Clin Chest Med 27 (2006) 579–589
Mechanisms of Acute Lung Injury/Acute Respiratory
Benjamin T. Suratt, MDa,b,*, Polly E. Parsons, MDa,b
Vermont Lung Center, Burlington, VT 05405, USA
Division of Pulmonary and Critical Care, Fletcher Allen Health Care,
Burlington, VT 05405, USA
In the 4 decades that have elapsed since changing emphasis on and controversy over the
Ashbaugh and colleagues  ﬁrst described acute importance of these many mechanisms. Such
lung injury/acute respiratory distress syndrome controversy continues because of the continued
(ALI/ARDS), tremendous strides have been lack of clinical advances despite these many
made in dissecting the pathophysiology of this discoveries. The inability to ﬁnd a unifying theory
disease. This understanding unfortunately has for ARDS pathogenesis almost certainly reﬂects
not translated into similar advances in the ability the complexity of a disease with a common end
to predict, prevent, or treat this disease that af- point but many causes.
fects nearly 200,000 patients a year in the United
States alone . The American-European Consen-
sus Conference (AECC) on ARDS formally de- Common end point: failure of the alveolar
ﬁned ARDS as severe arterial hypoxemia (with capillary membrane
a partial pressure arterial oxygen/fraction of in- Decades of clinical and animal research have
spired oxygen [PaO2/FIO2] ratio of 200 or less) in established increased permeability edema as the
the presence of bilateral alveolar inﬁltrates with- primary physiologic abnormality in the early stages
out evidence of elevated left atrial pressure . of ALI/ARDS. Such edema diﬀers from high-
This deﬁnition recognizes a clinical syndrome, ir- pressure or hydrostatic edema (eg, congestive heart
respective of speciﬁc molecular, immunologic, or failure) in that increased permeability edema is
physical events that may precede it. Thus, the def- driven primarily by a failure of the alveolar struc-
inition reﬂects a common ﬁnal end point in the tures that normally retain plasma within the alveo-
acute phase of this disease: respiratory compro- lar capillaries (the alveolar capillary membrane
mise caused by failure of the alveolar capillary [ACM]). Failure of the ACM in ALI/ARDS allows
barrier and the resulting development of protein- proteinaceous ﬂuid to ﬂood the alveolar airspaces
aceous edema. and contributes directly to the impairment of gas
From the formidable body of research exam- exchange and loss of lung compliance that charac-
ining the advent and perpetuation of this central terize this disorder. The ACM is formed by two
injury, several pathophysiologic themes have separate components: the capillary endothelium and
arisen, and the history of this work reﬂects ever- the alveolar epithelium; the function of both is
disrupted in ALI/ARDS (Fig. 1). The mechanisms
leading to the failure of the ACM are multiple but
This work was supported by Grants No. HL084200
can be divided roughly into those aﬀecting the cap-
and NCRR P20 RR15557 from the NIH.
* Corresponding author. Vermont Lung Center, 149 illary endothelium and those aﬀecting the alveolar
Beaumont Avenue, HSRF 230, Burlington, VT 05405. epithelium.
E-mail address: email@example.com Damage to the alveolar capillary endothelium
(B.T. Suratt). has long been recognized as a key feature of the
0272-5231/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
580 SURATT & PARSONS
Fig. 1. The normal alveolus (left) and the injured alveolus in the acute phase of acute lung injury and the acute respi-
ratory distress syndrome (right). In the acute phase of the syndrome there is sloughing of both the bronchial and alveolar
epithelial cells with the formation of protein-rich hyaline membranes on the denuded basement membrane. Neutrophils
are shown adhering to the injured and activated capillary endothelium and marginating through the interstitium into the
air space, which is ﬁlled with protein-rich edema ﬂuid. In the airspace an alveolar macrophage is secreting cytokines,
interleukin-1b, -6, -8, and -10, (IL-1b, 6, 8, and 10) and tumor necrosis factor-a (TNF-a), which act locally to stimulate
chemotaxis and activate neutrophils. Neutrophils can release oxidants, proteases, leukotrienes, and other proinﬂamma-
tory molecules. The inﬂux of protein-rich edema ﬂuid into the alveolus has led to the inactivation of surfactant. (From
Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med 2000;342(18):1339; with permission.
Ó 2000, Massachusetts Medical Society.)
acute phase of ALI/ARDS. Ultrastructural stud- to localized inﬂammation or injury (eg, bacterial
ies demonstrate endothelial cell swelling and pneumonia), this process becomes dysregulated
widening of the intercellular junctions , and ra- and uncontrolled in ALI/ARDS . Among the
dionuclide studies have conﬁrmed the presence of phenotypic and functional changes that occur,
profound capillary leak in these patients . More endothelial cell contraction and deranged vaso-
recent work, however, has established that endo- motor response contribute to the development
thelial structure and function may be altered in- of capillary leak; the expression of adhesion mol-
dependently of cellular injury in a process ecules and cytokines promulgates alveolar injury,
called ‘‘endothelial activation’’ that seems to pre- as discussed later .
cede and promote such injury. Endothelial acti- As with alveolar endothelial injury, the pres-
vation may occur in response to a wide range ence of epithelial damage, characterized by necro-
of stimuli implicated in ALI, including cytokines, sis and often profound disruption, has been
thrombin, lipopolysaccharide, and other micro- appreciated as a cardinal ﬁnding of ALI/ARDS
bial products, and extreme changes in blood for many years . The multiple consequences of
pressure. Whereas endothelial activation seems this injury and their central importance to the de-
to be part of a limited and reversible response velopment of edema in ALI have been recognized
MECHANISMS OF ALI/ARDS 581
only more recently. The alveolar epithelial lining endothelial and epithelial components of the alve-
is composed predominantly of ﬂat type I epithelial olar membrane and represents a loss of both bar-
cells that provide a thin surface for gas exchange rier function and ﬂuid resorption.
and a tight barrier against ﬂuid extravasation
into the airspace. This epithelial barrier function
seems to be even more critical than that played
by the endothelial surface of the ACM . Type Although increased permeability edema is the
II alveolar epithelial cells, although relatively central physiologic event in the development of
few in number, provide another crucial protection ALI/ARDS, the events leading to the failure of
against edema formation: the resorption of air- the alveolar capillary membrane are numerous
space ﬂuid . The degree of loss of this function and complex (Fig. 2). This diversity of pathways
in ALI/ARDS has been shown to correlate with invoked by the many inciting injuries leading to
poorer prognosis . This same study found ALI/ARDS and inﬂuenced by multiple host fac-
that increased rates of alveolar ﬂuid clearance in tors is the basis for much of the heterogeneity of
patients who had ALI/ARDS were associated disease manifestations and outcomes. One of the
with female sex, nonsmoking status, and ALI/ more intuitive examples of the potential diﬀer-
ARDS risk factors other than sepsis, suggesting ences among pathways leading to ALI is reﬂected
that some clinical heterogeneity may result from by the AECC categorization of risk factors for the
varying rates of alveolar ﬂuid clearance. Taken to- development of ALI/ARDS, which groups these
gether, the dual functions of the alveolar epithe- factors into direct and indirect (or pulmonary
lium underscore the dynamic nature of edema and extrapulmonary) injuries to the lung . Ap-
formation in which increases in ACM permeabil- parent diﬀerences in the radiographic and physio-
ity rapidly overtake the diminished resorptive logic manifestations of ALI caused by direct
capacity of the alveolus, and alveolar ﬂooding oc- versus indirect insults (and possibly in response
curs. Thus, the proteinaceous edema that charac- to some therapies) have led to ongoing contro-
terizes ALI/ARDS results from disruption of both versy about whether these groupings should be
Fig. 2. Pathophysiologic mechanisms of acute lung injury and the acute respiratory distress syndrome. In this schema-
tized view of the pathophysiologic pathways of ALI/ARDS, direct injuries to the lung damage the alveolar capillary
membrane and initiate local and subsequently systemic inﬂammatory cascades. Indirect injuries initiate the pathophys-
iologic pathways of ALI/ARDS primarily through release of systemic cytokines. Following both direct and indirect ini-
tiators of ALI/ARDS, the release of systemic inﬂammatory mediators activates circulating neutrophils and the vascular
endothelium of the lung, leading to pulmonary microvascular sequestration of neutrophils and inﬂammatory injury to
the ACM. Such injury results in failure of ACM barrier function and ﬂooding of the alveoli with proteinaceous edema
ﬂuid. Both ACM injury and alveolar edema cause surfactant loss and dysfunction, which promote alveolar instability
and collapse, driving further edema formation and alveolar injury, particularly in the setting of mechanical ventilation.
Not depicted is the role of the coagulation cascade, which promotes alveolar collapse, capillary obstruction with micro-
thrombi, and further release of inﬂammatory mediators.
582 SURATT & PARSONS
considered separate entities for the purposes of indirect causes of ALI/ARDS have conﬁrmed
further research and clinical trials [11,12]. the critical role of neutrophil recruitment in the in-
It seems logical that a direct injury to the jury to the alveolar membrane , and although
ACM, such as aspirated stomach acid, would studies in neutropenic patients have demonstrated
diﬀer from an indirect mechanism, such as extra- the occurrence of ALI/ARDS in the apparent ab-
thoracic trauma, and in fact ALI/ARDS mortality sence of neutrophils [22,23], it is widely recognized
from these two mechanisms diﬀers substantially that such patients worsen dramatically with the
. There also, however, are signiﬁcant diﬀerences resolution of neutropenia . The mechanisms
in mortality among indirect causes themselves: for governing neutrophil traﬃcking to the lung in
instance, sepsis with an extrapulmonary source ALI have been investigated intensely. Neutrophils
has a greater risk of mortality than extrathoracic are recruited to the lung through the release of sol-
trauma . Furthermore, there seems to be no uble mediators that activate these cells through
overall diﬀerence in mortality between direct and cell surface receptors. Some soluble factors, such
indirect causes of ALI/ARDS [12,13]. In animal as lipopolysaccharide (as seen in gram-negative
models examining direct injuries such as aspiration, sepsis), are triggers of the innate immune system,
the direct chemical injury to the ACM is only a small whereas others are endogenous signaling mole-
component of the resultant pathophysiologic pro- cules such as complement fragments, lipid factors,
cess . Thus, the relationship between inciting in- and, most importantly, cytokines.
jury and ALI/ARDS pathophysiology (at least as The role of cytokine signaling in ALI/ARDS is
reﬂected by mortality) seems to be more complex. vast and complex , but it can be schematized
Recent epidemiologic studies of ALI/ARDS have using the paradigm of the innate immune response
suggested that additional factors such as age, sex, . The cascade of cytokine release that occurs
and race may inﬂuence both the risk of developing during such a response begins with the activation
the disease and subsequent mortality, and these of monocytes and macrophages in the lung (in di-
ﬁndings probably also reﬂect diﬀerences in the rect injury) or in blood or other tissues (in indirect
pathophysiology of the disease as manifested in injury). This activation stimulates the release of
these patients [2,15–17]. Therefore, any apprecia- the ﬁrst wave of cytokine signal: tumor necrosis
tion of ALI/ARDS pathophysiology must take factor-a (TNF-a) and interleukin-1b (IL-1b),
in to account the numerous underlying pathways also known as ‘‘early-response cytokines.’’ These
thus far implicated in this disease and the grow- key pleiotropic cytokines act upon leukocytes
ing number of host and other factors that seem and other cell types (such as the alveolar epithe-
to inﬂuence their expression. lium and endothelium and lung ﬁbroblasts) to ini-
tiate a cascade of secondary cytokines and to
release other soluble mediators, diversifying and
amplifying the inﬂammatory signal. Critical
Leukocytes and soluble mediators
among these secondary cytokines are those acting
Much of the evolved understanding of ALI/ primarily on neutrophils (CXC cytokines, the
ARDS pathophysiology has centered on the role most important of which is IL-8) and monocytes
of leukocytes and, most recently, the numerous (CC cytokines, eg, macrophage inﬂammatory
soluble mediators that both drive their recruit- protein-1), which serve both to activate and to
ment and inﬂuence their behavior (see Fig. 1). As recruit these cells. Recent evidence suggests that
with the initial appreciation of alveolar epithelial inﬂammation in ALI/ARDS results not so much
and endothelial injury in ALI/ARDS, early histo- from the unopposed actions of these proinﬂam-
logic studies also identiﬁed neutrophils as a likely matory cytokines as from the net balance of a mi-
key participant in this injury [1,4]. Since this ob- lieu of both pro- and anti-inﬂammatory mediators
servation was made, studies have demonstrated (eg, IL-10, soluble TNF-a receptor, and IL-1 re-
neutrophils to be the predominant cell type in ceptor antagonist) . Also crucial during this
bronchoalveolar lavage of patients who have inﬂammatory cascade are cytokines (including
ALI/ARDS, and the persistence of this ﬁnding TNF-a and IL-1b) that activate the vascular
seems to correlate with increased mortality endothelium, particularly that of the lung, leading
[18,19]. Examination of bronchoalveolar lavage to expression of leukocyte adhesion molecules .
has also implicated the release of injurious neutro- Activation of both circulating neutrophils
phil contents, such as proteases and oxidants, in and the vascular endothelium leads to rapid
the process . Animal models of direct and sequestration of neutrophils to the microvascular
MECHANISMS OF ALI/ARDS 583
beds of the lung (Fig. 2). Unique among the tissues converts plasminogen to the ﬁbrinolytic factor,
of the body, the lung uses both classic leukocyte plasmin. Plasmin, in turn, cleaves the small
adhesion mechanisms in this process (selectin- amount of ﬁbrin that is formed through the co-
and integrin-mediated endovascular adhesion and agulation cascade, which is largely quiescent in
diapedesis ) and also a ﬁlter-like form of me- the resting lung. In the setting of injury to the al-
chanical arrest, in which neutrophils stiﬀen after veolar capillary membrane, this balance is over-
activation and can no longer pass through the whelmed by the leakage of coagulation factors
small capillaries of the alveoli [29,30]. This mecha- into the interstitium and airspace of the alveolus.
nism seems to be accentuated by the release of im- This imbalance, coupled with the expression of
mature marrow neutrophils, which are larger and procoagulant molecules (such as tissue factor)
less deformable [31,32], and may in part explain by the injured endothelial and epithelial compo-
the elevated risk of ALI/ARDS in patients during nents of the ACM and an increase in inhibitors
the recovery from neutropenia [24,33]. of ﬁbrinolysis (such as plasminogen activator-1,
Given the importance of the innate immune or as plasminogen-activator inhibitor type 1 [PAI-
system in the development of ALI/ARDS, it is 1]), leads to unopposed procoagulant activity
easy to appreciate how genetic variation involving . The resulting formation of ﬁbrin-platelet
cytokines and their receptors might lead to microthrombi in the capillaries and ﬁbrin-rich
signiﬁcant variability in the expression of ALI/ proteinaceous casts in the airspaces of the lung
ARDS after a given injury. For example, poly- contributes signiﬁcantly to ventilation-perfusion
morphisms in the gene for TNF-a have been mismatching in this disease. What has only re-
shown to be associated with increased susceptibil- cently been appreciated is the interaction be-
ity to and mortality from ALI/ARDS , and tween this imbalance of alveolar coagulation
multiple other polymorphisms in both pro- and and ﬁbrinolysis and the inﬂammatory processes
anti-inﬂammatory mediators and receptors such of ALI/ARDS.
as IL-1 receptor antagonist and toll-like recep- Recent studies have emphasized the multiple
tor-4 (the receptor for lipopolysaccharide) have eﬀects of thrombin and ﬁbrin deposition on local
been shown to inﬂuence the clinical manifesta- and systemic cellular activation and cytokine
tions of sepsis and possibly ALI/ARDS . Fur- production . Fibrin and thrombin seem to ac-
thermore, the pivotal role of neutrophils and their tivate both neutrophils and the vascular endothe-
cytotoxic contents in ALI/ARDS may explain, in lium directly, leading to augmented expression of
part, the epidemiologic observations that chronic adhesion molecules on both cell types and, in the
alcoholism, which is associated with a profound case of the endothelium, to increased capillary
deﬁciency in the antioxidant glutathione, is associ- permeability . Such activation also drives the
ated with increased susceptibility to and mortality further release of TNF-a and IL-1b and of other
from ALI/ARDS , whereas diabetic patients, cytokines, such as IL-8, which further propagate
who manifest defects in neutrophil activation the inﬂammatory response . Platelets seem to
and function, are at a lower risk of developing have similar proinﬂammatory actions in ALI/
ALI/ARDS . ARDS but also may have some ameliorating ef-
fects on the capillary leak, making their net eﬀect
in this process unclear . Although trials of sev-
eral anticoagulant therapies for ALI/ARDS have
Coagulation and platelets
yielded mixed results , recent success with acti-
The appearance of intra-alveolar hyaline mem- vated protein C in the treatment of sepsis has sug-
branes and microvascular thrombi has been rec- gested that this therapy may have a role in the
ognized as a histologic hallmark of acute ALI/ treatment of ALI/ARDS and has conﬁrmed that
ARDS for decades (see Fig. 1) [1,4], but the full such an approach may indeed attenuate both the
implications of these ﬁndings for the pathophys- thrombotic and inﬂammatory events of ALI/
iology of ALI/ARDS have been appreciated only ARDS [41,42]. Furthermore, studies of genetic
recently. Both ﬁndings reﬂect the exuberant de- polymorphisms of coagulation proteins, such as
position of ﬁbrin in the lung and represent a pro- PAI-1 and thrombospondin, have related these
found imbalance in production and degradation polymorphisms to variations in both proinﬂam-
of this molecule. In the uninjured lung a net ﬁ- matory cytokine levels and clinical outcomes for
brinolytic state is maintained through the action patients who have sepsis and ALI/ARDS
of urokinase plasminogen activator, which [43,44]. This ﬁnding also suggests that genetic
584 SURATT & PARSONS
variation may inﬂuence the clinical expression of mortality in those who had ALI/ARDS caused by
ALI/ARDS through polymorphisms of the cyto- indirect injury . Furthermore, genetic studies
kines themselves and of the coagulation pathways have suggested that certain polymorphisms in
as well . the SP-B protein may confer a greater risk for
the development of ALI/ARDS, particularly after
direct lung injury . Thus, the importance of
surfactant dysfunction may be inﬂuenced by the
mechanism of injury and by host factors.
One of the earliest theories about the patho-
physiology of ALI/ARDS concerned lung surfac-
tant and its dysfunction during ALI/ARDS . Ventilator-induced lung injury
The role of surfactant in this disease has since
been appreciated to be complex and, as regards Since the early observations of Webb and
possible therapeutic applications, somewhat per- Tierney  that high tidal volume ventilation in
plexing. Surfactant is a lipoprotein complex, rats leads to proteinaceous lung edema and
composed of phospholipids, neutral lipids, and histologic ﬁndings consistent with ALI/ARDS,
surfactant proteins (SP-A, B, C, and D), that is se- mechanical ventilation has been suspected to in-
creted by the type II alveolar epithelium and lines ﬂuence the manifestation of ALI/ARDS in hu-
the alveolar surfaces. Its functions are both bio- mans. How such an interaction might be
physical and immunologic, in that it serves to de- particularly accentuated in ALI/ARDS during
crease surface tension at the air/liquid interface even physiologic levels of ventilation was sug-
(maintaining alveolar patency), and participates gested subsequently by radiographic studies dem-
in critical elements of innate host defense . onstrating the extreme heterogeneity of alveolar
Studies of bronchoalveolar lavage in patients patency in this disease . These studies revealed
who have ALI/ARDS have demonstrated that that aeration of the lung varies markedly, not only
surfactant is altered quite early in the course of by lung region (with widespread patchy consolida-
the disease: the composition and form of surfac- tion, particularly in the more dependent areas) but
tant changes, and it develops abnormal surface also by the respiratory cycle. Thus, in the course
tension properties [47,48]. These changes seem to of cyclic tidal volume delivery, three areas may
be driven by the inﬂux of edema ﬂuid and serum be considered to exist :
proteins into the alveoli and worsened by injury
1. Fluid-ﬁlled or collapsed areas in which the al-
to the type II epithelium (as described previously),
veoli never inﬂate during the respiratory cycle
ultimately leading to alveolar instability and col-
and which lead to a decrease in total lung
lapse, as well as to altered immune function (see
Fig. 1). Alveolar collapse decreases lung compli-
2. Patent areas in which alveoli may become
ance, worsens hypoxemia, and draws additional
overdistended during inspiration because of
edema ﬂuid in to the alveoli , driving a vicious
shunting of delivered breath away from col-
cycle of further surfactant dysfunction and alveo-
lar edema (see Fig. 2).
3. Atelectatic areas in which the alveoli repeat-
Given the importance of surfactant in the
edly open and close with the respiratory cycle
pathophysiology of ALI/ARDS, the use of re-
because of alveolar instability, particularly
combinant surfactant to treat ALI/ARDS has
at lower levels of positive end-expiration
been of great interest [50,51]. Unlike the positive
results seen in neonatal respiratory distress syn-
drome (nRDS), however, trials of surfactant in Therefore, mechanical ventilation in ALI/
ALI/ARDS have been disappointing. These re- ARDS may lead to recurrent overdistension of
sults seem to underscore the relatively greater im- the patent alveoli and to shear injury in the areas
portance of impaired surfactant production in exposed to cyclic atelectasis.
nRDS versus surfactant function in ALI/ARDS Alveolar overdistension and shear seem to
. Post hoc analysis of the most recent trial of perpetuate and augment lung injury in three
surfactant replacement for ALI/ARDS suggested important ways: local and systemic inﬂammatory
that such therapy reduced mortality in patients signaling, direct disruption of the ACM, and
who had ALI/ARDS secondary to direct injury impairment of alveolar ﬂuid clearance (see
(pneumonia or aspiration) but may have increased Fig. 2). Animal studies have shown that
MECHANISMS OF ALI/ARDS 585
mechanical ventilation, particularly at high tidal as much by these corrective responses as by the
volumes, leads to both pulmonary and systemic initiating and ongoing injuries discussed previously.
release of inﬂammatory cytokines, including Resolution of ALI/ARDS involves the termination
TNF-a, IL-1b, and IL-6, and the recruitment of of the inﬂammatory response, the clearance of both
neutrophils to the lung. This process seems to be ﬂuid and debris from the alveoli, and the repair of
driven, in part, by cyclical stretching of lung mac- the ACM (see Fig. 1). Clinical and animal studies
rophages and type II pneumocytes, which pro- have suggested that an imbalance of corrective
motes activation of these cells [56,57]. Clinical and injurious forces or dysregulation of the repair
studies conﬁrm that inﬂammatory cytokines, in- process itself may lead to pathophysiology in the
cluding IL-6 and IL-8, are released into the lung later phase of ALI/ARDS and to poor outcome
and blood during mechanical ventilation of pa- from this disease .
tients who have ALI/ARDS, and that this eﬀect Fluid resorption from the alveolar airspace
is attenuated by the use of lower tidal volumes and interstitium is dependent on the active trans-
(6 mL/kg) than traditionally used (10–12 mL/kg) port of sodium (and with it, water) by the type II
[58–60]. In addition to the possible proinﬂamma- epithelium and reconstitution of the ACM barrier.
tory eﬀects of mechanical ventilation, direct me- Failure of ﬂuid resorption, even as measured quite
chanical injury to the ACM may occur at high early in the disease process, is predictive of
tidal volumes as the result of cell membrane dis- mortality from ALI/ARDS . Thus, regenera-
ruption [61,62] and mechanically transduced acti- tion of the alveolar epithelium structure and func-
vation of the alveolar endothelium . These tion, as described later, is a critical event in the
processes seem to be exacerbated by concomitant recovery from ALI/ARDS. The removal of cellu-
ALI/ARDS, so that even tidal volumes consid- lar and other debris from the airspace is also im-
ered to be physiologic may provoke signiﬁcant di- portant, both in the restoration of gas exchange
rect injury to the lung . Last, mechanical and in the resolution of inﬂammation and repair
ventilation in ALI/ARDS may reduce airspace of the damaged epithelium. Apoptosis (or con-
edema clearance further through alteration of trolled cell death) is the central mechanism of
type II pneumocyte function . cell clearance for both injured alveolar epithelium
The observation that mechanical ventilation, and recruited inﬂammatory cells. Apoptosis seems
even at physiologic tidal volumes, may compound to be a double-edged sword in ALI/ARDS: al-
the severity of lung injury in ALI/ARDS led to though the induction of apoptosis may contribute
a series of clinical trials examining low tidal to epithelial injury in early ALI/ARDS , it also
volume ventilation strategies for this disease. is critical in the removal of neutrophils by airspace
The largest of these trials demonstrated that a macrophages . Furthermore, the clearance of
6-mL/kg (ideal body weight) tidal volume strategy apoptotic cells from the airspace seems to alter
led to a reduction in ALI/ARDS mortality from the cytokine milieu and attenuate the inﬂamma-
40% to 31% compared with a conventional tidal tory response in the lung . In addition to the
volume strategy of 12 mL/kg . The signiﬁcance removal of cellular debris, the clearance of insolu-
of this ﬁnding cannot be overemphasized, because ble proteinaceous debris, particularly hyaline
to date it represents the only speciﬁc intervention membranes, is essential for gas exchange and for
ever shown in a large-scale trial to reduce mortal- restoration of the normal architecture of the
ity from this disease signiﬁcantly. It also suggests alveolus.
that yet another factor may dramatically inﬂuence Repair of the ACM is a complex process
the pathophysiology and outcome of ALI/ARDS: involving local cellular proliferation and, proba-
physician practice. bly, the recruitment of circulating progenitor cells.
As the clearance of airspace edema and debris
progresses, repair of the alveolar structure occurs
through the orderly replacement of its cellular
Late-phase pathophysiology: dysregulated
constituents. The denuded alveolar epithelium is
and maladaptive repair
replaced through a proliferation of the remaining
Even as the multiple pathophysiologic processes type II pneumocytes, which spread to cover the
of early ALI/ARDS promulgate alveolar injury and exposed basement membrane and subsequently
edema, reparative processes are initiated within the transdiﬀerentiate into type I cells. At the same
lung that drive toward resolution of the disease. time, the alveolar interstitium is restored through
Thus, the course of ALI/ARDS may be determined the orderly expansion of ﬁbroblasts and
586 SURATT & PARSONS
deposition of extracellular matrix proteins, the The frustration of generations of clinicians and
alveolar microcirculation is recanalized, and the researchers in developing consistent paradigms for
damaged endothelium is replaced. Although re- the understanding and treatment of ALI/ARDS
pair to the damaged lung in ALI/ARDS can lead underscores the fundamental variability of mech-
to substantial, although incomplete, restoration of anisms that underlie this condition. Recent prog-
lung function in survivors , divergence from ress in the ﬁelds of immunology, molecular
this choreographed sequence of events may lead biology, genetics, and epidemiology has allowed
to severe consequences. The best described of greater appreciation of the multiple factors inﬂu-
these consequences is the ‘‘ﬁbroproliferative encing this disease. In the end, each individual
phase’’ of ALI/ARDS that may develop approxi- aﬀected by ALI/ARDS must be seen as represent-
mately 5 to 7 days into the disease course and is ing a unique combination of genetics, comorbid
characterized by an exuberant ﬁbrotic response conditions, inciting events, and factors yet to be
 that has been likened to wound healing in the discovered that drive the process of ALI with
alveolar airspace . This maladaptive repair re- variable contributions from the many pathophys-
sponse seems to be promoted by severe injury to iologic mechanisms elucidated in the past 40
the ACM and extensive hyaline membrane forma- years.
tion, which lead to migration of myoﬁbroblasts
into the alveoli and subsequent formation of gran-
ulation tissue with physiologic compromise .
Recent studies indicate that these migrating cells
may, in fact, be recruited from the bone marrow  Ashbaugh DG, Bigelow DB, Petty TL, et al. Acute
, suggesting that blockade of this cytokine-me- respiratory distress in adults. Lancet 1967;2(7511):
diated recruitment might represent a future thera- 319–23.
peutic approach to this devastating complication  Rubenfeld GD, Caldwell E, Peabody E, et al. Inci-
. dence and outcomes of acute lung injury. N Engl
J Med 2005;353(16):1685–93.
Circulating precursor cells also seem to be
 Bernard GR, Artigas A, Brigham KL, et al. The
critical in the normal restoration of the ACM. American-European Consensus Conference on
Animal models of ALI/ARDS have suggested ARDS. Deﬁnitions, mechanisms, relevant outcomes,
that the repair of alveolar endothelium (and and clinical trial coordination. Am J Respir Crit Care
perhaps epithelium) requires bone marrow– Med 1994;149(3 Pt 1):818–24.
derived cells and that failure of these cells to  Bachofen M, Weibel ER. Structural alterations of
migrate to the lung may yield signiﬁcant anatomic lung parenchyma in the adult respiratory distress
derangement [73,74]. The clinical relevance of syndrome. Clin Chest Med 1982;3(1):35–56.
such ﬁndings is further suggested by the marked  Raijmakers PG, Groeneveld AB, Teule GJ, et al.
elevation of circulating endothelial precursors Diagnostic value of the gallium-67 pulmonary
leak index in pulmonary edema. J Nucl Med 1996;
found during ALI/ARDS and by the apparent
correlation between lower circulating levels of  Zimmerman GA, Albertine KH, Carveth HJ, et al.
these cells and subsequent mortality from the dis- Endothelial activation in ARDS. Chest 1999;116
ease . Higher lung levels of vascular endothe- (1 Suppl):18S–24S.
lial growth factor (VEGF), a chemoattractant  Piantadosi CA, Schwartz DA. The acute respiratory
for endothelial precursors, are associated with re- distress syndrome. Ann Intern Med 2004;141(6):
covery from ALI/ARDS , whereas certain ge- 460–70.
netic polymorphisms of VEGF associated with  Wiener-Kronish JP, Albertine KH, Matthay MA.
lower VEGF production seem to confer elevated Diﬀerential responses of the endothelial and
risk for the development of the disease . epithelial barriers of the lung in sheep to Escher-
ichia coli endotoxin. J Clin Invest 1991;88(3):
Thus, the inﬂuence of host factors seems to extend
to the recovery phase of ALI/ARDS as well.  Matthay MA, Folkesson HG, Clerici C. Lung epi-
thelial ﬂuid transport and the resolution of pulmo-
nary edema. Physiol Rev 2002;82(3):569–600.
Summary  Ware LB, Matthay MA. Alveolar ﬂuid clearance is
impaired in the majority of patients with acute
The pathophysiology of ALI/ARDS represents lung injury and the acute respiratory distress syn-
a complex and protean expression of multiple drome. Am J Respir Crit Care Med 2001;163(6):
processes culminating in a common end point. 1376–83.
MECHANISMS OF ALI/ARDS 587
 Rocco PR, Zin WA. Pulmonary and extrapulmo-  Strieter RM, Belperio JA, Keane MP. Cytokines in
nary acute respiratory distress syndrome: are they innate host defense in the lung. J Clin Invest 2002;
diﬀerent? Curr Opin Crit Care 2005;11(1):10–7. 109(6):699–705.
 Callister ME, Evans TW. Pulmonary versus extrap-  Park WY, Goodman RB, Steinberg KP, et al. Cyto-
ulmonary acute respiratory distress syndrome: dif- kine balance in the lungs of patients with acute respi-
ferent diseases or just a useful concept? Curr Opin ratory distress syndrome. Am J Respir Crit Care
Crit Care 2002;8(1):21–5. Med 2001;164(10 Pt 1):1896–903.
 Eisner MD, Thompson T, Hudson LD, et al. Eﬃ-  Albelda SM, Smith CW, Ward PA. Adhesion mole-
cacy of low tidal volume ventilation in patients cules and inﬂammatory injury. FASEB J 1994;8(8):
with diﬀerent clinical risk factors for acute lung in- 504–12.
jury and the acute respiratory distress syndrome.  Worthen GS, Schwab BD, Elson EL, et al. Mechan-
Am J Respir Crit Care Med 2001;164(2):231–6. ics of stimulated neutrophils: cell stiﬀening induces
 Modelska K, Pittet JF, Folkesson HG, et al. Acid- retention in capillaries. Science 1989;245(4914):
induced lung injury. Protective eﬀect of anti-inter- 183–6.
leukin-8 pretreatment on alveolar epithelial barrier  Doerschuk CM, Mizgerd JP, Kubo H, et al. Adhe-
function in rabbits. Am J Respir Crit Care Med sion molecules and cellular biomechanical changes
1999;160(5 Pt 1):1450–6. in acute lung injury: Giles F. Filley Lecture. Chest
 Zilberberg MD, Epstein SK. Acute lung injury in the 1999;116(1 Suppl):37S–43S.
medical ICU: comorbid conditions, age, etiology,  Kubo H, Graham L, Doyle NA, et al. Complement
and hospital outcome. Am J Respir Crit Care Med fragment-induced release of neutrophils from bone
1998;157(4 Pt 1):1159–64. marrow and sequestration within pulmonary capil-
 Moss M, Mannino DM. Race and gender diﬀerences laries in rabbits. Blood 1998;92(1):283–90.
in acute respiratory distress syndrome deaths in the  Van Eeden SF, Kitagawa Y, Klut ME, et al. Poly-
United States: an analysis of multiple-cause mortal- morphonuclear leukocytes released from the bone
ity data (1979-1996). Crit Care Med 2002;30(8): marrow preferentially sequester in lung microves-
1679–85. sels. Microcirculation 1997;4(3):369–80.
 Barnes KC. Genetic determinants and ethnic dispar-  Lee CK, Gingrich RD, Hohl RJ, et al. Engraftment
ities in sepsis-associated acute lung injury. Proc Am syndrome in autologous bone marrow and periph-
Thorac Soc 2005;2(3):195–201. eral stem cell transplantation. Bone Marrow Trans-
 Fowler AA, Hyers TM, Fisher BJ, et al. The adult plant 1995;16(1):175–82.
respiratory distress syndrome. Cell populations  Gong MN, Zhou W, Williams PL, et al. -308GA and
and soluble mediators in the air spaces of patients TNFB polymorphisms in acute respiratory distress
at high risk. Am Rev Respir Dis 1987;136(5): syndrome. Eur Respir J 2005;26(3):382–9.
1225–31.  Moss M, Burnham EL. Chronic alcohol abuse,
 Steinberg KP, Milberg JA, Martin TR, et al. Evolu- acute respiratory distress syndrome, and multiple or-
tion of bronchoalveolar cell populations in the adult gan dysfunction. Crit Care Med 2003;31(4 Suppl):
respiratory distress syndrome. Am J Respir Crit S207–12.
Care Med 1994;150(1):113–22.  Moss M, Guidot DM, Steinberg KP, et al. Diabetic
 Cochrane CG, Spragg R, Revak SD. Pathogenesis patients have a decreased incidence of acute respira-
of the adult respiratory distress syndrome. Evidence tory distress syndrome. Crit Care Med 2000;28(7):
of oxidant activity in bronchoalveolar lavage ﬂuid. 2187–92.
J Clin Invest 1983;71(3):754–61.  Gunther A, Mosavi P, Heinemann S, et al. Alveolar
 Matthay MA, Zimmerman GA. Acute lung injury ﬁbrin formation caused by enhanced procoagulant
and the acute respiratory distress syndrome: four de- and depressed ﬁbrinolytic capacities in severe pneu-
cades of inquiry into pathogenesis and rational man- monia. Comparison with the acute respiratory dis-
agement. Am J Respir Cell Mol Biol 2005;33(4): tress syndrome. Am J Respir Crit Care Med 2000;
319–27. 161(2 Pt 1):454–62.
 Ognibene FP, Martin SE, Parker MM, et al. Adult  Idell S. Coagulation, ﬁbrinolysis, and ﬁbrin deposi-
respiratory distress syndrome in patients with severe tion in acute lung injury. Crit Care Med 2003;31
neutropenia. N Engl J Med 1986;315(9):547–51. (4 Suppl):S213–20.
 Laufe MD, Simon RH, Flint A, et al. Adult respira-  Abraham E. Coagulation abnormalities in acute
tory distress syndrome in neutropenic patients. Am J lung injury and sepsis. Am J Respir Cell Mol Biol
Med 1986;80(6):1022–6. 2000;22(4):401–4.
 Azoulay E, Darmon M, Delclaux C, et al. Deteriora-  Laterre PF, Wittebole X, Dhainaut JF. Anticoagu-
tion of previous acute lung injury during neutrope- lant therapy in acute lung injury. Crit Care Med
nia recovery. Crit Care Med 2002;30(4):781–6. 2003;31(4 Suppl):S329–36.
 Puneet P, Moochhala S, Bhatia M. Chemokines in  Nick JA, Coldren CD, Geraci MW, et al. Recombi-
acute respiratory distress syndrome. Am J Physiol nant human activated protein C reduces human
Lung Cell Mol Physiol 2005;288(1):L3–15. endotoxin-induced pulmonary inﬂammation via
588 SURATT & PARSONS
inhibition of neutrophil chemotaxis. Blood 2004;  Pugin J, Dunn I, Jolliet P, et al. Activation of human
104(13):3878–85. macrophages by mechanical ventilation in vitro. Am
 Hoﬀmann JN, Vollmar B, Laschke MW, et al. J Physiol 1998;275(6 Pt 1):L1040–50.
Microhemodynamic and cellular mechanisms of ac-  Ranieri VM, Suter PM, Tortorella C, et al. Eﬀect of
tivated protein C action during endotoxemia. Crit mechanical ventilation on inﬂammatory mediators
Care Med 2004;32(4):1011–7. in patients with acute respiratory distress syndrome:
 Menges T, Hermans PW, Little SG, et al. Plasmino- a randomized controlled trial. JAMA 1999;282(1):
gen-activator-inhibitor-1 4G/5G promoter poly- 54–61.
morphism and prognosis of severely injured  ARDS Clinical Trials Network. Ventilation with
patients. Lancet 2001;357(9262):1096–7. lower tidal volumes as compared with traditional
 Maloney J, Strong D, Morrison J, et al. Increased tidal volumes for acute lung injury and the acute re-
prevalence of variant polymorphisms in the throm- spiratory distress syndrome. The Acute Respiratory
bospondin-1 (TSP1) gene in acute lung injury. Am Distress Syndrome Network. N Engl J Med 2000;
J Respir Crit Care Med 2002;165(8):A473. 342(18):1301–8.
 Russell JA. Genetics of coagulation factors in acute  Parsons PE, Eisner MD, Thompson BT, et al. Lower
lung injury. Crit Care Med 2003;31(4 Suppl): tidal volume ventilation and plasma cytokine
S243–7. markers of inﬂammation in patients with acute
 Wright JR. Immunoregulatory functions of sur- lung injury. Crit Care Med 2005;33(1):1–6 [discus-
factant proteins. Nat Rev Immunol 2005;5(1): sion: 230–2].
58–68.  Dreyfuss D, Soler P, Basset G, et al. High inﬂation
 Gregory TJ, Longmore WJ, Moxley MA, et al. Sur- pressure pulmonary edema. Respective eﬀects of
factant chemical composition and biophysical activ- high airway pressure, high tidal volume, and positive
ity in acute respiratory distress syndrome. J Clin end-expiratory pressure. Am Rev Respir Dis 1988;
Invest 1991;88(6):1976–81. 137(5):1159–64.
 Lewis JF. Surfactant in lung injury: how important  West JB, Tsukimoto K, Mathieu-Costello O, et al.
is it? Journal of Organ Dysfunction 2006;2:20–3. Stress failure in pulmonary capillaries. J Appl Phys-
 Albert RK, Lakshminarayan S, Hildebrandt J, et al. iol 1991;70(4):1731–42.
Increased surface tension favors pulmonary edema  Parker JC, Ivey CL, Tucker A. Phosphotyrosine
formation in anesthetized dogs’ lungs. J Clin Invest phosphatase and tyrosine kinase inhibition modu-
1979;63(5):1015–8. late airway pressure-induced lung injury. J Appl
 Anzueto A, Baughman RP, Guntupalli KK, et al. Physiol 1998;85(5):1753–61.
Aerosolized surfactant in adults with sepsis-induced  Frank JA, Gutierrez JA, Jones KD, et al. Low tidal
acute respiratory distress syndrome. Exosurf Acute volume reduces epithelial and endothelial injury in
Respiratory Distress Syndrome Sepsis Study Group. acid-injured rat lungs. Am J Respir Crit Care Med
N Engl J Med 1996;334(22):1417–21. 2002;165(2):242–9.
 Spragg RG, Lewis JF, Walmrath HD, et al. Eﬀect of  Lecuona E, Saldias F, Comellas A, et al. Ventilator-
recombinant surfactant protein C-based surfactant associated lung injury decreases lung ability to clear
on the acute respiratory distress syndrome. N Engl edema in rats. Am J Respir Crit Care Med 1999;
J Med 2004;351(9):884–92. 159(2):603–9.
 Gong MN, Wei Z, Xu LL, et al. Polymorphism in  Martin TR, Nakamura M, Matute-Bello G. The role
the surfactant protein-B gene, gender, and the risk of apoptosis in acute lung injury. Crit Care Med
of direct pulmonary injury and ARDS. Chest 2004; 2003;31(4 Suppl):S184–8.
125(1):203–11.  Hussain N, Wu F, Zhu L, et al. Neutrophil apoptosis
 Webb HH, Tierney DF. Experimental pulmonary during the development and resolution of oleic acid-
edema due to intermittent positive pressure ventila- induced acute lung injury in the rat. Am J Respir Cell
tion with high inﬂation pressures. Protection by pos- Mol Biol 1998;19(6):867–74.
itive end-expiratory pressure. Am Rev Respir Dis  Huynh ML, Fadok VA, Henson PM. Phosphatidyl-
1974;110(5):556–65. serine-dependent ingestion of apoptotic cells pro-
 Gattinoni L, Caironi P, Pelosi P, et al. What has motes TGF-beta1 secretion and the resolution of
computed tomography taught us about the acute re- inﬂammation. J Clin Invest 2002;109(1):41–50.
spiratory distress syndrome? Am J Respir Crit Care  Herridge MS, Cheung AM, Tansey CM, et al. One-
Med 2001;164(9):1701–11. year outcomes in survivors of the acute respiratory
 Frank JA, Matthay MA. Science review: mecha- distress syndrome. N Engl J Med 2003;348(8):
nisms of ventilator-induced injury. Crit Care 2003; 683–93.
7(3):233–41.  Bitterman PB. Pathogenesis of ﬁbrosis in acute lung
 Dunn I, Pugin J. Mechanical ventilation of various injury. Am J Med 1992;92(6A):39S–43S.
human lung cells in vitro: identiﬁcation of the mac-  Hashimoto N, Jin H, Liu T, et al. Bone marrow-
rophage as the main producer of inﬂammatory me- derived progenitor cells in pulmonary ﬁbrosis.
diators. Chest 1999;116(1 Suppl):95S–7S. J Clin Invest 2004;113(2):243–52.
MECHANISMS OF ALI/ARDS 589
 Phillips RJ, Burdick MD, Hong K, et al. Circulating  Burnham EL, Taylor WR, Quyyumi AA, et al. In-
ﬁbrocytes traﬃc to the lungs in response to CXCL12 creased circulating endothelial progenitor cells are
and mediate ﬁbrosis. J Clin Invest 2004;114(3): associated with survival in acute lung injury. Am J
438–46. Respir Crit Care Med 2005;172(7):854–60.
 Ishizawa K, Kubo H, Yamada M, et al. Hepato-  Thickett DR, Armstrong L, Millar AB. A role for
cyte growth factor induces angiogenesis in injured vascular endothelial growth factor in acute and re-
lungs through mobilizing endothelial progenitor solving lung injury. Am J Respir Crit Care Med
cells. Biochem Biophys Res Commun 2004;324(1): 2002;166(10):1332–7.
276–80.  Medford AR, Thickett DR, Keen L, et al. Frequency
 Yamada M, Kubo H, Kobayashi S, et al. Bone of vascular endothelial growth factor (VEGF) 936 C/T
marrow-derived progenitor cells are important for polymorphism in patients with or at risk of adult re-
lung repair after lipopolysaccharide-induced lung spiratory distress syndrome (ARDS). Am J Respir
injury. J Immunol 2004;172(2):1266–72. Crit Care Med 2002;165(8):A474.