INTRODUCTION

TOP ABSTRACT INTRODUCTION MODULATION OF LUNG FLUID... MECHANISMS INVOLVED IN THE... TREATMENT STRATEGIES TO... REFERENCES

REABSORPTION OF ALVEOLAR EDEMA fluid is essential for resolution of pulmonary edema and may have significant impact on the prognosis of patients with pulmonary edema (62). Since the initial observations made by Matthay and colleagues in the early 1980s, several studies have demonstrated that resolution of alveolar edema depends on active ion transport across the alveolar epithelium (2, 33). The physiological regulation and identification of several cellular and molecular mechanisms involved in transepithelial ion transport are relatively well understood. However, it is only recently that a number of investigators have evaluated regulation of transport in injured lungs and determined whether the resolution can be modulated in the presence of lung injury. In this brief review, we will discuss how alveolar fluid clearance is altered in injured lungs and which mechanisms may be responsible for modulating the rate of alveolar fluid clearance in injured lungs.

	MODULATION OF LUNG FLUID CLEARANCE IN THE INJURED LUNG

TOP ABSTRACT INTRODUCTION MODULATION OF LUNG FLUID... MECHANISMS INVOLVED IN THE... TREATMENT STRATEGIES TO... REFERENCES

Modulation of lung liquid clearance can occur in the injured lung because the alveolar and distal airway epithelia are remarkably resistant to injury (39). Furthermore, migration of neutrophils into the air spaces of the lung, a phenomenon associated with lung injury, is not always associated with an increase in permeability to proteins and dysfunction of the alveolar epithelium (29). Even when lung endothelial injury occurs, the alveolar epithelial barrier may retain its normal impermeability to protein and its normal fluid transport capacity. For example, when intravenous or intra-alveolar endotoxin was used to produce lung endothelial injury in sheep (63) and rats (45), the permeability to protein across the lung epithelial barrier was not altered. Furthermore, histologically the ultrastructural appearance of the alveolar epithelium cell was normal, whereas there were signs of endothelial cell injury as shown by the increased number of pinocytic vesicles (Fig. 1). In these injury models in which the alveolar epithelium is intact, there was no impairment of alveolar fluid clearance. However, when there is some evidence of increased permeability of the alveolar epithelium to protein, the rate of alveolar fluid clearance was modestly reduced (63).

View larger version (154K): [in this window] [in a new window]   Fig. 1.   Ultrastructural appearance of lungs obtained 4 h after instillation of autologous serum and the intravenous infusion of 5 µg/kg Escherichia coli endotoxin in 1 sheep (A) compared with the ultrastructural appearance of lungs obtained 4 h after instillation of serum with alveolar endotoxin (100 µg/kg) (B). In both conditions, although the alveolar spaces (A) are filled with the instilled serum, the endothelial cells (EN), interstitial cells (IC), and epithelial cells (EP) have retained their normal morphological characteristics. There are, however, increased numbers of pinocytotic vesicles (arrows), suggesting the presence of initial signs of injury to the endothelial cells. C, alveolar capillary. [From Wiener-Kronish et al. (63).]

Interestingly, in mild-to-moderate lung injury, with a modest increase in the wet-to-dry ratio (from 5 to 6 g H₂O/g dry lung), the capacity of the alveolar epithelium to transport salt and water is not only preserved but may even be upregulated. When septic shock was induced in rats (45), even though endothelial injury and interstitial pulmonary edema occurred, alveolar epithelial fluid transport was increased by 32% in septic rats. This enhanced fluid clearance was inhibited by instillation of amiloride (10⁴ M) or propranolol (10⁴ M) into the distal air spaces, demonstrating that the stimulated fluid clearance depended on endogenous -adrenergic agonist stimulation of alveolar epithelial sodium transport. After short-term hemorrhagic shock, alveolar fluid clearance can also be enhanced by catecholamine-dependent mechanisms (37). In addition, lung fluid clearance can also be stimulated in lung injury by catecholamine-independent mechanisms. In bacterial pneumonia induced by the instillation of Pseudomonas into the distal airways, there is a 48% increase in alveolar fluid clearance that can be inhibited by amiloride but not propranolol (47). Interestingly, this effect can be inhibited by an anti-tumor necrosis factor- (TNF-) antibody (47). TNF- is also important in stimulating alveolar fluid clearance after intestinal ischemia-reperfusion (5) and has been shown to be a TNF- receptor-dependent mechanism on the basis of studies of isolated human lung epithelial cells (21). Alveolar fluid clearance can also be stimulated in subacute models of hyperoxic lung injury (23) by upregulating sodium transport in the lung (58). These experimental results suggest that, when the integrity and function of the alveolar epithelium are preserved, alveolar liquid clearance can be stimulated even in presence of lung interstitial or mild alveolar edema.

In contrast, in the presence of more severe lung injury, with a greater increase in the wet-to-dry ratio (>6 g H₂O/g dry lung), there is often a decrease in alveolar liquid clearance. For example, when severe septic shock was induced in animals by infusion of Pseudomonas, a heterogeneous response was observed (46). In one group of sheep, the alveolar epithelial barrier remained resistant to injury, with restriction of the edema fluid to the pulmonary interstitium. In this group, alveolar liquid clearance remained normal or even slightly increased (46). In the other group, a more severe systemic and pulmonary endothelial injury was associated with alveolar flooding, a marked increase in epithelial permeability to protein, and an inability to transport fluid from the air spaces of the lung (46). This relationship between an increase in lung epithelial permeability and altered alveolar and lung fluid clearance has also been observed in other models. Acid aspiration-induced lung injury (26, 38) and smoke inhalation injury (27) both cause an increase in alveolar epithelial permeability to protein and a 40-50% reduction in net alveolar fluid clearance (Fig. 2). Ventilator-associated lung injury that also decreases lung fluid clearance is associated with an increased protein permeability of the alveolar epithelium to protein and small solutes (18, 28). There are also preliminary data indicating that, in the transplanted lung, there is an increase in alveolar epithelial permeability to protein and a decrease in alveolar fluid clearance (56). Although the inability to remove excess fluid from the air spaces in these models is probably related in part to an increase in paracellular permeability secondary to the injury to the alveolar epithelial barrier, these injuries may also diminish the capacity of the alveolar epithelial cells to transport ions normally. In some models, there is an altered function or expression of the transport pathways involved in sodium transport (28). Because alteration in alveolar liquid clearance can occur in different models of lung injury, these data suggest that multiple mechanisms could lead to a dysfunction of the alveolar epithelium. Furthermore, on the basis of the heterogeneous responses observed in alveolar liquid clearance in the septic shock model, we can hypothesize that the variability in the systemic response to injury could also be a determinant of alveolar liquid clearance.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Alveolar fluid clearance (% of instilled volume) was measured by the increase in the unlabeled alveolar protein concentration in rabbits with intact pulmonary blood flow (A) and in rabbits without vascular filtration (B). The 50% decrease in alveolar fluid clearance occurs in the presence or absence of pulmonary blood flow and could be inhibited by anti-interleukin-8 (Anti-IL-8) pretreatment. Values are means ± SD. *P < 0.05 from saline-instilled rabbits. [From Modelska et al. (38).]

Although the primary barrier to alveolar fluid clearance is the lung epithelial barrier, it is important to identify the major pathways for clearance of fluid once it has reached the interstitial space. Although some of the clearance occurs through the lymphatic system, the most important pathway for removal of excess lung interstitial fluid is the pulmonary circulation (1). Under some pathological conditions, the pleural space may serve as an important clearance pathway for pulmonary edema (32). Although the pulmonary circulation is important for the clearance of interstitial edema, it is not essential. Lung fluid clearance is preserved in the absence of pulmonary blood flow (25) and a significant reduction in bronchopulmonary anastomotic flow (51). More recently, it has been shown that alveolar fluid clearance is present in isolated nonperfused lungs (51, 24), including the human lung (48, 50).

Although the pulmonary circulation is not essential for alveolar fluid clearance, could changes in perfusion pressure influence alveolar fluid clearance? Recent data suggest that alveolar fluid clearance can be modulated by perfusion pressure. Campbell et al. (6) reported that alveolar fluid clearance persisted in the presence of a moderate increase in left atrial pressure. However, -adrenergic stimulation could not stimulate alveolar fluid clearance when there was a persistent elevation of the left atrial pressure (6). This inhibitory effect may have been attributed to atrial natriuretic factor (ANF), which has been shown to modulate the Na⁺-K⁺-ATPase in vitro (6) and to decrease sodium transport in the lung (41). However, other mechanisms besides ANF could also be involved. For example, Saldias et al. (52) reported that increased left atrial pressure in isolated perfused rat lungs decreased alveolar fluid clearance. Because there are no functional lymphatics in this model, it is possible that increased interstitial volume and pressure could slow alveolar fluid clearance. In fact, recently, Fukuda et al. (20) reported that an increase in interstitial volume can limit alveolar fluid clearance in in situ mouse lungs with no ventilation and no perfusion. Thus, although alveolar fluid clearance depends primarily on active ion transport in the distal lung epithelium, it can be modulated by the pulmonary circulation and the interstitial pressure generated by interstitial edema fluid.

These observations raise questions regarding the importance of the techniques used to measured alveolar liquid clearance in the injured lung. Most of the experimental data on alveolar liquid clearance in injured lung is obtained by measuring the change of proteins concentration in a test solution instilled in the injured lung. This method, which has been validated first in normal lung (3), has also been validated in multiple models of lung injury (33). Although the presence of an increase in epithelial permeability can confound this measurement, the use of labeled vascular and alveolar space proteins can be used to adjust for bidirectional protein movement (47). Furthermore, because the concentration of protein in the instilled solution in the lung is equal to the protein concentration present in the circulation, it is unlikely that an increase in protein concentration in edema fluid can be due to an increase movement of protein from the interstitial space. However, movement of liquid from the interstitial space to the alveoli could potentially dilute the protein in the instilled solution. In these circumstance, it can be difficult to determine whether the decrease in clearance is due to a decreased reabsorption of liquid or an enhanced movement of liquid from the interstitial space. This is why a decrease in alveolar protein concentration cannot be used to calculate net fluid accumulation in the air spaces. A rise in alveolar protein concentration, especially above the level of plasma protein concentration, means there must have been net alveolar fluid clearance. In clinical studies that estimate alveolar edema fluid clearance in patients, investigators have determined whether fluid clearance is intact on the basis of a rise in edema fluid protein concentration and also calculated the rate of fluid clearance over time (34, 61, 62).

	MECHANISMS INVOLVED IN THE UPREGULATION OF ALVEOLAR FLUID CLEARANCE IN THE INJURED LUNG

TOP ABSTRACT INTRODUCTION MODULATION OF LUNG FLUID... MECHANISMS INVOLVED IN THE... TREATMENT STRATEGIES TO... REFERENCES

Several mechanisms may upregulate the fluid transport capacity of the distal pulmonary epithelium, even after moderate-to-severe epithelial injury. The mechanisms involved may be different depending on the type of injury. Several studies have determined how alveolar fluid clearance may increase in hyperoxic lung injury. Two different models of hyperoxic lung injury have been studied: one of them is subacute, in which rats were exposed to 85% O₂ for 7 days, and the second is acute, in which the rats were exposed to 100% O₂ for 60 h (10). In the acute model of hyperoxic lung injury, Nici et al. (40) reported an increased expression of Na⁺-K⁺-ATPase mRNA and protein in O₂-exposed animals compared with controls. The same group has shown that this adaptive response of the Na⁺-K⁺-ATPase in alveolar type II cells is associated with an enhanced active sodium transport in vivo (7). However, work by Olivera et al. (43) demonstrated a decreased active sodium transport at the end of the hyperoxic exposure and an increase 7 days postexposure. In the subacute model of hyperoxia (85% O₂ × 7 days), Sznajder et al. (58) showed increased active sodium transport and alveolar fluid clearance at the end of the exposure period. Furthermore, this group of investigators reported that the enhanced sodium transport was associated with an increased Na⁺-K⁺ -ATPase activity in the alveolar type II cells at the end of hyperoxic exposure (42). It has also been reported that lung Na⁺-K⁺-ATPase activity was increased during recovery from thiourea-induced pulmonary edema, another type of oxidant-induced lung injury (65). The activity of the enzyme began to increase 4 h after induction of edema, with maximal activation being reached at 12 h (65). The increased activity was also associated with an elevated quantity of the enzyme in the lung and the alveolar type II cells at 12 h (65). Oxidant stimulation may modulate the activity or expression of these transport proteins in this model of lung injury (7, 64).

In experimental models of sepsis and hemorrhagic shock, the increase in alveolar fluid clearance is catecholamine dependent. Although catecholamines can stimulate the activity of either sodium channels through increased cAMP (31) or stimulate the activity (57) or membrane insertion (4) of the Na⁺-K⁺-ATPase, it has been recently reported that catecholamines, mainly -adrenergic agonists, can also modulate expression of the epithelial sodium channel (ENaC) as well as expression of the Na⁺-K⁺-ATPase (35). There is also interesting recent data showing that cAMP stimulation of lung epithelial fluid clearance may depend on chloride transport by the cystic fibrosis transmembrane conductance regulator (15). Glucocorticoids are other potent stress hormones that can modulate expression of the sodium channel and the Na⁺-K⁺- ATPase (11) and stimulate alveolar fluid clearance (17). Thus modulation of expression of sodium transport components by stress hormones is another mechanism by which alveolar fluid clearance could increase during acute lung injury.

There are probably other mechanisms by which alveolar fluid clearance can be modulated in lung injury. Recently, Folkesson et al. (16) demonstrated that alveolar fluid clearance can be stimulated in subacute lung injury from bleomycin and that the increased sodium transport occurs in the presence of decreased expression and function of amiloride-sensitive sodium channels (probably primary ENaC) in alveolar type II cells. The results indicated that the increased number of alveolar type II cells was primarily responsible for the twofold increase in alveolar fluid clearance (Fig. 3). Stimulation of alveolar fluid clearance by TNF- in lung injury secondary to pneumonia (47) or ischemia-reperfusion injury to the intestine (5) are other injuries in which activation of sodium transport is not related directly to an endogenous stress hormone. In fact, it has been suggested recently that TNF- might directly stimulate the sodium channel in alveolar epithelial cells (22). Overall, there are multiple mechanisms that can upregulate alveolar fluid clearance in lung injury.

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Schematic diagram illustrating 3 potential alveolar environments that can be encountered in an injured lung. In regions where there is significant injury to the epithelium, epithelial repair will be needed before stimulation of ion transport can be achieved. In regions where normal epithelial function is preserved, pharmacological stimulation is possible. Finally, in regions where a proliferative response is present, there could be an intrinsic enhanced clearance. [From Berthiaume et al. (2).]

	TREATMENT STRATEGIES TO INCREASE ALVEOLAR FLUID CLEARANCE IN THE INJURED LUNG

TOP ABSTRACT INTRODUCTION MODULATION OF LUNG FLUID... MECHANISMS INVOLVED IN THE... TREATMENT STRATEGIES TO... REFERENCES

Even if alveolar fluid clearance is enhanced in some patients with lung injury, more than 50% of patients with lung injury have impaired fluid clearance (62). Because patients with maximal alveolar fluid clearance have a lower mortality (62), we can hypothesize that improving alveolar liquid clearance in patients with defective clearance could potentially modify the evolution of lung injury. Multiple pharmacological or molecular tools have been used to stimulate alveolar fluid clearance.

The best studied agents are cAMP agonists, i.e., -adrenergic receptor agonists. -Adrenergic agonist therapy enhances alveolar fluid clearance in multiple models of lung injury, including hyperoxic lung injury (23, 49, 53) and ventilator-induced lung injury (54), and can even stimulate alveolar fluid clearance when it is administered as an aerosol (19). Endogenous catecholamines are also essential for the stimulation of alveolar fluid clearance in shock (sepsis or hemorrhagic) (45, 37). However, the potential use of -adrenergic agonists as therapy to enhance the resolution of pulmonary edema might be limited by multiple factors. First, prolonged stimulation of -adrenergic receptors with endogenous catecholamines could desensitize the -receptors and prevent their stimulation with exogenous catecholamines. However, because the enhanced alveolar liquid induced by -adrenergic agonist returns rapidly to a normal level after circulating catecholamine levels return to baseline (8), this limitation seems unlikely except in patients with prolonged, unstable shock. Other circulating factors could, however, limit the action of the -adrenergic agonists. For example, in the presence of left atrial hypertension, it is possible that ANF could inhibit the stimulatory effect of the -adrenergic agonist on alveolar fluid clearance (6). In prolonged hemorrhagic shock and resuscitation, cAMP agonists may not stimulate alveolar fluid clearance (36), because of oxidant-mediated injury (36, 44). Other agents might then be used to stimulate alveolar fluid clearance. Vasoactive agents, such as dobutamine or dopamine, have been shown to stimulate alveolar fluid clearance (9). Growth factors, such as epidermal growth factor, transforming growth factor-, and keratinocyte growth factor, can also stimulate vectorial sodium transport (9). Because keratinocyte growth factor has also been shown to stimulate alveolar fluid clearance and to have additive effects to -adrenergic agonists (60) and even to have a protective effect in some models of lung injury (30, 59), it may be an excellent alternative to -adrenergic agonist therapy. However, it still remains to be shown whether these agents can modulate alveolar fluid clearance and lung injury if they are administered after the insult. One final approach that could be used to manipulate alveolar fluid clearance is gene therapy. Overexpression of the -subunit of the Na⁺-K⁺-ATPase (14) or the ₂-adrenergic receptor (12) in the lung can stimulate alveolar fluid clearance and improve survival of animals with hyperoxic lung injury (13). Overexpression of both the - and -subunits of Na⁺-K⁺-ATPase can also decrease edema formation in a model of thiourea-induced lung injury (55). However, more data will be needed to determine whether this therapeutic approach is effective only as a pretreatment or whether it could in fact modulate alveolar fluid clearance and lung injury once it is used as a therapeutic agent in lung injury. The results suggest that novel therapeutic strategies might be designed to accelerate alveolar fluid clearance and the resolution of pulmonary edema in the future. It is, however, likely that more than one treatment strategy will be necessary because the injury process is heterogeneous from one patient to the others or even in the same patient. Pharmacological stimulation of alveolar liquid clearance would be possible in zones where the epithelium is intact. In other areas of the lung, where significant injury of the lung has occurred, epithelial repair will be necessary before stimulation of alveolar liquid clearance can be possible (Fig. 3).

In conclusion, over the past two decades, major progress has been made in understanding the mechanisms responsible for alveolar edema clearance. The challenge in the future will be to identify specific cellular mechanisms involved in sodium, chloride, and net fluid transport and to develop pharmacological treatments that will be useful to enhance alveolar fluid clearance and resolution of clinical pulmonary edema.