Respiratory Failure Part II: Acute Respiratory Distress Syndrome
Gilman B. Allen
Polly E. Parsons
Introduction
Acute lung injury (ALI) and the acute respiratory distress syndrome (ARDS) represent a continuum of severity for the same pathologic condition, both being defined by noncardiogenic pulmonary edema and hypoxemia in the setting of direct or indirect lung injury. Because ARDS, by definition, simply represents a more severely advanced form of ALI, the term “ALI” can be used as a comprehensive term for both conditions. ALI represents a common pathologic endpoint of various potential insults to the lung that almost invariably lead to hypoxemic respiratory failure requiring support with mechanical ventilation. Despite the confirmed success of protective mechanical ventilation strategies in lowering mortality [1,2] and ongoing efforts to discover other effective interventions [3,4,5,6], treatment of this condition remains largely supportive, and ALI continues to be a major source of morbidity and mortality in the intensive care unit [7,8]. Fortunately, an enormous body of research already exists on the pathogenesis of this condition, and advances continue to develop with regard to our understanding of ALI, its prognostic implications, and how to best manage the condition medically.
Definition
ALI is defined as a diminished arterial oxygen pressure (PaO2) to fractional inspired oxygen (FiO2) ratio (P to F (P:F) ratio less than 300), bilateral airspace disease on chest radiograph, and pulmonary edema from increased permeability, the latter defined by evidence of normal cardiac function [9]. ARDS is simply a subset of ALI having a more severely diminished P:F ratio (less than 200). However, because the P:F ratio can be affected by arbitrary ventilator settings [10], and because many
studies have shown that indices of oxygenation are not strongly predictive of outcome [11,12,13], this differentiation may be of limited clinical relevance. Furthermore, the definition of ALI and ARDS has undergone significant evolution over the years, and limitations of this definition still exist [14], which can confound the interpretation of older research results and contribute added challenges to the design of new studies.
studies have shown that indices of oxygenation are not strongly predictive of outcome [11,12,13], this differentiation may be of limited clinical relevance. Furthermore, the definition of ALI and ARDS has undergone significant evolution over the years, and limitations of this definition still exist [14], which can confound the interpretation of older research results and contribute added challenges to the design of new studies.
Table 47.1 Recommended Criteria for Acute Lung Injury (ALI) and Acute Respiratory Distress Syndrome (ARDS) | ||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||
In response to the recognized limitations in determining the incidence and outcomes of ALI, a committee of leading investigators in the field met in 1994 to develop a consensus between the American Thoracic Society and the European Society of Intensive Care Medicine. The most current definition of ALI derives from this consensus [9] and defines the condition as the acute onset of hypoxemia and noncardiogenic pulmonary edema (see Table 47.1). Although the source of hypoxemia in ALI is multifactorial, it is one of the most easily gauged markers of “lung injury” in the intensive care unit and thus an important component of the definition. Despite its limited prognostic value, the more inclusive P:F ratio of less than 300 can serve to identify patients earlier in their course [11], thus expediting delivery of critical life saving interventions before progression to ARDS. In ALI, the pulmonary edema is the result of capillary leak, a parameter that is difficult to measure in the clinical setting. Accordingly, noncardiogenic pulmonary edema is defined using clinical parameters, which include the presence of “bilateral infiltrates” consistent with pulmonary edema on chest radiograph and either a pulmonary artery wedge pressure (PAWP) less than 18 mm Hg (when measured) or no clinical evidence of left atrial hypertension [9]. However, because the group recognized that ALI does not always exist exclusively without heart failure, the consensus more explicitly defines ALI as “a syndrome of inflammation and increased permeability that is associated with a constellation of clinical, radiologic, and physiologic abnormalities that cannot be explained by, but may coexist with, left atrial or pulmonary capillary hypertension”[9].
Despite the great lengths taken to clarify the current definition of ALI, it is not without its shortcomings, particularly because it does not delineate the cause of hypoxemia (i.e., alveolar damage) or clearly establish the presence of increased permeability [14]. Unfortunately, easily employed tests for microvascular permeability are not yet available, and what degree of permeability is needed to reliably predict the presence of alveolar damage is not known [14]. The boundaries for the P:F ratio are also arbitrary. The consensus committee recognized the difficulty in interpreting this ratio in the setting of different levels of positive end-expiratory pressure (PEEP) [15], and thus decided to not include this parameter in their definition. It would also be impractical to base the clinical definition of ALI upon histologic findings given the often critical condition of patients and their poor candidacy for biopsy by the time of clinical diagnosis. Nevertheless, the histopathology of ALI has been well characterized and is, in many ways, descriptive of its pathogenesis.
Histopathology
Despite having many different potential etiologies [16,17,18], the histologic findings of ALI are fundamentally uniform and are collectively described by the term, diffuse alveolar damage (DAD) [19]. DAD represents a continuum of changes that can be temporally divided into exudative, proliferative, and fibrotic phases [19,20], between which considerable overlap exists. The exudative phase of DAD is the earliest phase, during which clinical symptoms first develop and lung mechanical changes become manifest [21]. This phase typically occupies the first week and is characterized by epithelial and endothelial cell death, neutrophil sequestration, platelet–fibrin thrombi, interstitial edema, and exudates within the airspaces, which consist of fluid, protein, and cellular debris [19]. These exudates compact into dense, protein-rich hyaline membranes that stain strongly with eosin and line the alveoli and alveolar ducts (Fig. 47.1A). During the second week of injury, the proliferative phase ensues, which is characterized by organization of the intra-alveolar exudates and proliferation of type II alveolar cells, fibroblasts, and myofibroblasts. During this phase, it is common to find areas of squamous metaplasia and granulation tissue occluding alveolar ducts in a manner similar to that of organizing pneumonia (Fig. 47.1B) [22].
The fibrotic phase has classically been considered the later phase of remodeling that occurs in patients who survive past 3 or 4 weeks [19]. However, studies suggest an increase in the fibrotic response to ALI as early as 24 hours from presentation [23], and histologic evidence can be seen within the first 2 weeks of diagnosis [24]. Because such overlap exists between the fibrotic and proliferative phases, the two are often described together as the fibroproliferative phase. On histology, alveolar septa are expanded and airspaces filled with sparsely cellular connective tissue [19]. Such airspace connective tissue formation can either resolve or progress to the point of complete airspace obliteration [24], fibrosis, and even honeycombing [22]. Regardless of severity, there is evidence that increased fibroproliferative signaling [23] and fibrosis [24] predict worse outcomes.
Radiographic Findings
The diagnostic criteria of ALI require bilateral infiltrates on frontal chest radiograph [9]. These infiltrates will often initially appear as heterogeneous opacities, but later become more homogenous over hours to days [25] (see Fig. 47.2A). Although some have recommended using criteria such as cardiac
silhouette size and vascular pedicle width to differentiate cardiogenic from noncardiogenic edema, this differentiation has proven difficult [26]. Furthermore, the seemingly straightforward interpretation of bilateral infiltrates can be obscured by factors such as atelectasis, effusions, or isolated lower lobe involvement, all of which contribute to low interobserver agreement [27].
silhouette size and vascular pedicle width to differentiate cardiogenic from noncardiogenic edema, this differentiation has proven difficult [26]. Furthermore, the seemingly straightforward interpretation of bilateral infiltrates can be obscured by factors such as atelectasis, effusions, or isolated lower lobe involvement, all of which contribute to low interobserver agreement [27].
 Figure 47.1. A: Histologic lung specimen from ARDS patient, showing red blood cells and neutrophils within the alveolar space and characteristic hyaline membranes (arrow) consistent with diagnosis of diffuse alveolar damage (DAD). B: Hematoxylin and eosin stained, 60×; demonstrates distal airspace granulation tissue (asterisks) consistent with organizing pneumonia. [Images were graciously provided by Dr. Martha Warnock.] |
 Figure 47.2. A plain chest radiograph from a patient with ARDS [generously provided by Dr. Jeff Klein]. B, C: Computed tomography images of the chest from patients with ARDS [Images reproduced with permission from Goodman LR, Fumagalli R, Tagliabue P, et al: Adult respiratory distress syndrome due to pulmonary and extrapulmonary causes: CT, clinical, and functional correlations. Radiology 213:545–552, 1999.]. B: Diffuse patchy regions of consolidation with a predominance of ground glass infiltrates and small effusion (arrow). C: A predominance dense consolidation (arrow), particularly at the bases, with sparse areas of ground glass. |
Prior to computed tomography (CT) scanning, the pulmonary edema seen on chest radiograph was widely believed to be a diffuse process. However, CT imaging has demonstrated
the distribution of ALI to oftentimes be heterogeneous and patchy, with areas of normal-appearing, aerated lung interspersed among areas of mixed ground glass opacity and consolidation, the latter being concentrated in the more gravitationally dependent regions of the lung [28] (see Fig. 47.2B, C). Despite this pattern, a recent study using positron emission tomography (PET) to map cellular metabolic activity demonstrated that diffuse inflammatory change can be detected even in areas of the lung that appear spared radiographically [29]. Some investigators have also used PET imaging and magnetic resonance imaging (MRI) to estimate pulmonary microvascular leak and assist in the differentiation between high permeability and hydrostatic pulmonary edema [30,31,32], but these methods have yet to be adopted in clinical practice.
the distribution of ALI to oftentimes be heterogeneous and patchy, with areas of normal-appearing, aerated lung interspersed among areas of mixed ground glass opacity and consolidation, the latter being concentrated in the more gravitationally dependent regions of the lung [28] (see Fig. 47.2B, C). Despite this pattern, a recent study using positron emission tomography (PET) to map cellular metabolic activity demonstrated that diffuse inflammatory change can be detected even in areas of the lung that appear spared radiographically [29]. Some investigators have also used PET imaging and magnetic resonance imaging (MRI) to estimate pulmonary microvascular leak and assist in the differentiation between high permeability and hydrostatic pulmonary edema [30,31,32], but these methods have yet to be adopted in clinical practice.
Epidemiology
The estimated incidence of ALI worldwide has been variable in the past due to its wide range of causes and previously nonuniform definition. The first estimate by the National Institutes of Health (NIH) projected an annual incidence of 75 cases per 100,000 in the United States [33]. Two subsequent cohort studies in Scandinavia and Australia, respectively, estimated an annual incidence of 18 and 34 cases per 100,000 [34,35], but these studies were limited in size and case inclusion. A much lager pool of prospective cases from the NHLBI-sponsored ARDS Network yielded a conservative estimate of 64.2 cases per 100,000 person-years [36]. A more recent and significantly larger prospective cohort study from King County in Washington State estimates an annual incidence of 78.9 cases per 100,000 person-years [7], which is more in accordance with the ARDS Network and original NIH estimates, and is likely to be the most accurate estimate to date for incidence in the United States.
In patients at risk of developing ALI, the onset of ALI is typically swift, with a median duration of 1 day (interquartile range 0 to 4 days) from the time of risk factor development to the time of diagnosis [37]. The known causes and risk factors for the development of ALI have been well characterized [16,17,18] (see Table 47.2), and can be categorized as ensuing from either direct or indirect injury to the lung [16,38]. This differentiation is justified by the demonstration of differing physiologic properties between ALI of a direct or indirect nature [38], and by the varied outcomes associated with different causes of ALI [7,11,13]. It is now well established that sepsis is the most commonly identified cause of ALI, and is associated with the worst outcome overall [7,13,18], while trauma-related ALI has a significantly lower mortality [7]. These differences in mortality may be in part due to differences in pathogenesis [39]. Other risk factors for the development of ALI following a known insult include a history of alcoholism [40,41,42], recent chemotherapy [41], delayed resuscitation [41], and transfusion with blood products [43,44,45,46]. The latter condition, commonly referred to as “transfusion-related ALI” (i.e., TRALI), may be more likely to develop following transfusion with fresh frozen plasma and platelets than with packed red blood cells (PRBCs) [44]. Curiously, in those at clinical risk for developing ALI, the diagnosis of diabetes mellitus has been shown to confer protection from ALI, providing about half the relative risk as that of nondiabetic patients [41,47].
Table 47.2 Clinical Disorders Associated with the Development of Ali and ARDS, Subcategorized Into Those Commonly Associated with Direct and Indirect Injury to the Lung | ||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||
Pathogenesis
An understanding of the pathogenesis of ALI is perhaps best imparted through a reflection on the predominant pathologic findings on histology. First and foremost, ALI is a condition triggered by injury to the alveolar epithelium and capillary endothelium. The insult can be initially isolated to either the epithelium, as in the case of aspiration, or to the endothelium, as in most forms of indirect ALI such as sepsis. However, injury is generally detected in both the endothelium and epithelium by the time of diagnosis [19,48]. This injury invariably leads to a leakage of plasma proteins into the alveolar space. Many of these plasma proteins in turn activate procoagulant and proinflammatory pathways that lead to the fibrinous and purulent exudates seen on histology. Through increased transcription and release of proinflammatory cytokines, and an increased expression of cell surface adhesion molecules, a profound acute inflammatory response ensues. This is heralded by epithelial cell apoptosis and necrosis [49], further activation of inflammatory cascades, and a robust recruitment of neutrophils [50]. The increased expression of tissue factor and other procoagulant factors ultimately leads to coagulation within the microvasculature and airspaces, accompanied by a suppression of fibrinolysis, which helps perpetuate the microthrombi and fibrinous exudates that are pathognomonic for ALI.
Injury to the alveolar epithelium plays a critical role in the pathogenesis of ALI. Through the loss of tight junctions and barrier function, plasma proteins and edema fluid seep into the alveolar space, leading to increased shunt fraction, higher alveolar surface tension, and a greater propensity for alveolar collapse. Clearance of both protein and fluid are crucial to the resolution of ALI. Indeed, a greater alveolar fluid clearance (AFC) rate is associated with fewer days of mechanical ventilation and lower mortality in patients with ALI [51]. The type I alveolar epithelial cell (pneumocyte) plays an important role in barrier function, while the type II pneumocyte is the primary source of surfactant production and is known to participate in AFC. Although type I pneumocytes comprise 99% of the alveolar surface area and are presumed to participate in AFC, their exact role in this process remains undefined [52]. AFC occurs by fluid following a sodium concentration gradient established by active sodium transport at the basolateral membrane via Na, K-ATPase activity [53]. Despite the demonstrated impairment of AFC in the setting of lung injury [54], areas of preserved AFC can coexist with injury and epithelial barrier disruption [55], making AFC a potential target for interventional therapy (see “Management” section).
The resorption of protein from the alveolar space is believed to occur more slowly than AFC, and is differentially regulated
depending on the burden of protein present. Alveolar albumin transport occurs primarily via receptor-mediated endocytosis at low concentrations, but occurs primarily via passive paracellular diffusion when present in higher concentrations, as in the case of ALI [56]. Removal of larger insoluble proteins such as fibrin can take much longer and require degradation [56].
depending on the burden of protein present. Alveolar albumin transport occurs primarily via receptor-mediated endocytosis at low concentrations, but occurs primarily via passive paracellular diffusion when present in higher concentrations, as in the case of ALI [56]. Removal of larger insoluble proteins such as fibrin can take much longer and require degradation [56].
On the other side of the alveolar capillary interface, injury to the endothelium results in increased permeability, release of inflammatory molecules, expression of cell adhesion molecules, and activation of procoagulant pathways. Although endothelial injury is detectable under electron microscopy [19], gross endothelial damage may be seen only sparingly [48,57]. Increased microvascular permeability has been widely demonstrated in ALI [32,58,59], but this may be more due to a functional alteration or activation of intact endothelium than due to actual cell lysis or necrosis. Endothelial cells can be activated by factors such as thrombin or endotoxin to increase surface expression of the potent neutrophil-tethering molecules called selectins [60] or to release preformed von Willebrand factor (vWF) [61] and potent neutrophil activating factors [62]. Endothelial cell activation of binding molecules on neutrophils can in turn promote their binding to the endothelium and transmigration into areas of injury. Furthermore, when endothelial cells are tethered to activated neutrophils, such interaction can promote neutrophil degranulation [63], further contributing to local injury and inflammation. The important role of endothelial activation in ALI is highlighted by the finding that elevated plasma levels of vWF have been shown to predict the development of ALI in patients at risk [64] and are associated with worse outcomes [65] and fewer organ failure-free days in established ALI [65].
Although widely accepted to play a key role in the pathogenesis of ALI [50,66,67], the neutrophil is not essential for the development of ALI, as evidenced by the development of ALI in the setting of neutropenia [68]. However, ALI can worsen during the recovery from neutropenia and after administration of the neutrophil growth and releasing factor, G-CSF [69]. Furthermore, neutrophil recruitment to the lung has been shown to be a crucial factor in experimentally induced ALI as demonstrated by attenuated pathology under neutrophil-depleted conditions [70,71].
Activated leukocytes and endothelial cells can also contribute to another recognized pathologic manifestation of ALI: dysregulated intravascular and extravascular coagulation [72,73]. Surface expression of tissue factor by alveolar macrophages and endothelial cells can activate the extrinsic coagulation cascade through factor VII [73], activating thrombin and generating fibrin [72]. Extravascular alveolar fibrin arising from increased procoagulant activity and impaired fibrinolysis [74,75] has been well described in ALI [48]. Fibrin formation and clearance in the lung is in part governed by the differential activity of fibrinolysis promoters and inhibitors [74,75,76]. Plasminogen activators enzymatically convert plasminogen to active plasmin, the key protease involved in fibrinolysis. Plasminogen activator inhibitor-1 (PAI-1) can prevent fibrinolysis via direct binding and inhibition of plasminogen activators [77]. PAI-1 inhibition of fibrinolysis in the BAL fluid of ALI patients was first recognized in 1990 [75]. Since then the importance of PAI-1 in ALI has been further recognized in that elevated plasma and edema fluid levels of PAI-1 are associated with higher mortality in ALI patients [78]. However, studies examining the direct role of PAI-1 in animal models of ALI have yielded mixed results [79,80]. With respect to the initial process of coagulation and fibrin generation, the activation and expression of tissue factor (TF) has received notable attention due to its known interaction with factor VIIa and downstream generation of thrombin. TF expression has been shown to be increased on the surface of alveolar epithelial cells and macrophages in patients with ALI, and is accompanied by increased procoagulant activity in the edema fluid [81].
Numerous additional pathways have been implicated in the pathogenesis of ALI, but an attempt to cover each in depth would extend beyond the intended breadth of this chapter. In brief, lipopolysaccharide (i.e., endotoxin) has long been recognized as a reliable initiator of ALI [82], particularly in the settings of sepsis and pneumonia, and the mechanisms of its action have been extensively elaborated [83]. Oxidant-mediated injury through the generation of reactive oxidant species is also a well-recognized pathway for injury in ALI [84]. The cytoprotective role of the heat-shock response in ALI, particularly through heat shock protein 70, is also widely acknowledged [85,86]. Dysregulated cell death and apoptosis through the release and accumulation of soluble Fas ligand is also thought to contribute to ALI and may also become a potential future target for therapeutic intervention [49,87]. The role of mechanical ventilation in contributing to the development and worsening of ALI is now also widely recognized and its mechanisms extensively researched [88,89].
Pathophysiology
Because of the accumulation of extravascular lung water (i.e., pulmonary edema), the physiologic derangements of ALI invariably manifest as refractory hypoxemia [90], decreased respiratory compliance [91], and a propensity for alveolar closure [92]. As alveolar edema fluid and protein accumulate within the alveoli, physiologic shunt develops as blood flows through capillary units and perfuses alveoli that are either filled with fluid, or have collapsed from the resulting increase in surface tension (see Fig. 47.3A). Hypoxic vasoconstriction, the normal autoregulatory reflex that helps match ventilation and perfusion by shunting capillary blood flow away from poorly ventilated regions of the lung, is severely impaired within the diseased regions of the lung [93]. Hence, physiologic shunt is accentuated by an imbalance of flow to the poorly ventilated lung regions [93]. Increased vasoconstriction and scattered microthrombi within well-ventilated lung regions contribute to physiologic dead space or “wasted ventilation” via diminished blood flow to aerated lung [93] (see Fig. 47.3B). The combined effects of these derangements result in refractory hypoxemia and increased minute ventilation requirements, which explain the often challenging demands of managing these patients in the intensive care unit.
Overall, the average pulmonary vascular resistance is commonly elevated in patients with ALI [94,95], likely the result of a reduction in total luminal diameter of the vascular bed, stemming from hypoxia and thrombotic obstruction [95,96]. This in turn leads to the common finding of pulmonary
hypertension in these patients, which can alter right ventricular loading and function [94,97], and predicts higher mortality in afflicted patients [97]. Because elevated pulmonary artery pressures could in theory contribute to increased pulmonary edema [94,98] and right heart strain, it is unclear whether pulmonary hypertension is directly contributing to mortality or simply a marker of disease severity [95].
hypertension in these patients, which can alter right ventricular loading and function [94,97], and predicts higher mortality in afflicted patients [97]. Because elevated pulmonary artery pressures could in theory contribute to increased pulmonary edema [94,98] and right heart strain, it is unclear whether pulmonary hypertension is directly contributing to mortality or simply a marker of disease severity [95].
 Figure 47.3. A: The edema fluid-filled alveolus and a neighboring collapsed alveolus, both with unrestricted blood flow, contributing to physiologic shunt. Double-headed (arrow) represents potential for fluid-filled alveolus to collapse and re-expand during normal tidal ventilation. B: The effect of a microthrombus (black oval) obstructing blood flow to a functioning alveolus, contributing to physiologic dead space. |
The mechanical manifestations of ALI present mainly as a decrease in respiratory compliance. This is primarily due to a decrease in lung compliance, particularly in the more direct forms of ALI such as pneumonia. However, contribution from the chest wall and abdominal compartment can be significant under conditions such as trauma and peritonitis [38]. The reduction in lung compliance reflects the collective contribution of changes in the intrinsic elastic properties of the remaining aerated lung and a reduction in resting lung volume via alveolar flooding and collapse. The increased elastic properties of the aerated lung result from increased tissue stiffness due to interstitial edema and increased alveolar surface tension, but the contribution from interstitial edema is thought to be negligible relative to that from alveolar edema [99]. The increase in alveolar surface tension is thought to develop from the increased surface forces generated by a greater abundance of alveolar lining fluid and a decrease in surfactant activity [100]. This loss in surfactant activity is believed to result from inhibitory binding of surfactant by plasma proteins [101] and cholesterol [102], and decreased production of functionally active surfactant by type II pneumocytes [3,103]. To further complicate matters, the biomechanical effects of mechanical ventilation alone can alter the structure and biophysical properties of surfactant [104,105], an unfortunate consequence of a typically mandatory intervention for this condition.
Lower resting lung volumes in ALI result from persistently fluid filled or collapsed alveoli, leading to what has been colloquially referred to as “baby lung” [106]. The affected regions of the lungs are often so diseased that they may remain fluid-filled or collapsed throughout each tidal inflation [107] and hence contribute negligibly to compliance. In fact, CT imaging has demonstrated respiratory compliance to be more closely linked to the amount of aerated lung [108], lending some to assert that compliance is more of a direct measure of aerated lung volume than tissue stiffness [106]. As a result, tidal volumes delivered to the heterogeneously fluid-filled and atelectatic lung are shunted preferentially to more compliant, aerated regions of the lung [109]. This is one of the main postulated mechanisms through which mechanical ventilation can overdistend and injure the remaining regions of “normal lung” and lead to ventilator-induced lung injury (VILI) [88].
At the bedside, the reduction in compliance is typically observed as an increase in peak and plateau airway pressures but can also be seen as an expansion in the hysteresis of pressure–volume (PV) curves obtained during graded inflation of the lung (see Fig. 47.4). The decrease in slope of the inspiratory limb of the PV curve represents a decrease in volume obtained for any given change in pressure, and hence a decrease in compliance.
 Figure 47.4. Simulated pressure volume curve obtained from typical acute lung injury patient, with pressure recorded during slow inflation to total lung volume. Lower inflection point (LIP) marked at point of sudden change in slope of inflation curve. Point of maximal curvature (PMC) also marked at point of maximal change in slope of deflation curve. |
Management
Mechanical Ventilation
Mechanical Ventilation and Low Tidal Volumes
The early presentation of ALI is chiefly characterized by hypoxemic respiratory failure and the almost invariable need for support with mechanical ventilation. Because the greatest danger posed to patients with ALI is the development of multiorgan failure [110], establishing supportive ventilation modes that optimize hemodynamic function and oxygen delivery remain important objectives in the management of these patients. Prior to the late 1960s, endotracheal intubation and positive pressure mechanical ventilation were primarily used for supporting patients during general anesthesia. It was during this time that investigators first noted that larger tidal volumes could reduce the shunt associated with atelectasis during general anesthesia [111]. Soon afterward, the benefits of a larger tidal volume on shunt were demonstrated in animal models of ALI [112]. Because many of the techniques used for the support of patients with acute respiratory failure were originally adopted from general anesthesia practice, employing tidal volumes of 10 to 15 mg per kg became the standard for improving oxygenation and ventilation in patients with ALI [113,114].
We now know that idealized oxygenation and normal physiologic pH and PaCO2 can come at a cost when employing higher tidal volumes in patients with ALI. After VILI was induced with higher tidal volumes in animal models [88,115], small retrospective and prospective uncontrolled trials suggested a benefit from limiting tidal volume and peak airway pressures in patients with ALI [116,117]. Numerous larger, randomized trials comparing traditional and lower tidal volumes have since been conducted, each trial differing in its methodology and results [1,2,118,119,120]. The largest randomized, multicenter trial to date, conducted by the ARDS Network, ultimately demonstrated a significant reduction in mortality when using a tidal volume of 6 mL per kg of predicted ideal body weight and a target plateau pressure of 30 cm H2O or less (mortality 31.0%) as opposed to a tidal volume of 12 mL per kg and a target plateau pressure less than 50 cm H2O (mortality 39.8%) [1].
In an effort to better understand the protection conferred by low tidal volumes, investigators have studied how this strategy modulates the inflammatory cascades associated with ALI and VILI. Evidence now exists to support the theory that low tidal volume ventilation improves outcomes at least in part through reduced activation of the inflammatory cascades associated with VILI and multiorgan failure. For instance, among patients enrolled in the ARDS Network trial of low tidal volume, it was found that higher plasma levels of soluble receptors for tumor necrosis factor-α (TNF-α) were associated with higher mortality and fewer organ-failure free days [121]. Furthermore, the lower tidal volume strategy was associated with lower levels of soluble TNF-α receptor I [121]. In another study from the same patient population, elevated plasma levels of interleukin (IL)-6, 8, and 10 were also linked to increased mortality while lower tidal volume was associated with a greater drop in IL-6 and IL-8 by day 3 of enrollment [122].
Many studies of low tidal volume ventilation adopted a strategy of permissive hypercapnia, in which investigators
tolerated a reduction in minute volume and an ensuing increase in PaCO2 to achieve lower target tidal volumes and airway pressures [117,118,120]. Most studies suggest that this strategy is safe [117,118], but the actual safety of this practice is not yet entirely known. Although some animal studies have demonstrated a potential protection by hypercapnic acidosis [123,124], others suggest that hypercapnic acidosis may worsen ALI and VILI [125,126]. Some guidelines acknowledge permissive hypercapnia as an acceptable practice when necessary to limit tidal volumes, but also stress that its use is limited in patients with preexistent metabolic acidosis, and contraindicated in patients with increased intracranial pressure [127]. Because no firm guidelines have been established, current options range from a allowing for an arterial pH as low as 6.8 [117], to increasing respiratory rate up to 35 and buffering with intravenous bicarbonate when pH drops below 7.3 [1]. Despite ongoing controversy [128] and the delayed adoption low tidal volume strategy in clinical practice [129,130], the current evidence has led professional societies to recommend the use of lower tidal volumes at goal plateau pressures less than 30 cm H2O in patients with established ALI [127]. Because calculations based on total body weight may be partly responsible for the documented underuse of lower tidal volumes for patients with ALI [129], the importance of using predicted ideal body weight (IBW), based upon measured height and sex, cannot be overstressed. IBW (in kg) for males is calculated as 50 + 0.91 ((height in cm)—152.4), and for females as 45.5 + 0.91 ((height in cm)—152.4) [1]. Although no firm guidelines exist regarding patients without established ALI, there is clinical evidence that a low tidal volume strategy may help prevent progression to ALI in patients at risk [131,132]. Yet to be determined is whether a more optimal or “best” strategy exists beyond that employed in the ARDS Network sponsored study. Although data suggest that tidal volumes lower than 6 mL per kg may confer even greater protection from VILI [133], there is no general consensus on this practice. However, the authors note that in the original ARDS Network trial, the lower tidal volume assignment started with a goal of 6 mL per kg, but patients in this arm were oftentimes adjusted to as low as 4 mL per kg as needed to maintain plateau pressures less than 30 cm H2O [1].
tolerated a reduction in minute volume and an ensuing increase in PaCO2 to achieve lower target tidal volumes and airway pressures [117,118,120]. Most studies suggest that this strategy is safe [117,118], but the actual safety of this practice is not yet entirely known. Although some animal studies have demonstrated a potential protection by hypercapnic acidosis [123,124], others suggest that hypercapnic acidosis may worsen ALI and VILI [125,126]. Some guidelines acknowledge permissive hypercapnia as an acceptable practice when necessary to limit tidal volumes, but also stress that its use is limited in patients with preexistent metabolic acidosis, and contraindicated in patients with increased intracranial pressure [127]. Because no firm guidelines have been established, current options range from a allowing for an arterial pH as low as 6.8 [117], to increasing respiratory rate up to 35 and buffering with intravenous bicarbonate when pH drops below 7.3 [1]. Despite ongoing controversy [128] and the delayed adoption low tidal volume strategy in clinical practice [129,130], the current evidence has led professional societies to recommend the use of lower tidal volumes at goal plateau pressures less than 30 cm H2O in patients with established ALI [127]. Because calculations based on total body weight may be partly responsible for the documented underuse of lower tidal volumes for patients with ALI [129], the importance of using predicted ideal body weight (IBW), based upon measured height and sex, cannot be overstressed. IBW (in kg) for males is calculated as 50 + 0.91 ((height in cm)—152.4), and for females as 45.5 + 0.91 ((height in cm)—152.4) [1]. Although no firm guidelines exist regarding patients without established ALI, there is clinical evidence that a low tidal volume strategy may help prevent progression to ALI in patients at risk [131,132]. Yet to be determined is whether a more optimal or “best” strategy exists beyond that employed in the ARDS Network sponsored study. Although data suggest that tidal volumes lower than 6 mL per kg may confer even greater protection from VILI [133], there is no general consensus on this practice. However, the authors note that in the original ARDS Network trial, the lower tidal volume assignment started with a goal of 6 mL per kg, but patients in this arm were oftentimes adjusted to as low as 4 mL per kg as needed to maintain plateau pressures less than 30 cm H2O [1].
Recruitment
The physiologic abnormalities in ALI can, in some patients, be reversed by a recruitment maneuver (RM), typically delivered as a sustained deep inflation with the intention of reopening collapsed regions of the lung. However, because of the unusually high surface tension within affected alveoli, the benefit is often transient [134,135], especially if not followed by sufficiently high levels of PEEP [136]. The potential impact of RMs on morbidity and mortality is not trivial. In fact, because derecruitment leads to an effectively smaller ventilated lung, investigators have proposed the use of “open lung” strategies [137] with periodic delivery of RMs to limit regional overdistention and minimize injury from atelectasis and cyclic alveolar reexpansion [88]. The long-term effect of atelectasis in humans is unclear, but prolonged periods of atelectasis have been shown to promote vascular leak and right ventricular failure in rodents [138]. On the other hand, periodic RMs also have the potential to worsen oxygenation by shunting blood flow to poorly aerated regions [139] and impair cardiac output by limiting venous return and cardiac preload [140,141]. Furthermore, RMs could conceivably contribute to lung injury through excessive overdistention [142] or repeated opening of collapsed lung.
Despite encouraging findings from animal studies [143,144], clinical studies have yielded mixed results regarding beneficial effects of RMs on oxygenation and lung function [134,141,145]. Although earlier clinical studies demonstrated the benefits of recruitment to be negligible or short-lived [134,140], recent larger trials have demonstrated more promising improvements in lung function and oxygenation but still failed to demonstrate any reduction in mortality [146,147]. Although no guidelines currently exist, it is important to note that patients with ALI of differing origin [38,136,148] and stages of injury [141] vary in their response to RM, and it may help to first differentiate responders from nonresponders [141,148]. When performed, RMs are traditionally delivered as sustained inflations with peak inflation pressures limited to between 30 and 40 cm H2O, and held for a period ranging from 15 to 40 seconds [2,141,144].
Positive End-Expiratory Pressure
PEEP is another widely employed strategy shown to retard alveolar derecruitment in the injured lung. Several studies have demonstrated the ability of PEEP to prevent or delay alveolar derecruitment [149,150] and attenuate VILI [115,151]. However the protective effect of higher PEEP was called into doubt after a multicenter randomized trial failed to demonstrate an improvement in outcomes using a higher PEEP strategy during low tidal volume ventilation in ALI patients [152]. In this NHLBI-sponsored trial, higher levels of PEEP were arbitrarily coupled to each step-wise increment in FiO2 requirement during low tidal volume ventilation [152]. The study failed to demonstrate any benefit in mortality or ventilator-free days with higher PEEP [152], but potential underpowering of this study has left room for continued debate [153]. In addition, since the amount of recruitable lung varies significantly among ALI patients [154], some have suggested that setting PEEP levels without first determining the level of recruitable lung may offset the potential benefits of PEEP. In a recent randomized trial, the selection of PEEP was more patient-directed and set at a level required to maintain plateau pressures of 28 to 30 cm H2O [147]. This higher PEEP strategy again failed to demonstrate a reduction in mortality, but did demonstrate lasting improvements in oxygenation and compliance and an increase in ventilator-free and organ failure-free days [147]. Others have shown that more directly targeting PEEP to transpulmonary pressure by measuring esophageal pressures may be a safer and more effective means of determining optimal PEEP [155].
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
