| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Pulmonary and Critical Care Medicine Service, Walter Reed Army Medical Center, Washington, DC.
Correspondence to: Andrew Shorr, MD, MPH, Pulmonary and Critical Care Medicine, Walter Reed Army Medical Center, 6900 Georgia Ave NW, Washington, DC;
 |
Abstract |
|---|
 TOP Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Design: Prospective observational study.
Setting: Medical ICU (MICU) of a tertiary care, academic medical center.
Patients: Individuals admitted to the MICU.
Interventions: Within 24 h of MICU admission, patients had DD status determined and interleukin (IL) levels (IL-6, IL-8, and IL-10) and tumor necrosis factor (TNF)-
measured. The strength of the DD level was also noted. Subjects were then monitored prospectively to determine mortality rate and the incidence of organ failure.
Measurement and results: The study cohort included 79 patients (mean age, 65.2 years; 54.5% male patients). DD was present in 53.2% of subjects. The DD reaction was weak (1+) in 15 patients and strong (2+) in 27 patients. The TNF-, IL-6, and IL-8 levels all increased in parallel with the increasing strength of the DD level. IL-10 levels did not differ based on DD status. Similarly, the severity of illness as measured by the APACHE (acute physiology and chronic health evaluation) II score was highest among those with higher DD levels: 24.7 ± 6.2 for those with 2+ DD vs 17.2 ± 3.1 and 11.5 ± 2.7 for those with 1+ DD and no circulating DD, respectively (p < 0.001). For patients lacking DD, the mortality rate was 8.1%, compared to 13.3% and 55.6% for those with 1+ and 2+ DD levels, respectively (p < 0.001). No patient without DD had multisystem organ failure (MSOF) develop, while the incidence of MSOF also increased with increasing DD levels. As a screening test for mortality, the DD performed as well as the APACHE II system.
Conclusions: The coagulation system is active in critically ill patients, and DD levels correlate with activation of the proinflammatory cytokine cascade. The absence of a relationship between DD and anti-inflammatory cytokines (IL-10) suggests that the presence of DD may reflect the imbalance between proinflammatory and anti-inflammatory cytokines. DD identifies patients at increased risk for both MSOF and death.
Key Words: ARDS • critical illness • cytokine • d-dimer • death • outcomes • sepsis
D-dimer (DD) results from the destruction of cross-linked fibrin and therefore is a measure of clot formation and lysis. In turn, the presence of DD indicates activation of the coagulation system. DD is often measured in the evaluation of patients with suspected disseminated intravascular coagulation, and is emerging as a screening test for venous thromboembolism.1 2 DD has also been shown to correlate with outcomes in a number of diseases, including closed head injury and systemic lupus erythematosus.3 4 More specifically, endothelial injury and dysregulation of coagulation homeostasis have been noted in several diseases that may necessitate treatment in an ICU, including variceal hemorrhage, sepsis, and ARDS.5 6 7 Investigators hypothesize that such diseases result in a microvascular coagulopathy that may lead to the formation of microthrombi.8 These microthrombi occlude the vascular beds of various organs and may contribute to their eventual failure.
In an earlier study,9 we demonstrated that DD correlated with mortality in a diverse group of critically ill medical patients. In a similar, larger trial, Kollef and colleagues10 noted that DD levels were associated not only with mortality but also with the development of severe sepsis or septic shock, multisystem organ failure (MSOF), and vascular thromboses. As with DD, protein C, an important mediator of the coagulation system, has been shown to correlate with clinical outcomes generally. In patients with neutropenia, protein C levels predict the subsequent development of sepsis.11 Finally, confirming the importance of the coagulation system in critical illness, a recently completed, large, prospective study12 of recombinant protein C revealed that this therapy may improve survival in patients suffering from severe sepsis or septic shock.
On a cellular level, multiple inflammatory cytokines may directly activate the coagulation system and simultaneously inhibit fibrinolysis. For example, both tumor necrosis factor (TNF)- and interleukin (IL)-6 are capable of disrupting coagulation homeostasis.13
14
Multiple studies15
16
have also shown that severity of illness, organ failure, and mortality correlate with serum levels of these cytokines. At present, however, it is not possible to measure easily these inflammatory cytokines. Systems for cytokine analysis are cumbersome and labor intensive. DD, however, can be measured by a variety of rapid, inexpensive, point-of-care assays. Given the link between coagulation and inflammation in critical illness, we hypothesized that DD status would correlate with levels of proinflammatory cytokines. Therefore, we performed a prospective, observational study to determine the relationship between DD and various cytokines, and to confirm the association between DD status and both severity of illness and hospital mortality in patients requiring care in an ICU.
|
Materials and Methods |
|---|
|
TOPAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
End Points
Eligible patients were identified, and after consent was obtained, blood samples were obtained for analysis (see below). Patients were then prospectively monitored after study entry until either hospital discharge or death. We focused on both biochemical and clinical end points. The primary biochemical end points were cytokine levels based on the blood samples obtained at enrollment. The primary clinical end points were severity of illness on hospital admission and hospital mortality. Secondary clinical end points were the incidence of either severe sepsis or septic shock, the diagnosis of ARDS, and the development of MSOF. All definitions for these end points were determined prospectively prior to the initiation of the study.
Severity of illness was measured via the APACHE (acute physiology and chronic health evaluation) II score.17
Severe sepsis and septic shock were defined in accordance with the conclusions of the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference.18
Consistent with the report of the American-European Consensus Conference Committee, ARDS was considered to be present if the subject had bilateral infiltrates acutely develop on chest radiograph in the absence of either a pulmonary capillary wedge pressure of > 18 mm Hg or clinical evidence of left atrial hypertension and had a PaO2/fraction of inspired oxygen ratio of 
200. MSOF required severe dysfunction of at least two organs.19
For definitions of organ dysfunction, we employed those proposed by Hebert et al20
and Rubin et al21
with modifications. Specifically, organ failures were defined as follows: (1) renal, serum creatinine level > 2 mg/dL or a twofold rise in the creatinine level from baseline value; (2) hepatic, bilirubin level > 2.0 mg/dL or an alkaline phosphatase level > 350 U/L; (3) pulmonary, requirement for mechanical ventilation or a PaO2 < 60 mm Hg while breathing an fraction of inspired oxygen > 0.5 or the use of > 10 cm H2O of positive end-expiratory pressure; (4) cardiovascular, systolic BP 90 mm Hg or need for treatment with any vasopressor; (5) hematologic failure, WBC count < 2,000/µL, or a platelet count < 75,000/µL, and/or an international normalized ratio of > 2.0; (6) neurologic, Glasgow Coma Scale score < 10 or a decrease in the Glasgow Coma Scale score by 3 if a primary CNS injury is present; and (7) GI, fresh blood from a nasogastric or orogastric tube and/or melena or fresh blood from the rectum accompanied by a fall in hemoglobin > 20 g/L requiring at least 2 U of packed RBCs in 24 h.
Assays
At the time of study enrollment, DD was measured on whole blood with a rapid, semiquantitative system (SimpliRed; AGEN Biomedical Limited; Brisbane, Australia). In the presence of DD, this assay leads to RBC agglutination. For each sample, at least 10 µL of blood was placed in a test well and mixed with reagent for 2 min. In accordance with the instructions of the manufacturer, if there was no agglutination, we confirmed the absence of DD by performing a negative control. This was completed with a different reagent in another reaction well adjacent to the test well. If agglutination indicating the presence of DD was noted, its intensity was graded as either weak (1+) or strong (2+) as outlined by the manufacturer. The stronger intensity of the reaction corresponds to higher levels of DD. The lower limit of detection of DD for the SimpliRed assay is 200 ng/mL. This level is in accordance with the upper limit of normal for DD. All DD assays were interpreted by one of two investigators (A.F.S. or S.J.T.). In other studies, the interobserver variability for this system has been minimal.10
Plasma samples for cytokine analysis were drawn contemporaneously with DD measurement. After immediate centrifugation, these were stored at - 70°C until cytokine levels were determined. Thus, the number of freeze-thaw cycles was minimized. The cytokine analysis was performed so that the technician performing the assays was blinded to the patients’ DD status and to their clinical outcomes. Concentrations of TNF-, IL-6, and IL-8 were measured by electroilluminescence. The sensitivity of this assay for each of these cytokines is 4.0 pg/mL, 5.0 pg/mL, and 5.0 pg/mL, respectively. IL-10 was measured by a electroilluminescence system. The sensitivity of this approach is 5.0 pg/mL.
Statistics
Continuous data are reported as mean ± SD. All comparisons were unpaired and two tailed. Normally distributed continuous variables were analyzed with the Student’s t test or analysis of variance. For continuous data that demonstrated a nonparametric distribution, we relied on the Wilcoxon rank sum test or the Kruskal-Wallis test. A 
2 test was performed to compare categorical variables unless the expected size of any one cell was small. In those instances we used the Fisher’s Exact Test. To determine the screening characteristics of the various biochemical markers (eg, cytokines, DD) at predicting study end points (eg, mortality), we constructed standard two-by-two tables. Similarly, we created receiver operating curves to facilitate comparisons of the various biochemical tests. We analyzed the area under curves (AUCs) to identify if any particular biochemical marker was superior at predicting the clinical end points of interest. Statistical analysis was completed using software (SPSS 9.0; SPSS; Chicago, IL). All p values 0.05 were assumed to indicate statistical significance.
|
Results |
|---|
|
TOPAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
level measured 45.71 ± 36.16 pg/mL, as compared to 82.20 ± 15.10 pg/mL in patients with weak agglutination reactions and 104.96 ± 50.67 pg/mL in those with strong reactions (p < 0.001). Similarly, IL-6 levels increased significantly with the increasing presence of DD. IL-6 was barely detectable in those without DD. In patients with mild DD reactions, the mean IL-6 level was 27.00 ± 53.99 pg/mL, while it was more elevated at 99.91 ± 30.31 pg/mL in patients with strong DD reactions (p = 0.019). A similar pattern was also observed with respect to IL-8 (Table 2)
. There was no correlation, however, between the anti-inflammatory cytokine we studied, IL-10, and DD status. IL-10 levels were equivalent across the spectrum of DD results. Since cytokine levels have been shown to be associated with survival in critically ill patients, we performed a logistic regression to control for baseline cytokine concentrations. After adjusting for hospital admission concentrations of TNF-, IL-6, IL-8, and IL-10, the odds ratio for in-hospital mortality for patients with the highest DD levels was 10.44 (95% confidence interval [CI], 1.40 to 78.15).
|
We noted like relationships between DD and both MSOF and mortality (Fig 1 ). No patient in whom DD was absent had MSOF develop. For those with mild DD elevations, one subject (6.7%) had MSOF; for those with higher DD levels, 29.6% had MSOF (p < 0.001). The median number of failing organs also correlated with DD: no organs failed in the DD-negative patients, one organ failed in those with weak DD, and three organs failed in those with high DD levels (p < 0.001) The in-hospital mortality rate was also significantly higher for patients with greater elevations in circulating DD: 55.6% of patients with high DD levels died as opposed to 13.3% of those with moderate DD elevations and 8.1% of subjects lacking DD (p < 0.001). The relative risk for mortality in patients with any detectable DD was 7.71 (95% CI, 2.03 to 29.19).
|
or IL-6 level at predicting MSOF. The AUC for the DD as a screening tool for MSOF was 0.87, while it was 0.88 for TNF- and 0.83 for IL-6 (p = not significant).
|
|
Discussion |
|---|
|
TOPAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Two earlier studies have addressed the relationship between DD and clinical outcomes in the ICU. In a smaller study9 using a different assay system, we noted that 43% of patients had circulating DD. We also observed a significant correlation between mortality and DD status. Unlike the SimpliRed system, for the earlier study we employed a latex agglutination assay. Although inexpensive and able to provide results rapidly, latex agglutination tests are prone to more interobserver variability. This present study builds on the initial effort, in that we focused on multiple clinically important end points rather than simply mortality. In a larger trial of 321 subjects, Kollef et al10 found that DD measured with the SimpliRed system not only correlated with mortality but also with the development of MSOF, sepsis, and vascular thromboses. In general, our findings confirm these earlier studies. Unlike Kollef et al,10 who relied on a complex 4-point scale to interpret the DD assay, we interpreted the DD as being either negative, weakly positive, or strongly positive. This manufacturer recommends this approach. Thus, our results add to the efforts of Kollef et al10 by showing that the relationship between clinical outcomes and DD status still holds when the DD is interpreted with a simpler, less complex scale. Additionally, one strength of our study is that despite using a simpler grading system for the intensity of agglutination, we nonetheless noted a step-up in both severity of illness and mortality with increasing DD levels. This confirms the significance of DD in critical illness by suggesting a possible dose-response relationship between the degree to which the coagulation system is activated as manifest by the DD level and multiple, distinct clinical outcomes.
More generally, our findings add to the growing evidence underscoring the importance of the coagulation system in critical illness. For example, Rubin and colleagues21 demonstrated that elevated levels of von Willebrand antigen, a coagulation system marker of endothelial injury, were predictive of the development of ARDS in patients with nonpulmonary sepsis. Unlike this study, we focused on a more diverse group of patients, and assessed end points other than only ARDS. Similarly, Idell and colleagues22 reported enhanced levels of tissue-associated factor VII in BAL fluid from patients with ARDS. Bredbacka et al23 postulated that patients with signs of early hypercoagulation, as measured by elevated soluble fibrin levels, would progress to have more organ failures develop and have worse outcomes than patients with normal fibrin levels. Subjects with high fibrin levels had an in-hospital mortality rate of 46% compared to 14% in those with normal fibrin levels. Although our findings are in overall agreement with their results, the studies are not directly comparable. Our study population included only medical patients, while 40% of the study cohort evaluated by Bredbacka et al23 were either postoperative or obstetrical patients. Highlighting the importance of the other aspects of the coagulation in critical illness, Boldt et al24 noted that serial measurement of protein C may be an important biomarker for distinguishing sepsis from other causes of critical illness, while Hartman and coworkers25 reported that > 85% of patients with severe sepsis had acquired protein C deficiency. Finally, and most importantly, efforts to restore coagulation homeostasis with infusions of protein C improve survival in select patients with severe sepsis and organ failure.12 Further studies are therefore warranted to correlate the relationship with DD status and protein C levels in hopes of developing a rapid, bedside test that will identify patients most likely to benefit from therapies directed at the coagulation system. At present, though, DD status alone should not be used to determine whether to offer treatments directed at the coagulation abnormalities common in critically patients.
The involvement of both proinflammatory and anti-inflammatory cytokines in critical illness is now widely appreciated. Levels of certain proinflammatory cytokines have been shown to correlate with severity of illness and to be predictive of mortality for critically ill patients. Presterel et al26
reported that daily alterations in the APACHE III score and the mortality probability model score paralleled changes in both TNF- and IL-6 levels. Additionally, TNF- levels at admission to the ICU in patients with sepsis are predictive of survival.27
Few studies, however, have expressly examined the relationship between DD and cytokine levels. Pape and coworkers28
demonstrated a correlation between markers of coagulation including DD and IL-6 in trauma patients, while Lopez-Aguirre et al29
confirmed a relationship between multiple markers of coagulation system activation and cytokines. However, our finding that proinflammatory cytokine levels are greater in those with higher DD levels shows the potential interrelationships between the coagulation system and inflammation. The absence of a relationship between IL-10 and DD suggests that DD does not reflect changes in the anti-inflammatory cascade that may also be active in critical illness.
For outcome prediction purposes, a number of scoring systems exist to aid clinicians, and include the APACHE score, the mortality probability model score, the multiple organ dysfunction score, and the sequential organ failure assessment score.12 30 31 32 Only some of these scoring systems incorporate measures of the coagulation system. Those that do address coagulation focus only on the platelet count. Despite this, we observed that the DD levels performed as well as the APACHE II score as a screening test for mortality. In light of the potential predictive power of DD, future efforts to develop organ-failure indexes and severity-of-illness scores should formally include measures of DD. This would hopefully aid in further refining the accuracy of these systems.
Our study has several limitations. First, it was restricted to one center and to only medically ill subjects. Although the relationship between DD and outcomes has been shown now in several studies, little data exist about DD in critically ill surgical patients. Surgery of and within itself will likely lead to the presence of DD in the circulation, and therefore the utility of DD in surgical patients may be less. Second, our study had a limited sample size. The impact of this on our observations can been seen in the range of the 95% CI surrounding the relative risk for mortality. Third, we measured DD only at hospital admission. Sequential assessments of DD status correlated with outcome would provide stronger evidence of a causal relationship between coagulation dysregulation, cytokine levels, and outcomes. In other words, the study design allows us to note an association but does not prove the existence of a causal relationship. Fourth, we used a semiquantitative DD assay system. If we had employed a purely quantitative system, the relationships we noted may have been less apparent. In other words, the cut points in the SimpliRed system could theoretically be ideal for the discriminating among the outcomes we measured or these thresholds could be the rather insensitive. Only a simultaneous comparison with a continuous, quantitative system would allow one to address this uncertainty. Finally, we did not rely on enzyme-linked immunosorbent assays for the measurement of cytokines. Chemillumenescence and electrochemillumenesence are thought to be comparable to traditional enzyme-linked immunosorbent assay systems.
In summary, we observed that DD status on admission to the ICU correlated with multiple clinical end points to include severity of illness, organ failure, and mortality. The presence of DD was also associated with proinflammatory cytokine levels. DD performed as well as the APACHE II score at predicting mortality.
|
Footnotes |
|---|
The opinions expressed herein are not to be construed as official or as reflecting the policy of either the Department of Defense or the Department of the Army.
Received for publication June 8, 2001. Accepted for publication October 3, 2001.
|
References |
|---|
|
TOPAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
: early predictors in patients with septic shock. Intensive Care Med 26,305-313[CrossRef][ISI][Medline]
This article has been cited by other articles:
|
|
 |
|
  K. M. Stamou, K. G. Toutouzas, P. B. Kekis, S. Nakos, A. Gafou, A. Manouras, E. Krespis, S. Katsaragakis, and J. Bramis Prospective Study of the Incidence and Risk Factors of Postsplenectomy Thrombosis of the Portal, Mesenteric, and Splenic Veins Arch Surg, July 1, 2006; 141(7): 663 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. McDermott, K. Liu, J. M. Guralnik, L. Ferrucci, D. Green, P. Greenland, L. Tian, M. H. Criqui, C. Lo, N. Rifai, et al. Functional decline in patients with and without peripheral arterial disease: predictive value of annual changes in levels of C-reactive protein and d-dimer. J. Gerontol. A Biol. Sci. Med. Sci., April 1, 2006; 61(4): 374 - 379. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Querol-Ribelles, J. M. Tenias, E. Grau, J. M. Querol-Borras, J. L. Climent, E. Gomez, and I. Martinez Plasma d-Dimer Levels Correlate With Outcomes in Patients With Community-Acquired Pneumonia Chest, October 1, 2004; 126(4): 1087 - 1092. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. McDermott, P. Greenland, J. M Guralnik, L. Ferrucci, D. Green, K. Liu, M. H Criqui, J. R Schneider, C. Chan, P. Ridker, et al. Inflammatory markers, D-dimer, pro-thrombotic factors, and physical activity levels in patients with peripheral arterial disease Vascular Medicine, May 1, 2004; 9(2): 103 - 105. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
