The amount of fluid recovered in mini–bronchoalveolar-lavage specimens was fairly consistent (mean [±SD] volume, 13±3 ml), and correction for the dilution therefore did not appear to be indicated. Moreover, the discriminative power of the test might even allow consideration of merely the presence or absence (rather than the concentration) of sTREM-1.
The accuracy of the sTREM-1 assay, with a sensitivity of 98 percent (95 percent confidence interval, 92 to 99) and a specificity of 90 percent (95 percent confidence interval, 81 to 96), did not differ between patients with community-acquired pneumonia, the diagnosis of which is generally straightforward, and patients with ventilator-associated pneumonia. Moreover, in the group of patients classified as having no pneumonia, either an alternative diagnosis was made (in 54 of 64 patients), or the patients recovered spontaneously, without receiving antimicrobial therapy. Within this no-pneumonia group, 11 patients with chronic obstructive pulmonary disease had persistent ventilatory abnormalities, which are common in such patients.
Finally, patients were classified by physicians who were unaware of the results of the sTREM-1 assay. We do not think that the lack of a gold-standard definition of ventilator-associated pneumonia (diagnosed in 31 percent of our patients) may have biased our conclusions.
Sébastien Gibot, M.D.
Bruno Levy, M.D., Ph.D.
Réanimation Médicale
54000 Nancy, France
s.gibot@chu-nancy.fr
Marie-Christine Béné, Pharm.Sci.D., Ph.D.
Laboratoire d'Immunologie
54000 Nancy, France
References
Pugin J, Auckenthaler R, Mili N, Janssens JP, Lew PD, Suter PM. Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic "blind" bronchoalveolar lavage fluid. Am Rev Respir Dis 1991;143:1121-1129.
Torres A, El-Ebiary M. Bronchoscopic BAL in the diagnosis of ventilator-associated pneumonia. Chest 2000;117:Suppl 2:198S-202S.
Allaouchiche B, Jaumain H, Chassard D, Bouletreau P. Gram stain of bronchoalveolar lavage fluid in the early diagnosis of ventilator-associated pneumonia. Br J Anaesth 1999;83:845-849.
The fact that these latter patients were considered not to have pneumonia even though they met the usual criteria is disturbing. Causative pathogens are identified in only 50 percent of patients with community-acquired pneumonia.2 Moreover, the sensitivity of bronchoalveolar-lavage findings for the diagnosis of ventilator-associated pneumonia is around 70 percent.3 Consequently, we wonder whether all the patients with negative bronchoalveolar-lavage specimens could be considered not to have pneumonia. And it is surprising that 11 patients with chronic obstructive pulmonary disease were considered to have an acute exacerbation without pneumonia, despite having new pulmonary infiltrates. Consequently, we wonder whether the no-pneumonia group used to validate the usefulness of sTREM-1 detection can really be considered a control group.
Georges Hugues, M.D.
Leroy Olivier, M.D.
Guery Benoit, M.D., Ph.D.
H?pital Chatilliez
59208 Tourcoing, France
References
Gibot S, Cravoisy A, Levy B, Bene MC, Faure G, Bollaert P-E. Soluble triggering receptor expressed on myeloid cells and the diagnosis of pneumonia. N Engl J Med 2004;350:451-458.
Niederman MS, Mandell LA, Anzueto A, et al. Guidelines for the management of adults with community-acquired pneumonia: diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001;163:1730-1754.
Torres A, El-Ebiary M. Bronchoscopic BAL in the diagnosis of ventilator-associated pneumonia. Chest 2000;117:Suppl 2:198S-202S.
To the Editor: With regard to the report by Gibot et al. concerning the levels of sTREM-1, sensitivity and specificity should be expressed with confidence intervals, to show the range within which true values are likely to lie.1 The authors used nonparametric tests; hence, they should have expressed their results as medians (with interquartile ranges) instead of means (with standard deviations).
With pneumonia defined by the association of histologic signs and positive lung-tissue cultures, the accuracy of mini–bronchoalveolar lavage in the diagnosis is somewhat poor (sensitivity, 78 percent; specificity, 86 percent).2 Bronchoscopy with direct examination of bronchoalveolar-lavage fluid with the use of a threshold of 1 percent of infected cells is easier to perform and less expensive than sTREM-1 detection with similar accuracy (sensitivity, 93.6 percent; specificity, 91.5 percent).3 The usefulness of sTREM-1 detection in the diagnosis of pneumonia is therefore unclear. Finally, measurements of sTREM-1 in mini–bronchoalveolar-lavage fluid should have been corrected by the dilution factor, since the observed concentrations of sTREM-1 are closely correlated with the volumes of the aspirates.4
Bernard Allaouchiche, M.D., Ph.D.
Edouard Herriot Hospital
69433 Lyons, France
Emmanuel Boselli, M.D.
H?tel Dieu Hospital
69002 Lyons, France
References
Harper R, Reeves B. Reporting of precision of estimates for diagnostic accuracy: a review. BMJ 1999;318:1322-1323.
Bregeon F, Papazian L, Thomas P, et al. Diagnostic accuracy of protected catheter sampling in ventilator-associated bacterial pneumonia. Eur Respir J 2000;16:969-975.
Timsit JF, Cheval C, Gachot B, et al. Usefulness of a strategy based on bronchoscopy with direct examination of bronchoalveolar lavage fluid in the initial antibiotic therapy of suspected ventilator-associated pneumonia. Intensive Care Med 2001;27:640-647.
Yamazaki K, Ogura S, Ishizaka A, Oh-hara T, Nishimura M. Bronchoscopic microsampling method for measuring drug concentration in epithelial lining fluid. Am J Respir Crit Care Med 2003;168:1304-1307.
The authors reply: We thank Hugues et al. and Allaouchiche and Boselli for their comments and want to take the opportunity to clarify our statement about the diagnostic performance of sTREM-1 assessment in patients with suspected pneumonia. We agree that the lack of a gold-standard definition for ventilator-associated pneumonia may have hampered the classification of some of our patients. Nevertheless, we used the most widely used definition1 along with a mini–bronchoalveolar-lavage analysis, which has been reported to compare favorably with traditional bronchoscopic bronchoalveolar-lavage examination.2
The assertion that bronchoscopic bronchoalveolar-lavage direct examination is similar in accuracy to sTREM-1 determination, as well as easier and less expensive, is not correct. First, bronchoscopic bronchoalveolar lavage is an invasive technique, which may cause a transient worsening of the patient's condition, and it must be performed by an experienced physician. Second, no cutoff level for bronchoalveolar-lavage cells containing intracellular bacteria has been clearly validated — should it be 1 percent? 5 percent? Third, Allaouchiche and Boselli showed that the agreement between Gram's staining of bronchoalveolar-lavage fluid and protected-specimen-brushing quantitative cultures was complete in only 51 percent of cases and was partial in 39 percent and absent in 10 percent.3
The amount of fluid recovered in mini–bronchoalveolar-lavage specimens was fairly consistent (mean [±SD] volume, 13±3 ml), and correction for the dilution therefore did not appear to be indicated. Moreover, the discriminative power of the test might even allow consideration of merely the presence or absence (rather than the concentration) of sTREM-1.
The accuracy of the sTREM-1 assay, with a sensitivity of 98 percent (95 percent confidence interval, 92 to 99) and a specificity of 90 percent (95 percent confidence interval, 81 to 96), did not differ between patients with community-acquired pneumonia, the diagnosis of which is generally straightforward, and patients with ventilator-associated pneumonia. Moreover, in the group of patients classified as having no pneumonia, either an alternative diagnosis was made (in 54 of 64 patients), or the patients recovered spontaneously, without receiving antimicrobial therapy. Within this no-pneumonia group, 11 patients with chronic obstructive pulmonary disease had persistent ventilatory abnormalities, which are common in such patients.
Finally, patients were classified by physicians who were unaware of the results of the sTREM-1 assay. We do not think that the lack of a gold-standard definition of ventilator-associated pneumonia (diagnosed in 31 percent of our patients) may have biased our conclusions.
Sébastien Gibot, M.D.
Bruno Levy, M.D., Ph.D.
Réanimation Médicale
54000 Nancy, France
s.gibot@chu-nancy.fr
Marie-Christine Béné, Pharm.Sci.D., Ph.D.
Laboratoire d'Immunologie
54000 Nancy, France
References
Pugin J, Auckenthaler R, Mili N, Janssens JP, Lew PD, Suter PM. Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic "blind" bronchoalveolar lavage fluid. Am Rev Respir Dis 1991;143:1121-1129.
Torres A, El-Ebiary M. Bronchoscopic BAL in the diagnosis of ventilator-associated pneumonia. Chest 2000;117:Suppl 2:198S-202S.
Allaouchiche B, Jaumain H, Chassard D, Bouletreau P. Gram stain of bronchoalveolar lavage fluid in the early diagnosis of ventilator-associated pneumonia. Br J Anaesth 1999;83:845-849.