|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TRANSLATIONAL PHYSIOLOGY
1Cardiovascular Research Institute and 3Division of Occupational and Environmental Medicine, Division of Pulmonary and Critical Care Medicine, University of California, San Francisco, California; 2University of Texas Health Center at Tyler, Department of Biochemistry, Tyler, Texas; 4Pulmonary and Critical Care Unit, Department of Medicine, Massachusetts General Hospital, Boston, and 5Massachusetts General Hospital Biostatistics Center, Boston, Massachusetts
Submitted 31 October 2005 ; accepted in final form 4 May 2006
| ABSTRACT |
|---|
|
|
|---|
acute respiratory distress syndrome; extracellular matrix; stretch injury
The extracellular matrix, the region of the lung between the alveolar epithelium and the vascular endothelium, provides structural support to the lungs through collagens, glycoproteins, and proteoglycans (17). Elevations of extracellular matrix proteins such as procollagen type I, procollagen type III, hyaluronan, and laminin have been reported in the bronchoalveolar lavage and pulmonary edema fluid of patients with ALI implicating damage to the extracellular matrix (4, 15, 18).
The extracellular matrix also facilitates passive recoil of the lungs via the unique stretch properties of elastin (19). The lung is a major source of elastin in the body, and in the absence of some pathological process, elastin is stable over the lifetime of an individual (11, 23, 26, 27, 31). The average adult, however, excretes between 1 and 5 mg of elastin per day in the urine arising from the lung as well as other tissues (26, 30). During lung epithelial and endothelial injury, as in ALI, elastin can be broken down by exposure to proteases such as neutrophil elastase (27). Elastin breakdown results in smaller fragments containing desmosine and isodesmosine, crosslinks unique to elastin (27). Desmosine-containing fragments are eliminated in the urine without further modification as small peptides and up to 15% as free desmosine (27). Formation of the crosslinks is a postribosomal event, and coupled with the fact that there is virtually no dietary absorption, inhalation, or other sources of desmosine in the body (26), urine desmosine is a selective marker of elastin breakdown. Desmosine has been studied as a marker of elastin breakdown in several chronic pulmonary conditions, including chronic obstructive pulmonary disease (COPD), cystic fibrosis, and chronic tobacco use (5, 12, 29). There is some variability in the desmosine levels when measured in certain chronic disease states such as COPD or cystic fibrosis compared with controls (2).
The purpose of the current study was to evaluate urine desmosine as a surrogate marker for elastin breakdown in a large group of patients with ALI. Urine desmosine levels were measured in the subjects enrolled in the recent, multicenter randomized controlled trial of lower tidal volume ventilation in ARDS (1). Three related hypotheses were tested similar to prior studies of biomarkers in patients with ALI included in the ARDS Network trial (8, 20, 21, 32). First, we evaluated whether baseline urine desmosine was higher in patients who died compared with those who survived. Second, we assessed whether urine desmosine levels correlated with clinical markers of disease severity and other outcome measures, such as ventilator-free days and organ failure-free days. Finally, we evaluated whether urine desmosine levels demonstrated an attenuated rise in the group of patients that received the 6-ml/kg predicted body weight ventilation strategy compared with those that received 12-ml/kg predicted body weight reflecting less stretch injury in patients who received the lower tidal volume ventilation strategy.
| METHODS |
|---|
|
|
|---|
|
Statistical analyses.
All analyses were done using SAS software. Sample results were reidentified, and information from the ARDS Network database was utilized for analysis. The ratio of urine desmosine to urine creatinine was used for all samples to control for the effect of urine dilution. This ratio has been shown in prior work to provide the same interpretation of the data as determining total urine desmosine in a 24-h urine collection (25). The ratio of urine desmosine-to-creatinine was not normally distributed in our dataset. We presented raw data results in Fig. 1 and ![]()
![]()
Table 5. Because of the abnormal data distribution, we presented the median and interquartile ranges. All statistical analysis was done using log transformation to create a more normally distributed dataset for statistical analysis. We evaluated the affect of the log of the urine desmosine-to-creatinine ratio on the risk of death using logistic regression and then adjusted for ventilator group assignment because ventilator group affected mortality. We compared baseline urine desmosine-to-creatinine ratios in patients with direct vs. indirect pulmonary insults to determine if there were differences based on the category of the medical condition that predisposed to ALI. Pearson correlation coefficients controlling for ventilator group were determined for the baseline log urine desmosine-to-creatinine ratio and other clinical outcomes including ventilator-free days and organ failure-free days. Pearson correlation coefficients were also determined for baseline log urine desmosine-to- creatinine with clinical variables known to correlate with disease severity in ARDS, including PaO2/FIO2, plateau airway pressure, non-pulmonary organ failures, APACHE III score, and age. Finally, the study group was separated by ventilator group assignment, i.e., 6 vs. 12 ml/kg tidal volume, to determine whether tidal volume specifically affected the log urine desmosine level. A repeated measures linear regression model was used to compare the average log of the urine desmosine-to-creatinine ratio on days 1 and 3 while controlling for baseline value. Results were considered to be statistically significant if the P values were <0.05.
|
|
|
|
|
| RESULTS |
|---|
|
|
|---|
Baseline predictive value. Higher baseline values of the ratio of urine desmosine to urine creatinine were associated with a higher risk of death (odds ratio 1.39 per log increment, 95% confidence interval 1.04-1.85; Table 2). We then tested the baseline predictive value of urine desmosine to urine creatinine in a multivariate analysis that included the ventilator group assignment. This multivariate analysis again demonstrated that the baseline log urine desmosine-to-creatinine ratio still had predictive value for mortality (Table 2). We were unable to compare the urine desmosine results to a control group of ventilated patients without ALI. We did determine that the mean value of desmosine-to-creatinine for all 579 patients in this study was 129 pmol desmosine/mg creatinine. This value was higher than the mean urine values reported in a study that compared COPD patients, COPD patients that continued to smoke, and age-matched normal controls that showed no statistical difference in mean desmosine levels between these three groups (mean 28.4 to 35.5 pmol desmosine/mg creatinine) (2). In addition, five healthy individuals provided urine samples over 3 days for desmosine analysis. The healthy individuals had a much lower concentration of urine desmosine-to-creatinine (mean 27.9 pmol desmosine/mg creatinine) than the mean levels of patients from the ARDS Network.
Baseline levels in direct vs. indirect pulmonary insult. Baseline levels of urine desmosine-to-creatinine were compared in study subjects who had a direct pulmonary insult predisposing to ALI vs. an indirect pulmonary insult using t-test comparison. The analysis demonstrated no significant difference in baseline urine desmosine in patients with direct vs. indirect pulmonary insult (P = 0.27).
Correlation with ventilator-free days and organ failure-free days. We evaluated whether urine desmosine would correlate with other outcomes measured in the original ARDS Network trial. Therefore, the correlation coefficients, controlling for ventilator group, were determined between the baseline log urine desmosine-to-creatinine ratio and specific organ failure-free days as well as ventilator-free days (Table 3). All of the correlations in this table were negative. These negative correlations mean that higher urine desmosine-to-creatinine ratios were associated with fewer ventilator-free days or organ failure-free days, i.e., higher log urine desmosine-to-creatinine was associated with worse outcomes. The correlations were weak but statistically significant for all variables except renal failure.
Correlation with markers of disease severity. Certain clinical characteristics have been identified as markers of disease severity in ALI, whereas others are predictive of mortality in ALI. We tested several of these variables using correlation coefficients to determine whether the baseline log urine desmosine-to-creatinine ratio correlated with these variables. Only two of these clinical variables, age and APACHE III score, showed a significant correlation with the baseline urine desmosine-to-creatinine ratio (Table 4).
Modulation of desmosine levels by tidal volume group. A repeated measures model was used to compare the changes in the mean of the log of urine desmosine to creatinine over the first three study days in each ventilator group. The rise in urine desmosine was attenuated by a tidal volume of 6 ml/kg compared with a tidal volume of 12 ml/kg (Table 5). The median values for both the 6- and the 12-ml/kg group increased over the course of the study, but the rise in the 6-ml/kg group was attenuated compared with the 12-ml/kg group (Fig. 1).
| DISCUSSION |
|---|
|
|
|---|
Prior work has suggested elastin breakdown occurs in ALI. Tenholder and coworkers (31) studied urine desmosine levels in patients with ARDS as a marker of elastin breakdown. The results of that study showed that ARDS patients had a higher ratio of urine desmosine to serum creatinine compared with patients with cardiogenic pulmonary edema. Thus the authors concluded that urine desmosine may be useful to discriminate between different kinds of pulmonary edema when the clinical picture is not clear. Fan and Nagle (9) investigated 100 patients who died with ARDS and in 30 patients found elastin-staining laminar structures within the interstitium that were remnants of elastic fibers. An intriguing finding of this study was the fact that elastin-staining laminar structures were only observed in cases of rapidly developing acute respiratory failure. The results of the current study in ARDS Network patients showed that higher levels of urine desmosine-to-creatinine correlated with worse clinical outcomes, including fewer ventilator-free days and organ failure-free days, suggesting that more severe damage to the extracellular matrix occurred in the most critically ill ALI patients.
There are several reasons why damage to the extracellular matrix might occur. The initial insult that leads to ARDS may directly affect the extracellular matrix. For example, a stimulus may be transmitted by a cytokine cascade leading to breakdown of matrix proteins. Alternatively, the extracellular matrix may be a bystander in ongoing damage to the vascular endothelium and alveolar epithelium that border the extracellular matrix. If this is true, then desmosine may be considered a marker of the severity of endothelial and epithelial injury that occurs in ALI. It is also possible that ventilation with positive pressure may lead to some mechanical damage to the extracellular matrix with stretch injury and breakdown of extracellular matrix proteins. Delivery of a 12-ml/kg tidal volume may overdistend the compliant, less injured alveoli, exacerbating the initial injury with concomitant stretch injury. Because most patients were treated with tidal volumes higher than 6 ml/kg before initiation of the ARDS Network study protocol, subjects may have received tidal volumes high enough to overdistend the functional regions of the lung and thereby may have suffered stretch injury before enrollment in the ARDS Network study.
The second important finding of this study was the attenuation in the rise in urine desmosine over the course of the trial. Subjects ventilated with the lower tidal volume ventilation not only had a lower mortality rate but they also had lower levels of the log urine desmosine-to-creatinine in day 1 and day 3 values as demonstrated by a strong statistical trend that nearly reached significance (P = 0.053). Desmosine is unique to elastin and elastin is in high abundance in the lung. Because the only difference between the groups in this study was the ventilation strategy, it is reasonable to infer that the elastin breakdown occurred in the lung. There is support for this interpretation in the literature. Janoff et al. (13) evaluated urine desmosine in sheep who underwent bronchoalveolar lavage with varying doses of elastase to determine whether urine desmosine could be used as a marker of lung elastin breakdown. Their results showed that baseline urine desmosine concentration in sheep urine was similar to humans. When elastase was instilled directly into the lung, the quantity of desmosine excreted above the baseline level increased linearly with increasing dose of elastase administered. Furthermore, urine desmosine positively correlated with a decrease in perfusion and ventilation in the lung compared with baseline, substantiating that the desmosine measured in the urine originated from injury to the lung. The authors of that study convincingly demonstrated that pathological changes to proteins in the lung could be reliably measured in the urine.
The attenuation in urine desmosine concentrations over 3 days in the current study probably reflected less damage to the extracellular matrix of the lung as a result of the lower tidal volume ventilation coupled with plateau pressure limitation (<30 cmH2O). Pierce et al. (22) studied desmosine levels in preterm lambs and found higher whole lung levels of desmosine and higher mRNA of tropoelastin in ventilated preterm lambs compared with normal term lambs. Furthermore, the preterm lambs with lung injury that received lower tidal volume ventilation (5 ml/kg) had an attenuated rise in desmosine compared with those that had received standard (15 ml/kg) tidal volume ventilation. The results of this experimental study demonstrated both a role for elastin breakdown in lung injury as well as the contribution of stretch injury to the quantity of elastin breakdown.
In the current study, desmosine did not correlate well with markers of disease severity, only showing weak correlations with APACHE III and age. The process of damage to the extracellular matrix may be different than the processes that affect other clinical parameters, such as compliance or oxygenation. Poor oxygenation may be the result of extensive epithelial injury with alveolar filling rather than the result of damage to the extracellular matrix. Alternatively, the urine desmosine measured in this study may reflect elastin breakdown occurring in other parts of the body such as skin or blood vessels. In either case, the results suggest that desmosine may be more useful in understanding the pathogenesis of ALI and less useful as a marker of disease severity.
Although the results of our study implicate a role for damage to the extracellular matrix in ALI patients, there were limitations to consider. First, we were only able to include 579 of the 861 patients included in the original trial. It is possible that the subset we studied was not representative of the whole group. However, the data in Table 1 demonstrate that this large subset was similar in terms of baseline characteristics to the original group included in the ARDS Network trial of lower tidal volume ventilation. Another limitation is that the study design did not allow for discrimination between elastin breakdown that occurred in the extracellular matrix of the lung vs. elastin breakdown that may have occurred in other elastin-rich tissues such as systemic blood vessels or skin tissue. However, since the lung is a major source of elastin in the body (31) and since the intervention in this study was a ventilator strategy, it appears most likely that lower elastin levels represent less damage to the extracellular matrix of the lung. Furthermore, in the Tenholder et al. study (31), the ARDS patients had a significantly higher level of urine desmosine to serum creatinine than critically ill subjects without ARDS and subjects with cardiogenic pulmonary edema. Our results suggest that elevation in urine desmosine, referenced to urine creatinine, is not simply the result of severe illness or pulmonary edema alone but rather is more specific to ALI. One final limitation is that we cannot distinguish association from causation. Damage to the extracellular matrix could be a major factor that worsens ALI and plays a role in ongoing lung damage in ALI. Alternatively, elastin breakdown may simply be a marker of damage that is occurring to tissues surrounding the extracellular matrix, specifically the vascular endothelium and the alveolar epithelium.
In summary, the results of this study provide evidence that worse clinical outcomes are associated with increased elastin breakdown in patients with ALI. In particular, elevated levels of urine desmosine early in the course of ALI are associated with higher mortality rates. Elastin breakdown is attenuated by a lung-protective lower tidal volume ventilation strategy. The results also illustrate the utility of urine samples as a noninvasive source of biological samples for the investigation of matrix breakdown in patients with ALI. Therefore, clinical investigations that utilize urine for biological evaluation should be of value in future studies of ALI.
| GRANTS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
Clinical Coordinating Center: Massachusetts General Hospital, Harvard Medical School, David A. Schoenfeld, Ph.D.,* B. Taylor Thompson, M. D., Marek Ancukiewicz, Ph.D., Douglas Hayden, M. A., Francine Molay, M. S. W., Nancy Ringwood, B. S. N., R. N., Gail Wenzlow, M. S. W., M. P. H., and Ali S. Kazeroonin, B. S.
NHLBI Staff: Dorothy B. Gail, Ph.D., Andrea Harabin, Ph.D.,* Pamela Lew, and Myron Waclawiw, Ph.D.
*Steering Committee: Gordon R. Bernard, M. D., Chair, Principal Investigator from each center as indicated by an asterisk.
Data and Safety Monitoring Board: Roger G. Spragg, M. D., Chair, James Boyett, Ph.D., Jason Kelley, M. D., Kenneth Leeper, M. D., Marion Gray Secundy, Ph.D., and Arthur Slutsky, M. D.
Protocol Review Committee: Joe G. N. Garcia, M. D., Chair, Scott S. Emerson, M. D., Ph.D., Susan K. Pingleton, M. D., Michael D. Shasby, M. D., and William J. Sibbald, M. D.
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. E. Levitt, M. K. Gould, L. B. Ware, and M. A. Matthay Analytic Review: The Pathogenetic and Prognostic Value of Biologic Markers in Acute Lung Injury J Intensive Care Med, May 1, 2009; 24(3): 151 - 167. [Abstract] [PDF] |
||||
![]() |
D. E. McClintock, L. B. Ware, M. D. Eisner, N. Wickersham, B. T. Thompson, M. A. Matthay, and the National Heart, Lung, and Blood Institute ARDS Higher Urine Nitric Oxide Is Associated with Improved Outcomes in Patients with Acute Lung Injury Am. J. Respir. Crit. Care Med., February 1, 2007; 175(3): 256 - 262. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |