| 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 University of California, San Francisco, CA.
Correspondence to: Michael A. Matthay, MD, Critical Care Medicine, M-917, University of California, 505 Parnassus Ave, San Francisco, CA 94143-0624; e-mail: mmatt{at}itsa.ucsf.edu
The overall objective of this summary will be to highlight the major clinical and experimental work that was presented at the 1998 conference on acute lung injury. This overview will focus on major themes that represent an important link to prior and current research and provide direction for future work in the field.
The summary is divided into five sections. The first section reviews new information regarding definitions and epidemiology of clinical acute lung injury and ARDS. The second section considers experimental and clinical studies that have provided new insights into the pathogenesis of acute lung injury. The third section reviews new data regarding the mechanisms that contribute to the resolution of acute lung injury. The fourth section considers the results of several clinical studies that examined the potential prognostic and pathogenic significance of biological markers that were measured in the plasma, edema fluid, or BAL of patients with or at risk of developing ARDS. The fifth section provides a brief overview of clinical trials that have been completed since the last Aspen Lung Conference on Acute Lung Injury in 1993, as well as a brief discussion of ongoing and potential future clinical trials.
Clinical Definitions and Epidemiology
Dr. Petty led off this year's conference with a brief historical perspective of the term, adult respiratory distress syndrome (ARDS). His original intention was not to exclude children; in fact, two of the first patients he described in his original description of ARDS were 11 and 15 years old.1 Therefore, Dr. Petty agreed that ARDS should refer to the acute respiratory distress syndrome.
Since the 1993 Aspen Lung Conference 5 years ago, a North American-European Consensus Conference considered the problems encountered by both clinicians and investigators in trying to define ARDS or clinical acute lung injury. After several meetings, there was a nearly unanimous recommendation that the following definitions be used for the diagnosis of clinical acute lung injury and for ARDS. Both definitions require the presence of bilateral pulmonary infiltrates on the chest radiograph. In the case of clinical acute lung injury, the cutoff for arterial oxygenation is a PaO2/fraction of inspired oxygen (FIO2) < 300 whereas for ARDS, the PaO2/FIO2 cutoff was < 200. The conference did not recommend routine measurement of pulmonary arterial wedge pressure to rule out cardiogenic or volume overload pulmonary edema. Instead, clinicians were advised to use history, physical examination, the chest radiograph, and noninvasive methods, such as a cardiac echocardiogram, to evaluate for the presence or absence of hydrostatic pulmonary edema. If clinical evaluation alone was insufficient, then pulmonary arterial catheterization was recommended.2
In 1988, investigators from the University of California at San Francisco proposed a four-point lung injury scoring system that provided a semiquantitative method for measuring the severity of pulmonary physiologic dysfunction based on arterial oxygenation, the level of positive end-expiratory pressure (PEEP), static thoracic compliance, and the extent of infiltrates on the chest radiograph.3 Although this four-point acute lung injury score has been widely used to quantify acute lung injury, there is no evidence that the severity of pulmonary dysfunction early in the course of acute lung injury (day 1) identifies which patients are less likely to survive. In two recent prospective studies on patients with acute lung injury, neither the composite four-point lung injury score nor its individual components were predictive of mortality when analyzed on day 1 of acute lung injury. In one study that included both medical and surgical patients with all the usual causes of acute lung injury except trauma, the most important predictors of a poor outcome were nonpulmonary organ system dysfunction that was present prior to admission to the ICU, chronic liver disease, and sepsis.4 In the other study of medical patients,5 chronic liver disease and sepsis were the most important predictors of a poor outcome. In both studies, the cutoff of a PaO2/FIO2 < 300 was used to define acute lung injury. Even with this less severe oxygenation cutoff, mortality remained high at 58% in both studies, thus providing partial validation of the North American-European consensus conference recommendation that acute lung injury be used clinically. Thus, the prior definitions of ARDS that required more severe oxygenation criteria did not necessarily define a different population of patients. There is also new evidence from the National Institutes of Health (NIH) ARDS Network that the incidence of nonpulmonary organ failure is similar in patients with a PaO2/FIO2 < 200 compared to those with a PaO2/FIO2 < 200 to 300.6 However, patients with a PaO2/FIO2 < 200 have a higher incidence of systolic hypotension (< 100 mm Hg). Based on preliminary analyses, all the clinical disorders associated with development of acute lung injury (ie, sepsis, pneumonia, aspiration) were similar regardless of the initial PaO2/FIO2 ratio, except trauma, which occurred more commonly in patients with a less severe oxygenation defect (PaO2/FIO2 of 200 to 300).
At this year's conference, several participants recognized the need to provide better guidelines for interpretation of the chest radiograph in patients with suspected clinical acute lung injury. The term, bilateral pulmonary infiltrates consistent with pulmonary edema, is qualitative and can be confused with atelectasis, a mass lesion, or pleural effusions. A recent study by Dr. Rubenfeld was carried out in which clinicians and investigators who are experts in the diagnosis of ARDS were asked to review several candidate chest radiographs to determine whether they met criteria for bilateral pulmonary infiltrates that were consistent with acute lung injury. Interestingly, there was considerable disagreement among the experts regarding what constitutes bilateral pulmonary infiltrates; also, the interobserver variability was high.7 A conference on this issue was recently convened in Seattle; considerable effort is being made to provide more uniform criteria for interpreting the chest radiograph in clinical acute lung injury.
Some additional methods for classifying patients with acute lung injury may be needed. Several attendees at the conference were concerned that the current definitions recommended by the North American-European Consensus Conference do not include the level of PEEP or a measure of lung compliance. Others were concerned that the definition does not specify the clinical disorder or disorders that are associated with the development of lung injury (eg, sepsis, pneumonia, aspiration of gastric contents, major trauma). This is a potentially important issue since most studies have found that sepsis-induced lung injury has a higher mortality than other clinical disorders.4 8 9 Furthermore, ventilated patients with severe hydrostatic pulmonary edema meet most of the pulmonary physiologic criteria for acute lung injury, except that hydrostatic pulmonary edema fluid has a lower protein concentration than edema fluid from patients with increased permeability pulmonary edema or acute lung injury.10 11 In fact, the Simplified Acute Physiology Score II is similar in patients with severe hydrostatic pulmonary edema when compared to ventilated patients with acute lung injury and increased permeability pulmonary edema.10 11 These observations emphasize the need for clinical methods to determine which patients with pulmonary edema have an increase in endothelial and epithelial permeability to protein, as some investigators have contended.12 Pulmonary edema fluid cannot be easily collected from all patients, and therefore remains primarily a research tool. BAL is also primarily a research tool. Other methods to measure lung vascular permeability are more invasive and difficult to apply to all patients. Therefore, there is no obvious solution to this problem currently.
There have been some interesting new observations regarding the epidemiology of clinical acute lung injury. Evidence from Dr. David Guidot and coworkers indicated that chronic alcoholics have an increased incidence of ARDS. The mechanism for this observation is not clear, although experimental data suggest that there may be abnormalities in alveolar epithelial surfactant production or alveolar epithelial permeability to protein. In addition, new data were presented that suggested the patients with diabetes mellitus may have a decreased incidence of ARDS.
From studies at Harborview Hospital at the University of Washington in Seattle, Dr. Kenneth Steinberg reported that there has been a decline in the fatality rate in patients with ARDS over the last few years.13 The explanation for the decline in ARDS mortality is not evident at this time. Changes in fluid management, altered ventilator management, enteral vs parenteral nutrition, more potent antibiotics, or a decrease in the use of invasive hemodynamic monitoring could be important.
New work presented from the University of Washington at Seattle indicated that there may be a persistent impairment in the functional status and mental acuity scores in survivors of ARDS.14 Previously, most studies have focused on recovery of pulmonary function, but the new work promises to provide a more global assessment of the functional status of patients who recover from ARDS.
There was also some discussion about the incidence of clinical acute lung injury in North America and Europe. Although a few studies have reported incidence rates for ARDS in Utah, the Canary Islands, and parts of Europe, most participants agreed that prospective, well-designed epidemiologic studies are needed to determine the actual incidence of acute lung injury. If the annual incidence in the United States is between 100,000 and 150,000 cases, then the mortality from ARDS is between 40,000 and 70,000 per year. There are approximately 40,000 to 45,000 annual deaths a year from breast cancer in the United States, so the clinical problem is of major importance to the biomedical community. A prospective epidemiologic study in King County is being planned by investigators at the University of Washington in Seattle.
Pathogenesis
Since the 1993 Aspen Acute Lung Injury Conference, there has been
substantial progress in understanding the complex inflammatory
mechanisms that mediate acute lung injury. The primary focus has
continued to be on neutrophil-mediated lung injury, although work
presented by Dr. Dayer emphasized the potential contribution of T
lymphocytes and monocytes, particularly in the later phase of lung
injury. There is new information on the important role of
anti-inflammatory cytokines, including interleukin-10 (IL-10) and
IL-1ra. Clinical studies presented by Dr. Thomas Martin illustrated the
complexity of the alveolar environment in the early phase of acute lung
injury. For example, in addition to proximal proinflammatory cytokines
such as tumor necrosis factor-
(TNF-) and IL-1, there are
cytokine inhibitors in the plasma and the airspaces of the lung. In
addition, IL-6 is present in abundance, and much of IL-6 is bound to
its receptor, GP130. Interestingly, IL-6 is one example of a cytokine
in which the cytokine and receptor complex have an agonist, not an
antagonist, effect.
Overall, there has been increasing awareness of the complexity of the
cytokine response in patients with acute lung injury, a finding that
Dr. Martin aptly termed "cytokine storm." Several additional
cytokines have been described, and the mechanisms that regulate their
production and neutralization are incompletely understood. In parallel
with the increased understanding of the large number of cytokines in
the acute inflammatory environment in ARDS, there is also new
information regarding the signaling effects of cytokines. Dr. David
Goeddel explained that TNF- signaling is a receptor-dependent
process that is mediated by several intracellular proteins, including
TRAFs, cIAPs, and death domain-containing proteins. Some of the TNF-
signaling pathways activate nuclear factor-
B, an important
modulator of the inflammatory response. The transcriptional effects of
TNF- and IL-1 signaling seem to overlap; future work promises to
determine which genes are differentially activated by TNF- and IL-1,
and how signaling regulates the timing and magnitude of the
inflammatory response.
More work has been done on the potential role of oxidants in mediating acute lung injury, although the field needs more direct evidence of oxidant-mediated lung injury. One of the challenges for the next few years will be to directly assemble evidence of oxidative injury in the lungs in patients with acute lung injury. There is experimental evidence that reactive oxygen species can cause injury to the endothelial and epithelial barriers of the lung. Furthermore, there is recent work from Matalon and colleagues15 that peroxynitrate accumulates in the alveolar compartment after both experimental and clinical lung injury. Also, there is evidence from Dr. Matalon's laboratory that peroxynitrate can directly alter the function of an epithelial sodium channel (ENaC), a critical sodium transport channel in pulmonary epithelium. New work presented from Dr. David Ingbar's laboratory indicated that nitrotyrosine was bound to the ß1-subunit of NaK-A adenosine triphosphatase (ATPase) following 60 h of hyperoxia in rats.
Although there is considerable evidence that neutrophils are an important source of reactive, oxygen species, Dr. Aron Fisher presented evidence that endothelial nicotinamide adenine dinucleotide phosphate (reduced form) (NADPH) oxidase is a source of reactive oxygen species in a rat lung ischemia model. In a different model, Dr. David Pearse presented evidence that inhibition of leukocyte NADPH oxidase in ischemia-reperfused sheep lungs could prevent injury.
An important area of basic research has been to understand the role of adhesion molecules in mediating neutrophil accumulation in the acutely injured lung. Dr. Claire Doerschuk presented evidence that decreased neutrophil deformability contributes to intrapulmonary sequestration in the lung under some pathologic conditions. The decreased deformability of neutrophils can be induced by C5a, leukotriene B4, IL-8, and bacterial products. Interestingly, neutrophils are trapped in the lung by CD11- and CD18-dependent mechanisms in experimental Gram-negative pneumonia in mice, but neutrophil sequestration occurs by CD11- and CD18-independent mechanisms in Gram-positive pneumonia, hyperoxic lung injury, and following acid-induced lung injury.16 17 Interestingly, a double mouse knockout of E and P selectins had no effect on neutrophil accumulation in the lung.
Although controversy still exists regarding the role of neutrophils in mediating all causes of acute lung injury, it is clear that neutrophils play a role in most cases of experimental-induced lung injury. However, the neutrophil can sequester in the pulmonary microcirculation and be chemoattracted to the alveoli without causing endothelial or epithelial injury, providing that the neutrophil is not activated.18
Dr. Guy Zimmerman discussed how degranulation of neutrophils can produce injury at the endothelial barrier. In addition, Dr. Zimmerman reviewed work from his own laboratory and other investigators that focused on identifying putative neutrophil degranulating factors. Interestingly, after an extensive search with expression cloning and library screening, he presented evidence that granulocyte-macrophage colony-stimulating factor was the most important degranulating factor. In addition, Dr. Kurt Albertine presented histopathologic evidence of granulocyte macrophage-colony-stimulating factor protein expression in the pulmonary venules of patients with ARDS. These morphologic studies emphasized the need for more pathologic studies of lungs of patients with ARDS in order to enhance understanding of the mechanisms in the pathogenesis of clinical acute lung injury. Dr. Zimmerman made the important observation that the new gene array technologies will make it possible to identify several candidate gene products in acute lung injury. However, as he explained, the challenge will be to determine the biological mechanisms, not just a novel complementary DNA.
In addition to ongoing work by several laboratories on the mechanisms of neutrophil adhesion in the lung, mediated by adhesion molecules as well as intrinsic biophysical properties of neutrophils (deformability) and the specific geometry of the pulmonary microcirculation, other progress has been made in understanding unanticipated functions of adhesion molecules. For example, Dr. Gregory Downey presented evidence that L-selectin plays an important role as a signaling molecule in addition to its contribution in the systemic capillaries to rolling, the initial step in adhesion. Dr. Downey's work demonstrated significant differences between suspended and adherent neutrophils. With L-selectin binding, there is an oxidative burst that alters the biophysical properties of neutrophils and CD11, 18 expression with a subsequent increase in intercellular adhesion molecule-1 (ICAM-1) expression.
Over the last 5 years, several experimental studies have provided some
evidence to support the hypothesis that ventilator-induced lung injury
may occur under some conditions.19
Although the
experimental models do not replicate all of the clinical conditions,
the studies are predicated on the premise that some alveoli in patients
with ARDS are exposed to high airway pressures and a high tidal volume
that may worsen or compound the lung injury. Slutsky and
Tremblay20
presented evidence that cytokines can be
upregulated in the alveolar epithelium in isolated rat lungs without
perfusion in the presence of a high tidal volume. The
immunohistochemistry data indicated that much of the TNF- was being
released from alveolar macrophages, although some originated from
pulmonary epithelium. From another perspective, Dr. Kevin Creamer
presented data in which an attempt was made experimentally to maintain
alveolar distention at an optimal level with higher levels of PEEP with
the hypothesis that this approach would protect the lung against
neutrophil accumulation and injury. Although there was improved lung
compliance with this approach, there was no change in the cellular
infiltration or the magnitude of acute lung injury.
Overall, the results from this conference indicate that chemokine
activity in the acute lung injury may have multiple biological effects.
On the one hand, proximal cytokines clearly prepare the endothelium and
the epithelium for neutrophil adhesion by upregulating endothelial and
epithelial-specific adhesion molecules. In addition, TNF- and IL-1
have important signaling properties that are mediated through nuclear
factor-B that result in secondary signals that amplify inflammation.
Several neutrophil chemotactic factors are released; the most potent
neutrophil chemotactic factors appear to be IL-8 and epithelial
neutrophil activating protein-78. The role of IL-6 is unclear, but
large quantities are present in the lung and the net biological effect
is not well understood. Interestingly, an IL-6 overexpressing
transgenic mouse preparation protected against hyperoxic lung injury.
Other cytokines, such as IL-10, also appear to have a protective
anti-inflammatory effect. Furthermore, Dr. Aaron Waxman from Yale
University presented evidence that an IL-11 overexpressing mouse
protected against hyperoxic lung injury, although it is not clear if
the protection was mediated by antiapoptotic effects or by other
mechanisms. Finally, there are other functions of cytokines that have
been recently appreciated, including a potential role for IL-8 in
angiogenesis and repair, as discussed by Dr. Robert Strieter.
Interestingly, Dr. Seamus Donnelly presented evidence that macrophage
migration inhibitory factor may have an important role in opposing the
glucocorticoid-induced downregulation of early cytokine expression.
Several other presentations provided important new insights into acute lung injury. For example, Dr. Paul Noble presented evidence that hyaluronan fragments induce plasminogen activation inhibitor-1 in murine alveolar macrophages, thus providing a potential mechanism for impaired fibronolytic activity in acute lung injury. Also, Dr. Mitchell Olman presented evidence that proteolytic fibrin fragments have an important role in regulating plasminogen activator inhibitor-1, a determinant of fibrinolysis in the injured alveoli with hyaline membrane formation. The potential contribution of several growth factors in remodeling of the fibrotic and injured airspaces was discussed by Dr. Galen Toews. Experimental studies point to a potential beneficial effect of several epithelial specific mitogens, such as hepatocyte growth factor and keratinocyte growth factor (KGF). However, the critical steps in repair and remodeling of the injured lung are not known. The process of fibrosis and its resolution in the lung is complex. For example, Dr. William Parks presented evidence that lung elastin expression is regulated by a unique posttranscriptional mechanism that involves binding to a developmentally regulated cystolic protein.
There was also evidence from Dr. Augustine Choi that hemoxygenase can protect against experimental hyperoxic lung injury using in vitro transfection in rats. Surprisingly, the protective effect may be partly mediated by carbon monoxide, a major byproduct of hemoxygenase catalysis. In fact, there were several exciting experimental approaches that prevented or attenuated hyperoxic lung injury in mice or rats. In light-hearted moments at the conference, there was animated discussion of how successful we are in preventing or treating acute lung injury from hyperoxia in mice; if only the same success could be achieved at the bedside in patients with ARDS. Even though the murine models have limitations, it is clear that the pulmonary scientific community is moving rapidly to using functional genomics to test putative mechanisms of acute lung injury as well as protective and therapeutic strategies.
Resolution of Lung Injury
Rapid progress has been made in understanding the basic transport mechanisms that regulate clearance of alveolar edema fluid.15 19 20 Sodium is actively transported, primarily by sodium channels on the apical surface of type II cells. Recently, an important ENaC was cloned. This channel is widely distributed in epithelia in several organs including the lung epithelium. After sodium is taken up by ENaC and other sodium channels on type II cells, the sodium pumps (NaK-ATPase), located on the basolateral surface, transport sodium to the interstitium of the lung. New evidence indicates that water crosses the alveolar epithelium through specific transcellular water channels (aquaporins), three of which have been localized to the endothelium and epithelium of the lung (aquaporins 1, 4, and 5).21 22 There is also new evidence that both catacholamine-dependent and independent mechanisms can upregulate alveolar epithelial fluid clearance.21 22 Some of these observations may have clinical significance. For example, there is evidence that ß-adrenergic agonist therapy can upregulate alveolar fluid clearance in some experimental models of lung injury.23 24 Interestingly, a lipid soluble ß2-agonist may be superior in potency to a standard water-soluble agonist.25 At this conference, evidence was presented by Dr. Phillip Factor from Dr. Jacob Sznajder's research group that transfection of rats with NaK-ATPaseB1 in an adenoviral vector resulted in a dramatic increase in survival in hyperoxic rats, apparently by an increase in alveolar edema fluid clearance. Also, there is new evidence that two commonly used vasopressor agonists, dobutamine and dopamine, can upregulate alveolar fluid clearance.26 27 The dobutamine effect is mediated by stimulation of ß2-receptors, while the dopamine effect appears to be mediated in part by dopamine receptor stimulation that translocates more NaKATPase to the basolateral membrane. Dr. Carmen Guerrero presented evidence that the dopamine stimulation may also be mediated by mitogen-activated protein kinase signaling that increases the transcription rates for NaK-ATPase. Additional studies on the mechanisms that regulate NaK-ATPase under hyperoxic conditions were presented by Dr. Christine Wendt from Dr. Ingbar's research group at the University of Minnesota. Her studies identified a promoter region in MDCK cells (a kidney cell line) that is responsible for upregulation NaK-ATPase ß1-gene expression in the presence of hyperoxia.
Biological Markers in Clinical Lung Injury
A substantial number of studies have been carried out over the last 15 years to examine the pathogenetic and prognostic value of biological markers in patients at risk for developing ARDS or in the early phase of ARDS.28 The measurements have been made in the plasma, pulmonary edema fluid, or in BAL fluid. At this year's conference, there were promising new data.
Dr. V. Newman presented evidence that an integral membrane antigen of alveolar epithelial type I cells (HTI-56) could be quantified in the plasma and the pulmonary edema fluid of patients with clinical lung injury. The levels of this type I cell antigen were higher in the pulmonary edema fluid and the plasma of ARDS patients compared to control patients with hydrostatic pulmonary edema. In addition, Drs. Kelly Greene, Robert Mason, and Polly Parsons presented new data that higher levels of surfactant protein A (SP-A) were detected in the plasma of at-risk patients who progressed to develop ARDS than in patients who did not progress. Dr. Lorraine Ware presented evidence that ICAM-1 was twofold higher in pulmonary edema fluid of acute lung injury patients than control patients with hydrostatic edema. In addition, those lung injury patients who had a prolonged duration of assisted ventilation had a significantly higher levels of ICAM-1 in their edema fluid.
Finally, Dr. Richard Goodman reported that CXCR2 is downregulated in neutrophils from patients with sepsis, while CXCR1 expression was not reduced. These results suggest that IL-8/CXCR1 may function as the dominant ligand/receptor pair in patients with sepsis. These data may have relevance for targeting anti-inflammatory strategies in patients with sepsis.
Clinical Trials
In the last 5 years, there have been several major clinical trials that have tested new therapeutic approaches for patients with ARDS. Although there was not time at this year's lung conference to discuss the details of these trials, this report includes a brief summary of the results of some of the major trials.
A major development since the 1993 Aspen Lung Conference on Acute Lung Injury was the commitment of the Lung Division of the NIH to establish an ARDS Network of 10 university medical centers in the United States to carry out phase III trials of new therapeutic interventions for ARDS. The ARDS Network was officially established in 1994; the first clinical trials began in March 1996. Since then, the Network has enrolled > 850 patients in an ongoing phase III randomized multicenter trial to test the potential benefit of a lung protective ventilator strategy. In fact, in March 1999, this ventilator trial was stopped for efficacy with a dramatic reduction in mortality from approximately 39% in the 12 mL/kg tidal volume group to 31% in the 6 mL/kg in the low tidal volume group (p < 0.01). There was also a significant reduction in the duration of mechanical ventilation in the low tidal volume group. The results of this landmark study were presented at the American Thoracic Society meeting in San Diego, CA, in April 1999.
The ARDS Network also initiated in 1997 a phase III trial to test the value of glucocorticoid therapy in the treatment of established ARDS (after 7 days of ARDS). The Network has tested two pharmacologic therapies for ARDS, ketoconazole and lysofylline, but neither was effective. The Network is considering several other clinical trials in ARDS over the next few years, including a plan to test the clinical value of the pulmonary arterial catheter in patients with ARDS.
The commitment of the Lung Division of the NIH to support prospective clinical trials has substantially advanced the opportunity to discover and identify effective new treatments for patients with ARDS. Furthermore, the Network is also currently in the process of beginning pathogenesis studies to examine the biological markers that may be of pathogenetic or prognostic value when measured at baseline or after the development of acute lung injury. In addition, several other phase II and phase III clinical trials have been carried out in Europe, South America, and North America that have tested new treatment strategies for patients with ARDS. The details of these trials will be briefly summarized.
The first major new treatment for ARDS was the extracorporeal membrane oxygenation trial in 1974. Extracorporeal membrane oxygenation was not superior to conventional therapy and the mortality rate in both groups of patients was 90%.19 Over 10 years later, an alternative approach was evaluated using extracorporeal carbon dioxide removal. Although initial uncontrolled studies in Europe were promising, ultimately an NIH-sponsored study in the United States demonstrated no effect of extracorporeal carbon dioxide removal on morbidity or mortality in patients with ARDS.22 As indicated, prior phase III studies also showed no benefit of prophylactic (PEEP) or of high-frequency ventilation.22
As already discussed, some experimental studies suggested that a lung protective ventilatory strategy using lower tidal volumes might attenuate the degree of acute lung injury.19 20 Therefore, several clinical investigators became interested in the possible benefit of a low tidal volume, pressure-limited ventilation strategy. Initially, observational and phase II studies reported favorable results with a low tidal volume strategy, but prospective randomized trials were not available. In 1998, two phase III trials were completed. A Canadian study reported no difference in morbidity or mortality in 105 patients with ARDS when they were ventilated with a tidal volume of 7 mL/kg compared to a control group with 10 mL/kg.29 In addition, another prospective study from Dr. Brochard and colleagues described 107 patients and reported that a similar low tidal volume strategy had no favorable impact on morbidity or mortality.30 Although both of these studies were randomized and involved multiple centers, the number of patients in each study was small, especially compared to the large trial of 861 patients in the American study. Therefore, the positive results of the recently completed American NIH phase III trial of low tidal volume (6 mL/kg vs 12 mL/kg) should provide a definitive answer to the hypothesis that a low tidal volume strategy can be protective when instituted within the first 36 h of acute lung injury. In the American trial, the differences in low stretch vs higher stretch between the two groups was greater because the low tidal volume group was ventilated with 6 mL/kg while the control group was ventilated with 12 mL/kg. Furthermore, a much larger number of patients were studied. Also, careful monitoring of the conduct of the trial was carried out. For example, preliminary evidence indicates that there was approximately a 90% compliance with the protocol-driven ventilatory strategy, which included a protocol for weaning and extubation.
An alternate ventilatory strategy was proposed by Amato and colleagues31 who recommended that a combination of pressure control, inverse ratio, low tidal volume, and higher PEEP levels, set according to the inflection point on a pressure-volume curve, could reduce morbidity and mortality in patients with ARDS. Their study was randomized, but included only 56 patients. Furthermore, there was major concern about several of the details of the study, including an extremely high mortality rate in the control group (71%). Theoretically, the potential advantage of this approach rests on increasing PEEP to a level that recruits more alveolar units, resulting in less injury from inflation and deflation of distal lung units during ventilation. Further studies will be needed to evaluate the potential value of higher levels of PEEP in the presence of a low tidal volume (6 mL/kg).
Another novel approach to ventilation of ARDS patients is partial liquid ventilation with perfluorocarbon. Since oxygen has a high solubility in perfluorocarbon, there may be improvement in gas exchange because of recruitment and preservation of functional residual capacity of the lung, allowing lower levels of inspired oxygen; there may also be some benefit from a washout of inflammatory mediators in the distal airspace of the lung. However, there were no major benefits of partial liquid ventilation in a phase II study of patients with ARDS. Further studies are in progress.
Another approach to improving oxygenation in patients with ARDS has been the use of prone position. Although the exact mechanisms that explain the improvement in oxygenation are unclear, periodic changes in body position and particularly the prone position can improve oxygenation, perhaps by reinflating atelectatic areas of the lung in ARDS. Not all patients respond to the prone position and more information is needed about the practical issues of using the prone position, including whether the improvements in oxygenation are sustained, and if the same improvements could be achieved with use of more frequent rotation of the patient to the left and right lateral positions. Several studies are in progress in the United States and Europe.
A number of pharmacologic strategies have been evaluated in clinical trials in the last 5 years for their potential benefit in patients with ARDS.22 Most of the treatments have been delivered early in the course of lung injury, with the expectation that early treatment could favorably modify the course of lung injury.
At the last conference, there was a brief overview of the clinical
trials that showed no benefit of high-dose glucocorticoid in ARDS. At
this year's conference, Dr. Seamas Donnelly presented new evidence
that may explain why glucocorticoids fail in early ARDS. Macrophage
inhibitor factor (MIF), has been shown to override
glucocorticoid-mediated inhibition of cytokine secretion;
interestingly, MIF is present in the BAL of ARDS patients and MIF
enhanced both TNF- and IL-8 secretion from alveolar macrophages from
ARDS patients. Thus, MIF may act as a mediator that promotes and
sustains the pulmonary inflammatory response in ARDS, and thus may
explain, in part, why glucocorticoids have been ineffective in the
acute phase of ARDS.
At the 1993 Aspen Lung Conference, there was considerable discussion of the potential value of inhaled nitric oxide (NO) as a treatment for patients with ARDS. Preliminary data had indicated that inhaled NO could improve oxygenation without causing systemic hypotension. Although there was concern about potential deleterious effects of inhaled NO from the generation of potentially damaging oxidant molecules, two prospective American and European trials have been carried out. The American phase II trial of inhaled NO was disappointing.32 The benefits of inhaled NO on arterial oxygenation were minimal and there was neither significant trend for an improvement in the duration of positive pressure ventilation nor a trend for a survival benefit. In addition, a phase III French study of inhaled NO showed no benefit on morbidity or mortality in patients with ARDS.
Several anti-inflammatory strategies have been evaluated. For many years, a leading hypothesis to explain acute lung injury has been the potential injurious effect of oxygen radicals on the lung. Recently a phase III American trial of procysteine, an analogue of N-acetylcysteine, was completed. N-acetylcysteine is a thiol-containing compound that acts as an oxygen-free radical scavenger and is a precursor for glutathione, an essential element of the antioxidant defense mechanism. N-acetylcysteine has been effective in preventing progressive liver injury from acetaminophen overdose, and the drug has minimal side effects. Unfortunately, a recently completed prospective phase III randomized placebo-controlled trial showed no benefit of treatment with procysteine. The results of the trial have not yet been published. Other pharmacologic agents that have shown no efficacy include prostaglandin E1, delivered with or without liposomes, ketoconazole, and ibuprofen (although this drug was primarily given for patients with sepsis). A variety of other anticytokine therapies have also been ineffective in reducing mortality in patients with sepsis.33
A large phase III trial was carried out to evaluate the potential benefits of aerosolized surfactant therapy (Exosurf; Glaxo Wellcome; Research Triangle Park, NC) in patients with ARDS from sepsis. Unfortunately, in this trial of > 700 patients, there was no beneficial effect of aerosolized surfactant in reducing morbidity or mortality from ARDS.34 However, the use of an aerosol strategy to deliver a surfactant preparation may not have been effective in delivering much more than 5% of the aerosolized dose of surfactant to the distal airspaces of the lung. The recent availability of recombinant SP-B and SP-C preparations makes it possible to consider testing these preparations, perhaps with direct instillation into the lung.
Although pharmacologic treatment of ARDS patients with several at-risk causes (sepsis, aspiration, pneumonia, trauma) has not yet been successful, encouraging results of a clinical trial of complement inhibition in patients undergoing lung transplantation were presented by Dr. Martin Zamora. Short-term complement inhibition with a soluble complement receptor 1 inhibitor (TP10) significantly reduced reperfusion lung injury and led to earlier extubation following lung transplantation.
Although no pharmacologic strategies have been proven beneficial in patients with early ARDS, anecdotal information from observational studies suggested that high doses of glucocorticoids given in patients with moderately severe, established ARDS, approximately 1 week after the diagnosis of ARDS, might improve outcome. At this year's conference, Meduri et al35 reported that high-dose glucocorticoids resulted in a reduction in mortality in a placebo-controlled trial of 24 patients. Unfortunately, 18 of the 24 patients received glucocorticoids since the randomization schedule was 2:1 for placebo to control. Furthermore, of the eight control patients, four crossed over to receive glucocorticoids. Therefore, interpretation of the study is made difficult by the limited number of patients and a crossover design with only four patients in the control group who did not receive glucocorticoid treatment. However, as mentioned above, the NIH 10 center ARDS Network currently is evaluating the same glucocorticoid strategy proposed by Meduri in a prospective phase III trial that has enrolled > 50 patients. In the NIH trial, there is a 1:1 randomization schedule, and the primary end point is hospital mortality. The results should be available by 2001.
Future clinical trials will evaluate the potential utility of anti-inflammatory strategies directed at proximal cytokines that recruit neutrophils to the airspaces of the lung. One potential target is IL-8, the most important chemotactic agent for neutrophils in the airspace of the lung. There are two potential drawbacks to this approach. First, anti-IL-8 therapy might increase the risk of infection, and also might blunt protective aspects of the host inflammatory response. Also, it may be difficult to give the therapy early enough in the patient's clinical course to have a therapeutic benefit. Another potential strategy would be to administer an anti-inflammatory cytokine, such as IL-10, an approach that is currently being evaluated in a phase II study.
An alternative approach would be to provide a treatment strategy designed to enhance the resolution of acute lung injury. For example, it might be possible to administer a recombinant surfactant therapy (SP-B or SP-C) directly into the distal airspaces of the lung. However, there are still potential problems with delivery of a surfactant preparation into the alveoli and potential confounding problems with the need to administer large volumes of fluid to deliver the surfactant by bronchoscopy or BAL. Another strategy would be to deliver ß2-adrenergic agonist therapy, an approach that might hasten the resolution of the alveolar edema and might also increase surfactant secretion from alveolar type II cells. A third strategy for future trials would be to administer KGF, an epithelial-specific mitogen. KGF has been shown to effectively prevent the development of acute lung injury in several experimental studies of acute lung injury. In view of the positive results of the American low tidal volume trial, other lung protective strategies may be evaluated, including higher levels of PEEP or partial liquid ventilation.
Conclusion
In summary, the 1998 Thomas L. Petty Acute Lung Injury Conference in Aspen, CO, was a great success. Basic and clinical investigators presented a broad range of new insights into the mechanisms responsible for acute lung injury, the pathophysiology that sustains lung injury, and potential opportunities for new therapeutic strategies. There has been significant progress in identifying potentially useful biological markers for both pathogenesis and prognosis, and major new insights have been obtained regarding the basic cellular and molecular mechanisms responsible for acute lung injury. Despite the progress, however, there is still much to be learned about the exact sequence of events that lead to lung endothelial and epithelial injury. It is clear that injury to alveolar epithelial barrier is a major determinant of outcome in ARDS, so that differential effects of injurious agents on both barriers are important to investigate. In addition, more research is needed to understand the fundamental mechanisms that regulate resolution of lung injury ranging from the reabsorption of alveolar edema fluid to the remodeling of the lung that occurs following fibrosing alveolitis. The new evidence that low tidal volume ventilation reduces mortality in ARDS constitutes the first major breakthrough in the treatment of ARDS since it was described by Dr. Petty and his colleagues 32 years ago.1 There is also some evidence that standard supportive therapy may have reduced mortality over the last 10 years.13 Finally, the establishment by the Lung Division of the NIH of an ARDS Network at 10 American university centers has directly led to the first breakthrough in a proven new treatment that reduces mortality in ARDS.
Footnotes
Supported in part by NIH HL51854 and HL51856.
References
This article has been cited by other articles:
|
|
 |
|
  M. Licker, M. de Perrot, A. Spiliopoulos, J. Robert, J. Diaper, C. Chevalley, and J.-M. Tschopp Risk Factors for Acute Lung Injury After Thoracic Surgery for Lung Cancer Anesth. Analg., December 1, 2003; 97(6): 1558 - 1565. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. E. Marik Pharmacologic Strategies for the Treatment of Acute Respiratory Distress Syndrome: The Horizon is Getting Closer J Intensive Care Med, November 1, 2002; 17(6): 326 - 328. [PDF]
|
|
|
|
 |
|
 R. G. Brower, L. B. Ware, Y. Berthiaume, and M. A. Matthay Treatment of ARDS Chest, October 1, 2001; 120(4): 1347 - 1367. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. B. WARE and M. A. MATTHAY Alveolar Fluid Clearance Is Impaired in the Majority of Patients with Acute Lung Injury and the Acute Respiratory Distress Syndrome Am. J. Respir. Crit. Care Med., May 1, 2001; 163(6): 1376 - 1383. [Abstract] [Full Text]
|
|
|
|
|
|
T. GEISER, K. ATABAI, P.-H. JARREAU, L. B. WARE, J. PUGIN, and M. A. MATTHAY Pulmonary Edema Fluid from Patients with Acute Lung Injury Augments In Vitro Alveolar Epithelial Repair by an IL-1{beta}-dependent Mechanism Am. J. Respir. Crit. Care Med., May 1, 2001; 163(6): 1384 - 1388. [Abstract] [Full Text]
|
|
|
|
 |
|
 L. B. Ware and M. A. Matthay The Acute Respiratory Distress Syndrome N. Engl. J. Med., May 4, 2000; 342(18): 1334 - 1349. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
