(Chest. 2004;126:1087-1092.)
© 2004
American College of Chest Physicians
Plasma d-Dimer Levels Correlate With Outcomes in Patients With Community-Acquired Pneumonia*
Jose M. Querol-Ribelles, MD;
Jose M. Tenias, MD;
Enric Grau, MD;
Jose M. Querol-Borras, MD;
Jose L. Climent, MD;
Emilio Gomez, MD and
Isidoro Martinez, MD
* From the Departments of Internal Medicine (Drs. Querol-Ribelles, Querol-Borras, Climent, Gomez, and Martinez), Preventive Medicine (Dr. Tenias), and Hematology (Dr. Grau), Hospital Lluis Alcanyis, Xativa, Spain.
Correspondence to: Jose M. Querol-Ribelles, MD, Department of Internal Medicine, Hospital Lluis Alcanyis, Crta. Xativa-Silla, Km 2, 46800-Xativa, Spain; e-mail: querol_jos{at}gva.es
 |
Abstract
|
|---|
Study objectives: The aim of this study was to investigate the prognostic value of plasma d-dimer levels in patients with community-acquired pneumonia (CAP).
Design: Prospective observational study.
Setting: Hospital Lluis Alcanyis of Xativa, Spain.
Patients: Consecutive adult patients admitted to the hospital with CAP from January 2000 to October 2002.
Measurements and results: A total of 302 patients were included. Plasma d-dimer was measured using an automated latex assay. The relationships between plasma d-dimer and prognostic variables included in the pneumonia severity index (PSI) were examined using univariate and multivariate linear and logistic regression analyses. d-Dimer levels were negative (ie, < 500 ng/mL) in 16.9% of the patients. In nonsurvivors, the d-dimer plasma level mean value was 3,786 ng/mL, while in survivors it was 1,609 ng/mL (p < 0.0001). A significant relationship was found between the presence of elevated d-dimer levels and the PSI and APACHE (acute physiology and chronic health evaluation) II score. Elevated d-dimer levels were associated with radiologic pneumonia extension. The d-dimer predictive value for mechanical ventilation therapy showed an area under the curve of 0.78 (95% confidence interval, 0.71 to 0.81).
Conclusions: d-Dimer plasma levels could be useful for predicting clinical outcome in patients with CAP.
Key Words: community-acquired pneumonia • d-dimer • outcomes • pneumonia severity index
|
Introduction
|
|---|
d-Dimer results from the fibrin breakdown after fibrinolytic system activation. Circulating d-dimer levels can be measured easily, and elevated levels have been detected in patients with disseminated intravascular coagulation, severe sepsis, thromboembolic events, pregnancy, liver disease, surgery, and trauma.123
Although the potential use of plasma d-dimer levels has been assessed as a screening test for venous thromboembolism, its role in other disorders has not been as well defined. In patients without disseminated intravascular coagulation or venous thromboembolic disease, high d-dimer levels may represent microvascular thrombosis or extracellular remodeling of fibrin. d-Dimer has been found to be a clinically significant marker for lymphovascular invasion and early tumor invasion in patients with solid tumors.45 In addition, several investigators have demonstrated that elevated d-dimer levels on the admission of critically ill patients to the ICU is associated with an increased risk of mortality.67
Little is known about the relationship between d-dimer levels and the clinical outcomes of patients with community-acquired pneumonia (CAP). Previous studies have supported the fact that both intravascular and extravascular coagulation is associated with acute and chronic lung injury. Intraalveolar activation of the coagulation cascade is thought to lead to fibrin deposition in the pulmonary interstitium and alveoli. In parenchymal lung and pleural disease, a transitional fibrin neomatrix constitutes part of the acute inflammatory response, and appears to initiate a sequence of events that leads to tissue remodeling and ultimately to fibrosis.8910 These observations prompted the present study that has examined the relationship between d-dimer levels and outcomes in patients with CAP.
|
Materials and Methods
|
|---|
Patients
From January 1, 2000, to October 31, 2002, in the Hospital Lluis Alcanyis (Xativa, Spain), blood samples were prospectively obtained from patients who were > 16 years of age and had received a diagnosis of CAP. CAP was defined as an opacity that was consistent with the presence of acute pneumonia on the chest radiograph (CXR) and was associated with respiratory symptoms, infectious syndrome, and lack of an alternative diagnosis. Blood was obtained from all patients in the emergency service from whom consent could be obtained before starting antibiotic treatment. Exclusion criteria included neutropenia due to chemotherapy, hematologic neoplasm, admission to a hospital within the previous 10 days, and high probability of a pulmonary embolism (PE).11
All patients were investigated using the pneumonia severity index (PSI) and acute physiology and chronic health evaluation (APACHE) II questionnaire.1213 Initial PSI and APACHE II scores were determined by physicians from the emergency department, and were reviewed in the next 24 h by one the authors (JMQR). Patients were treated in the hospital or at home by the implementation of CAP guidelines.14 The criteria are shown in Table 1
.
Study Protocol
d-Dimer was measured (Liatest D-DI quantitative automated latex system; Diagnostica Stago; Asnières-sur-Seine, France) on the day that the specimen was collected (STA; Diagnostica Stago). A value of > 500 ng/mL was considered to be a positive d-dimer level (ie, the cutoff level used in our hospital in the diagnosis of venous thromboembolic disease, which is equal to the upper limit of the reference range). Patients also underwent standard evaluation that included medical history, physical examination, CXR, and ECG. Blood cell counts, C-reactive protein measurement, and prothrombin time also were measured. Two blood samples for culture and a sputum sample for Gram staining and culture were collected before the antibiotic treatment was begun when the patient could expectorate normally. In the first 24 to 48 h, antigen detection tests for pneumococcus (Now Streptococcus pneumoniae Urinary Antigen Test; Binax; Portland, ME) and Legionella (Now Legionella pneumophila Urinary Antigen Test; Binax) in urine were performed, and a serum sample was drawn to complete the etiologic diagnosis, using it together with the sample collected after the first 4 weeks to make the serologic diagnosis for L pneumophila, Mycoplasma pneumoniae, Chlamydia pneumoniae, Coxiella burnetii, respiratory syncytial virus, adenovirus, influenza A, influenza B, parainfluenza 1, parainfluenza 2, and parainfluenza 3.
A CXR was obtained on hospital admission and was repeated 48 h later in all patients. Patients were assigned to the following pneumonia subgroups: segmental pneumonia (one lung segment involved); lobar pneumonia (more than one segment of the same lobe involved); and multilobar pneumonia (more than one lobe involved). Improvement was considered to be apparent when a CXR performed 48 h later showed a reduction of the consolidation. Worsening was considered to have occurred when segmental pneumonia became lobar or when lobar pneumonia became multilobar pneumonia.
All the patients admitted to the hospital were assessed daily and on the day of hospital discharge. The patients discharged from the hospital underwent a CXR and clinical assessment 30 days later.
Statistical Analysis
The analytical approach regarding the statistics was determined in order to detect significant relationships between patient outcomes and measurements of circulating d-dimer. Continuous variables were analyzed with the Student t test. Analysis of variance for normally distributed variables and the Kruskal-Wallis test for nonparametrically distributed variables were carried out. The Spearman rank correlation test was used for analysis of the linear relationship between d-dimer levels and quantitative variables.
The relationship between d-dimer and mortality was evaluated using a multivariate logistic regression analysis in which d-dimer levels and PSI scores were considered to be independent variables. The area under the curve (AUC) of the receiver operating characteristic was used for the d-dimer mortality predictive value. All p values of < 0.05 were assumed to indicate statistical significance.
|
Results
|
|---|
During the study period, 405 patients were initially analyzed. Of these, 20 patients were excluded from the study because they had a high probability of PE. In addition, 18 patients with CAP criteria were excluded from the study due to neutropenia (10 patients), hematologic cancer (6 patients), and previous hospital admission (2 patients). Finally, 65 patients were excluded from the study because d-dimer measurements had been performed 24 to 48 h after starting antibiotic treatment. Thus, the final study cohort consisted of 302 patients. There were no differences between patients included in the study and those excluded regarding age, gender, PSI, APACHE II score, and radiologic extension.
The mean age of patients was 73 years, 74.8% of the patients were men, and the in-hospital mortality rate was 12.6%. The mean (± SD) PSI was 109.6 ± 33.7, and the mean APACHE II score was 13.0 ± 6.8. Table 2
shows the median levels of d-dimer in the presence or absence of factors included in the PSI. d-Dimer status correlated significantly with stroke, mental disease, pH of < 7.35, glucose level of < 250 mg/dL, pleural effusion, and pulmonary extension.
Plasma d-dimer levels showed an asymmetric distribution (mean level, 1,034 ng/mL; range, 171 to 11,248 ng/mL). The d-dimer level was negative in 16.9% of the patients. At least 52.9% of the patients with d-dimer levels of > 500 ng/mL showed d-dimer levels between 1,000 and 2,000 ng/mL.
Although it was not very high, there was a statistically significant correlation between d-dimer levels and APACHE II score (r = 0.23; p < 0.001). d-Dimer levels increased in relationship to PSI (Table 3
), with a statistically significant correlation (r = 0.30; p < 0.001). d-Dimer levels were significantly higher in patients with lobar or multilobar pneumonia than in patients with segmental pneumonia (Table 4
). In addition, those patients who had a poor clinical course also had higher d-dimer levels than those patients who showed improvement. In fact, mechanical ventilation was needed more frequently in patients with high d-dimer levels (Table 4). The predictive value of d-dimer level for mechanical ventilation therapy showed an AUC of 0.78 (95% confidence interval [CI], 0.71 to 0.81). The median (ie, 25th to 75th interquartile range) d-dimer levels observed in patients with major complications who died were 3,629 ng/mL in those with acute respiratory failure, 5,794 ng/mL in those with respiratory distress syndrome, 2,860 ng/mL in those with severe sepsis, and 3,063 ng/mL in those with serious decompensation of their baseline disease.
A microbiological diagnosis was achieved in 96 of patients (31.4%), S pneumoniae was identified in 54 patients (18%), L pneumophila was identified in 12 patients (4%), Haemophilus influenzae was identified in 7 patients (2.3%), M pneumoniae was identified in 6 patients (2%), C pneumoniae was identified in 4 patients (1.4%), C burnetii was identified in 3 patients (1%), Staphylococcus aureus was identified in 3 patients (1.2%), M pneumoniae was identified in 2 patients (0.7%), and viruses were identified in 5 patients (1.7%). No correlation was found between d-dimer levels and pneumococcal etiology of pneumonia (Table 4).
d-Dimer levels strongly correlated with mortality. In nonsurvivors, the mean d-dimer level was 3,786 ± 2,646 ng/mL, while in survivors it was 1,609 ± 1,808 ng/mL (p < 0.0001). For a cutoff level of 500 ng/mL, the sensitivity for mortality was 97.4%, and the negative predictive value was 98.1%. In fact, only one patient died who had d-dimer levels of 422 ng/mL. d-Dimer levels showed a high mortality predictive value with AUCs of 0.80 (95% CI, 0.75 to 0.84). The logistic regression model showed an independent relationship between d-dimer levels and mortality. In addition, the mortality risk in patients in PSI categories IV and V significantly increased when d-dimer levels were found to be > 2,000 ng/mL (Fig 1
). However, an AUC of the combination of both PSI and d-dimer level of 0.92 (95% CI, 0.88 to 0.95) did not significantly increase the mortality predictive value of a PSI of 0.91 (95% CI, 0.88 to 0.95) [Fig 2
].
View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1.. The probability of death in relationship with d-dimer levels in PSI categories IV and V. Probabilities were calculated with a logistic regression model using PSI and d-dimer levels as independent variables.
|
|
View larger version (23K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2.. Receiver operating characteristic curves for d-dimer level (gray line; AUC, 0.78; 95% CI, 0.71 to 0.81), APACHE II score (broken line; AUC, 0.88; 95% CI, 0.84 to 0.92), and PSI (black line; AUC, 0.91; 95% CI, 0. 88 to 0.95).
|
|
|
Discussion
|
|---|
As far as we know, no relationship between d-dimer status and CAP has been reported. This prospective study found that d-dimer status was often positive in patients with newly diagnosed CAP. Moreover, this study has demonstrated that in-hospital mortality was significantly greater among patients with high d-dimer levels. d-Dimer levels correlated with well-established scoring systems for outcome prediction purposes such as APACHE II score and PSI. However, the relationship between d-dimer level and mortality was independent of these prognostic indicators, and even of others such as the type of antibiotic administered or the length of delay in its implementation.
Several investigators have addressed the relationship between d-dimer level and clinical outcomes in critically ill patients. Shorr et al7 found that high d-dimer levels in critically ill patients were associated with increased in-hospital mortality. In that study, 9 of the 14 patients admitted to the ICU with pneumonia had high d-dimer levels. Kollef et al6 measured d-dimer levels in 123 patients who had been admitted to a medical ICU and demonstrated that increased plasma d-dimer concentrations were associated with clinical outcomes. Other authors have examined d-dimer levels in different lung diseases, in which histologic studies have demonstrated fibrin deposition in the pulmonary interstitium and alveoli. One study found that d-dimer status was often positive in patients with sarcoidosis. More importantly, these authors found that d-dimer levels were associated with disease activity, pulmonary function test results, and levels of serum markers of inflammation.15
It is well-known that d-dimer results from the breakdown of intravascular fibrin and can serve as a marker for fibrinolytic system activity. Additionally, growing evidence suggests that fibrin degradation products may enter into the circulation by the action of the alveolar space fibrinolytic activity. Under inflammatory conditions, the alveolar hemostatic balance is shifted toward a predominance of procoagulant activity. In contrast, the fibrinolytic activity of the alveolar space was found to be markedly reduced under these conditions.89101617 Multiple inflammatory cytokines may be involved in endothelial injury and dysregulation coagulation and fibrinolysis. Several authors1819 have demonstrated a relationship between markers of coagulation system activation, including d-dimer and cytokines such as interleukin-6. On the other hand, in the BAL fluids from patients with ARDS, the urokinase concentrations, representing the predominant plasminogen activator in this compartment, were markedly decreased, whereas elevated activities of plasminogen activator inhibitor-1 and 
2-antiplasmin were consistently encountered.20 It has been demonstrated21 that urokinase may induce selective fibrinolysis into the alveolar compartment, confirming the pathologic relevance of alveolar fibrin degradation.
On the other hand, it has also been demonstrated22 that inflammatory cytokines are capable of activating coagulation and inhibiting fibrinolysis in patients with severe sepsis. In the Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis study,3 which was performed in patients who had been hospitalized due to severe sepsis with systemic inflammation and organ failure due to acute infection, in 53.6% of whom the origin was pulmonary infection, the baseline plasma d-dimer and serum interleukin-6 levels were elevated, as was the expression of the inflammatory and procoagulant host response to infection. In our study, which was performed in patients with CAP without severe sepsis criteria, we observed that d-dimer levels correlated with pleural effusion, pulmonary extension, and radiologic evolution rather than with systemic fibrinolytic activation markers, such as leukocytosis, serum C-reactive protein, or bacteremia. Other authors23 have observed similar results in critically ill patients due to severe pneumonia or ventilator-associated pneumonia. Moreover, they have found that d-dimer levels were higher in those patients with alveolar or interstitial pneumonia than in patients with bronchopneumonia.23
Our study has several limitations. First, patients with positive d-dimer status were not systematically assessed to detect the presence of venous thromboembolism. In fact, ventilation-perfusion lung scanning, helical CT scanning, or compression ultrasonography was performed to rule out pulmonary embolism (PE) or deep vein thrombosis only in those patients with a high pretest probability. Further studies using serial systematic evaluations for deep vein thrombosis and PE will make it possible to determine how many patients with venous thromboembolism are not identified when a CAP diagnosis is made. Other authors have compared d-dimer levels between patients with pneumonia and patients with a high probability of PE. They found that patients with pneumonia or PE showed higher d-dimer levels than a control group. They also concluded that d-dimer measurement was useless in the differential diagnosis between pneumonia and PE.24 Second, a systematic study of the prothrombotic changes in the hemostatic system of patients in whom CAP had been diagnosed was not performed. Prothrombotic changes in the hemostatic system associated with acute inflammation may provide the pathophysiologic link between respiratory infections and vascular disease. The measurement of fibrinogen, factor VII, prothrombin fragment 1 and 2, thrombin-antithrombin complexes, plasmin-antiplasmin complexes, tissue-type plasminogen activator, and plasminogen activator inhibitor-1 may help to establish the role of intravascular fibrinolytic activation in the increased d-dimer levels found in patients with CAP.
In summary, we observed that d-dimer status is frequently positive in patients with CAP. We found that d-dimer levels at the time of diagnosis of CAP correlated with clinical outcome. Our study suggests that d-dimer levels of > 2,000 ng/mL in patients in PSI category IV identify a subgroup of patients with CAP who are at increased risk for in-hospital death. Our findings add to the growing evidence underscoring the significance of intraalveolar fibrin remodeling in patients with CAP. The question of whether high d-dimer levels result from a systemic fibrinolytic activation or from intraalveolar fibrin degradation merits further research.
|
Footnotes
|
|---|
Abbreviations: APACHE = acute physiology and chronic health evaluation; AUC = area under the curve; CAP = community-acquired pneumonia; CI = confidence interval; CXR = chest radiograph; PE = pulmonary embolism; PSI = pneumonia severity index
Received for publication September 2, 2003.
Accepted for publication March 31, 2004.
|
References
|
|---|
- Ginsberg, JS, Wells, PS, Kearon, C, et al (1998) Sensitivity and specificity of a rapid whole-blood assay for d-dimer in the diagnosis of pulmonary embolism. Ann Intern Med 129,1006-1011[Abstract/Free Full Text]
- Wada, H, Sakuragawa, N, Mori, Y, et al Hemostatic molecular markers before the onset of disseminated intravascular coagulation. Am J Hematol 1999;60,273-278[CrossRef][ISI][Medline]
- Bernard, GR, Vincent, JL, Laterre, PF, et al Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001;344,699-709[Abstract/Free Full Text]
- Taguchi, O, Gabazza, EC, Yasui, H, et al Prognostic significance of plasma d-dimer levels in patients with lung cancer. Thorax 1997;52,563-565[Abstract]
- Blackwell, K, Haroon, Z, Broadwater, G, et al Plasma d-dimer levels in operable breast cancer patients correlate with clinical stage and axillary lymph node status. J Clin Oncol 2000;18,600-608[Abstract/Free Full Text]
- Kollef, MH, Eisenberg, PR, Shannon, W A rapid assay for the detection of circulating d-dimer is associated with clinical outcomes among critically ill patients. Crit Care Med 1998;26,1054-1060[CrossRef][ISI][Medline]
- Shorr, AF, Trotta, RF, Alkins, SA, et al D-dimer assay predicts mortality in critically ill patients without disseminated intravascular or venous thromboembolic disease. Intensive Care Med 1999;25,207-210[CrossRef][ISI][Medline]
- Idell, S Coagulation, fibrinolysis and fibrin deposition in acute lung injury. Crit Care Med 2003;31(suppl),S213-S220
- Abraham, E Coagulation abnormalities in acute lung injury and sepsis. Am J Respir Cell Mol Biol 2000;22,401-404[Free Full Text]
- Günther, A, Mosavi, P, Heinemann, S, et al Alveolar fibrin formation caused by enhanced procoagulant and depressed fibrinolytic capacities in severe pneumonia. Am J Respir Crit Care Med 2000;161,454-462[Abstract/Free Full Text]
- Wells, PS, Ginsberg, JS, Anderson, DR, et al Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998;129,997-1005[Abstract/Free Full Text]
- Fine, MJ, Auble, TE, Borowsky, LH, et al A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med 1997;336,243-250[Abstract/Free Full Text]
- Knaus, WA, Draper, EA, Wagner, DP, et al APACHE II: a severity of disease classification system. Crit Care Med 1985;10,818-829
- Barlett, JG, Dowell, SF, Mandell, LA, et al Practice guidelines for the management of community-acquired pneumonia in adults. Clin Infect Dis 2000;31,347-382[CrossRef][Medline]
- Shorr, AF, Hnatiuk, OW Circulating D dimer in patients with sarcoidosis. Chest 2000;117,1012-1016[Abstract/Free Full Text]
- Fuchs-Buder, T, De Moerloose, P, Ricou, B, et al Time course of procoagulant activity and D dimer in bronchoalveolar fluid of patients at risk for or with acute respiratory distress syndrome. Am J Respir Crit Care Med 1996;153,163-167[Abstract]
- Abraham, E Coagulation abnormalities in acute lung injury and sepsis. Am J Respir Cell Mol Biol 2000;22,401-404
- Shorr, AF, Thomas, SJ, Alkins, SA, et al D-dimer correlates with proinflammatory cytokine levels and outcomes in critically ill patients. Chest 2002;121,1262-1268[Abstract/Free Full Text]
- Pettila, V, Hynninen, M, Takkunen, O, et al Predictive value of procalcitonin and interleukin 6 in critically ill patients with suspected sepsis. Intensive Care Med 2002;28,1220-1225[CrossRef][ISI][Medline]
- Hasday, JD, Bachwich, PR, Lynch, JP, et al Procoagulant and plasminogen activator activities of bronchoalveolar fluid in patients with pulmonary sarcoidosis. Exp Lung Res 1988;14,261-278[ISI][Medline]
- Ruppert, C, Markart, P, Schmidt, R, et al Chemical crosslinking of urokinase to pulmonary surfactant protein B for targeting alveolarfibrin. Thromb Haemost 2002;89,53-64
- Bone, RC, Grodzin, CG, Balk, RA Sepsis: a new hypothesis for pathogenesis on the disease process. Chest 1997;112,235-243[Free Full Text]
- Levi, M, Schultz, MJ, Rijneveld, AW, et al Bronchoalveolar coagulation and fibrinolysis in endotoxemia and pneumonia. Crit Care Med 2003;31(suppl),S238-S242
- Jimenez-Castro, D, Perez-Rodriguez, E, Montaner, L, et al Diagnostic value of D-dimer in pulmonary embolism and pneumonia. Respiration 2002;IV,229-233
TOP
Abstract
Introduction
