Pediatric acute respiratory distress syndrome - current views.

Acute respiratory distress syndrome (ARDS) mainly involves acute respiratory failure. In addition to this affected patients feel progressive arterial hypoxemia, dyspnea, and a marked increase in the work of breathing. The only clinical solution for the above pathological state is ventilation. Mechanical ventilation is necessary to support life in ARDs but it itself worsen lung injury and the term is known clinically as 'ventilation induced lung injury' (VILI). At the cellular level, respiratory epithelial cells are subjected to cyclic stretch, i.e. repeated cycles of positive and negative strain, during normal tidal ventilation. In aerated areas of diseased lungs, or even normal lungs subjected to injurious positive pressure mechanical ventilation, the cells are at risk of being over distended, and worsening injury by disrupting the alveolar epithelial barrier. Further, hypercapnic acidosis (HCA) in itself confers protection from stretch injury, potentially via a mechanisms involving inhibition of nuclear factor κB (NF-κB), a transcription factor central to inflammation, injury and repair. Mesenchymal stem cells are the latest in the field and are being investigated as a possible therapy for ARDS.

Introduction

Significant morbidity and mortality has been confirmed due to acute respiratory distress syndrome (ARDS) worldwide (1). It is characterized by severe inflammation of the lung parenchyma with associated protein-rich pulmonary oedema leading to hypoxaemia refractory to supplemental oxygen, and respiratory failure (2). Most patients require mechanical ventilation (MeV) as a life-saving measure in an intensive care setting. Many patients frequently develop subsequent multi-organ failure (3). The mortality burden of ARDS approaches that of breast cancer and HIV (4), and could be as high as 60%. There is currently no known treatment, and management remains supportive.

Mechanical ventilation role

Mechanical ventilation has been the mainstay of life-supporting treatment in intensive care medicine (5). One of the first mechanical ventilators was in the form of an iron lung and a metal cylinder (5). Negative pressure applied in the chamber using a vacuum pump, was used to expand the chest. However, this mode is a difficult procedure with significant discomfort to the patient involved.

The modern positive pressure ventilators reduced mortality from 85 to 20% (6). In positive pressure ventilation, the airway pressure is applied at the patient's airway via an endotracheal or tracheostomy tube. This permits gas flow into the lungs until the end of the ‘breath’ delivered by the machine. Elastic recoil of the chest occurs when the airway pressure is zero, leading to passive exhalation.

Role in sustaining life in ARDS patients

Mechanical ventilation is crucial for patients with ARDS and other forms of respiratory failure (7). Traditionally, most patients, including those with ARDS that required mechanical ventilation were ventilated with tidal volumes of more than 9 ml/kg. This in turn helped in maintenance of normal oxygen and carbon dioxide levels along with minimal atelectasis (8,9). To achieve these goals, grossly high peak inspiratory pressures of up to 50 cm H2O were tolerated quite often in the absence of obvious complications of barotrauma, such as the development of pneumothoraces.

Ventilator-induced lung injury (VILI)

It was discovered that ventilating rat lungs with peak pressures of up to 45 cm H2O resulted in faster onset and more severe pulmonary oedema than when ventilating at pressures of 30 cm H2O, and that positive end-expiratory pressure (PEEP) could alleviate that damage (10). VILI can cause or worsen ARDS, and is in fact, a form of ARDS both clinically and pathologically. Animal lungs injured by mechanical ventilation show patterns of atelectasis, severe congestion and enlargement due to oedema. Contemporary ventilation strategies entail use of lower tidal volumes (6–8 ml/kg) leading to minimal lung over-distension and further injury. ARDS typically does not have a uniform distribution, and affects some segments of lung more severely than others. CT-guided studies have demonstrated particular segments of lung with relatively normal elastance (baby lung) (11,12). Therefore even ‘normal’ tidal volumes during mechanical ventilation could damage these smaller segments of aerated lung tissue. The Vt/baby lung ratio is more relevant than the Vt/kg ratio for VILI in this context.

Mechanisms of VILI

It was later appreciated that simple barotraumas, i.e. gross air leaks induced by large trans-pulmonary pressures, was not the only way that ventilators could damage lungs. Important mechanistic insights have emerged from extensive research to elucidate the potential mechanisms by which certain ventilator settings may cause specific biological damage. The concept of volutrauma is of the view that lung distention is the cause behind lung injury (13). Atelectrauma is a term used to describe repeated opening and reopening of atelectatic areas leading to high shear stress and subsequent necrosis, classically with high ventilator distending pressures and no PEEP (14). Biotrauma, a relatively recent concept, states that excessive release of proinflammatory mediators and immune system hyperstimulation is the primary mechanism by which injurious ventilation damages the lungs (15).

Ventilator-induced lung inflammation

Multiple animal models of VILI have demonstrated that injurious ventilation in the form of volutrauma or atelectrauma results in significant leukocyte sequestration and lung damage (16). Cytokines are an important group of mediators in this process. These are low-molecular-weight glycoproteins, produced by a number of cells, which communicate with cell-surface receptors to activate or inhibit the inflammatory cascade. Pro-inflammatory cytokines play an important role in immune cell recruitment and in the activation of macrophages and polymorphonuclear cells. Mechanical ventilation has the ability to release cytokines in lung cells by a number of mechanisms:

i) Mechanotransduction: the conversion of physical forces on the cell membrane/receptors into promotion of intracellular signal transduction pathways leading to increased synthesis and secretion of cytokines (17). ii) Cell necrosis may result from injurious stretch, leading to secretion of preformed cytokines, which then modulate the production of more cytokines (18). iii) Vascular endothelial cells may exert an increased inflammatory response as a consequence of increased pressure in the pulmonary vasculature by MV (19).

The localized inflammatory response resulting from these mechanisms could rapidly become systemic if the endothelial-epithelial barrier is disrupted secondary to necrosis, permitting ‘spillover’ of cytokines, endotoxins and bacteria to spread from alveoli into the systemic circulation. Previous studies have linked the systemic inflammatory response that accompanies aggressive mechanical ventilation with cytokine release from lung cells (20), and consequent translocation of cytokines into the bloodstream (21). This loss of pulmonary containment of inflammation, termed de-compartmentalization, explains the high mortality of patients with ARDS from multi-organ failure. This concept of a systemic effect from an initial lung injury is supported by direct evidence from ex vivo studies (22), in vivo animal studies (23) and indirect supportive data from human studies (24). It was observed in an earlier study that high tidal volumes and zero PEEP caused significant elevation in serum levels of TNF-α compared to rats ventilated at low tidal volumes with or without PEEP and high tidal volumes with PEEP (25). Of note, PEEP is shown to be protective here, as in the high tidal volume group with PEEP, there was no TNF-α response.

Cyclic stretch and transcriptional activation

Mechano-transduction also stimulates gene transcription following mechano-sensing. A growing body of evidence has implicated the mitogen-activated protein kinase (MAPK) pathway as integral in transduction of signals from the mechano-sensory apparatus to the nucleus and driving gene transcription (26). MAPKs are signaling enzymes activated by phosphorylation in response to a range of extracellular stimuli such as cytokines and growth factors (27).

These enzymes are responsible for the transduction of mitogenic and differentiation signals leading to activation of transcription factors (28). ERK1 and 2 are activated by cyclic stretch in alveolar epithelial cells (29), as well as in a variety of other cell types, including cardiac myocytes (30), endothelial cells (31), and vascular smooth muscle cells (32). Many cyclic stretch studies have demonstrated increased alveolar levels of IL-8, which is highly suggestive of increased transcription of this inflammatory cytokine by mechanical stretch (33). Cyclic stretch has also been demonstrated to induce activation of a powerful transcription factor nuclear factor κB (NF-κB).

NF-κB in injury

The NF-κB helps in harmonization of the inflammatory response. It regulates pro-inflammatory transcriptional programs that carry out this response. Its activation leads to the upregulation of adhesion molecules and chemokines in vascular endothelial cells and within the tissue. It is also essential for the production of antimicrobial effector molecules and for the survival of leukocytes in an inflammatory milieu. Although crucial for an intact host defense response, excessive activation of NF-κB leads to exaggerated inflammatory injury of lungs and other organs (34). Mechanical stretch activates the NF-κB pathway in cardiac myocytes (35), vascular endothelial cells (36) and alveolar epithelial cells (37).

The presence of this transcription factor in the pathogenesis of lung injury has been demonstrated in many experimental studies. Upregulation of NF-κB in response to stretch has been demonstrated in many in vitro and ex vivo lung studies (38). Increased activation of NF-κB occurs in alveolar macrophages in established ARDS (39). Using an isolated perfused lung model, Held et al demonstrated activation of NF-κB in response to LPS and ventilation with large tidal volumes (40). NF-κB regulation can take place at any point of its activation and postactivation pathway and probably plays a crucial role in the NF-κB-mediated inflammatory cascade. In ARDS for example, Moine et al (41) showed that NF-κB maintained an activated state in spite of increased IκB levels and reductions in Bcl-3 levels (another member of the IκB family). This implies that basic aberrations in NF-κB-related transcriptional mechanisms are likely central to the generation of the inflammatory cascade, which occurs in patients with sepsis and ARDS.

Mechanisms of protection by low tidal volume ventilation

Modulation of the inflammatory response. Protective ventilation has actually been shown to reduce the inflammatory response at a molecular level, compared to more traditional ventilation. It was found in a recent report that BAL levels of IL-1β, IL-6, IL-8, TNF-α, TNF-αsR55, TNF-αsR75, and IL-1ra decreased over time, as did plasma levels of IL-6. This demonstrated that the protective strategy partially attenuated local and systemic inflammation over time. The low stretch/lung rest arm was concomitant with minimal apoptosis and maximum risks of atelectasis. A study in the recent past noted a significant decrease in interleukin-6 and −8 levels (42). Another study demonstrated reversal of the injurious stretch-induced systemic inflammatory response, in terms of cytokine release, within 6 h of changing to a protective ventilation strategy, but no impact on the local pulmonary response (43). More recently, a study noted significant reduction in the levels of systemic inflammatory cytokines in the low tidal volume group.

Impact of lung ventilatory strategies over time

Mortality in ARDS has undoubtedly decreased over time according to a number of studies (44). Protective ventilation has played a major role in this reduction, but other advancements in critical care, such as better supportive care, have been influential. Petrucci and De Feo conducted an extensive meta-analysis in the form of a Cochrane review, which they updated for the third time in 2013 (45) in which they examine all trials to date that have studied the impact of lung protective ventilator strategies in ARDS. In this most recent update, the main conclusion was the utilization of lower tidal volume ventilation reduces relative risk of death. However, there is scarcity of information regarding conclusion in terms of morbidity and long-term outcomes.

The role of carbon dioxide in modulating lung injury

Injurious effects of hypocapnia. In contrast, traditional approaches to mechanical ventilation frequently resulted in hypocapnia, which can be severely damaging to lungs. A number of studies in the literature have identified hyperventilation and hypocapnia as risk factors for adverse outcomes in ARDS (46). A possible explanation could be vasoconstriction and hemoglobin oxygen dissociation curve left shift. Both hyperventilation and hypocapnia might be independent causes of bronchopulmonary dysplasia in infants (47). Hypocapnia reduces lung compliance (48), impairs functional surfactant production (49) and increases microvascular permeability in tracheal mucosa (46). Trimble et al, in 1971 demonstrated that hypocapnia increased work of breathing, aggravated ventilation/perfusion mismatch, increased airway resistance, increased the A-a O2 gradient and decreased the PaO2 in patients with ARDS (50). Another study found hypocapnia to be directly deleterious in lung injury, and a promoter of ischemia-reperfusion-induced acute lung injury (51).

Rationale for ‘permissive’ hypercapnia to reduce lung stretch

One of the consequences of instituting a low tidal volume ventilatory strategy is a gradually increasing PCO2 level greater than the upper limit of normal, termed hypercapnia. This ‘permissive hypercapnia’ (PaCO2 >45 mmHg) has been adopted as the preferred approach versus normo- or hypocapnic targets in the setting of ALI/ARDS. Having been first demonstrated to be effective in the setting of acute severe asthma (52) in 1984, the strategy was then trialled by Hickling (53) for ARDS in the 1990s. This hypercapnia is accompanied by an acidosis in the acute phase, which is gradually then subjected to renal and tissue buffering. Hickling demonstrated no detrimental effects of this hypercapnic acidosis (HCA) in experimental studies. In subsequent preclinical studies, induced HCA - by means of addition of CO2 to the inspired gases - reduced the severity of acute lung injury (ALI) induced by multiple factors (54,55). HCA has attenuated early lung (56) and systemic (57) sepsis. These preclinical studies have suggested a potential direct benefit of HCA. Moreover, the presence of an HCA at the time of randomization in patients subjected to high tidal volume ventilation was associated with better 28-day survival (58).

Therapeutic hypercapnia - further reducing lung injury?

Encouraged by the beneficial effects of permissive hypercapnia, some groups hypothesized that deliberately increasing the PaCO2, by either initially ventilating at very low tidal volumes, or by inhalation of small fractions of CO2, would also confer benefit in lung injury - termed ‘therapeutic hypercapnia’ (TH). Experimental studies exploring this hypothesis have achieved promising results. Laffey and colleagues demonstrated a protective effect in an in vivo lung reperfusion injury model (54), while Sinclair et al demonstrated improved lung mechanics, gas exchange and injury scores in rabbits ventilated with high tidal volumes and targeted hypercapnia versus rabbits with eucapnic targets (59). De Smet et al ventilated isolated perfused rat lungs for 2 h with low (7 ml/kg) or moderately high (20 ml/kg) Vt and 5% or 20% CO2, with lipopolysaccharide or saline added to the perfusate (60). HCA resulted in improved pulmonary edema, decreased lung stiffness, and markedly decreased levels of TNF-α and IL-6 in the lavage and perfusate. Of note, they found HCA to be beneficial, regardless of the tidal volume, suggesting that HCA has independent protective effects. More recent experimental studies have demonstrated that therapeutic hypercapnia improved lung function and reduced inflammation after one-lung ventilation in lobectomy patients (61). Its beneficial effects extend outside the lung to the heart; TH prevents the adverse effects of sustained exposure to iNO on RV systolic function by limiting IL-1-mediated NOS-2 upregulation and consequent nitration (62). Other studies have demonstrated that TH is neuroprotective via anti-apoptotic mechanisms and improves impaired spatial memory (63). A trial currently in progress will assess the feasibility and safety of targeting mild hypercapnia for 24 h following intensive care unit admission for cardiac arrest patients. It will also provide insight on whether such treatment improves neurological injury biomarkers concentrations, compared with normocapnia (64). The evidence is mounting in favour of the anti-inflammatory merits of CO2.

Mechanisms of action of HCA

One therapeutic mechanism is by the attenuation of free radical injury. Laffey and colleagues demonstrated anti-apoptotic effects and inhibition of nitric oxide-derived oxidant generation (65). The study by Takeshita et al provides strong supportive evidence that HCA attenuates LPS-induced NF-κB activation by suppression of IκBα degradation, which in turn leads to downregulation of ICAM-1 and IL-8 in pulmonary endothelial cells (66). Our own group replicated this finding and in addition demonstrated that HCA attenuated ventilation-induced lung injury independent of injury severity in Sprague-Dawley rats and decreased mechanical stretch-induced epithelial injury and death, via an NF-κB-dependent mechanism (67).

HCA and lung

HCA might also exert potentially deleterious effects in the setting of lung injury and may indeed worsen a pre-existing injury in certain situations. One group demonstrated that HCA worsens protein malfunction by potentiating protein nitration by the free radical, peroxynitrite (68). Other groups, including ours, have shown that HCA impairs bacterial killing in the setting of prolonged lung infection (69), delays plasma membrane resealing, an essential mechanism of cellular repair (70), and inhibits pulmonary epithelial wound healing by decreasing cell migration via an NF-κB-dependent mechanism (71). Dixon also demonstrated that HCA, along with inhibiting stretch-induced cytokine production, also impaired stretch-induced pulmonary surfactant release in rat alveolar type II cells (72). It can be concluded, therefore, that HCA exerts diverse effects, some of which may be beneficial and others harmful, depending on the context. In this manner, it may constitute a ‘double-edged sword’.

MSCs and lung injury

Mesenchymal stem cells (MSCs) have inbuilt ability of self-renewal and could classically differentiate into mesodermal lineages, including chondrocytes, adipocytes, and osteoblasts (73). There is increasing evidence that they might also differentiate into cells of non-mesodermal origin such as lung epithelial cells (74). MSCs offer unique advantage as they allow allogeneic administration due to their low immunogenicity (75). In this manner, the ‘immunomodulatory’ properties of adult stem cells like MSCs might be more successful. The alveolar fluid clearance limits and depletion of surfactant are commonly observed in ALI/ARDS (76). In the above situations, MSCs might restore epithelial and endothelial function, either by secretion of paracrine factors to enhance renewal of these tissues or by differentiation into these cell types. ALI/ARDS is also associated with multi-organ failure, which frequently is the reason for mortality in critically ill patients. MSCs have great therapeutic potential in the above cases (77). MSCs have also potential to secrete disease modifying molecules, which might augment their therapeutic effects. As they home to sites of inflammation following intravenous administration after tissue injury (78), MSCs might also act as vector for genetic therapies (79). The intratracheal administration provides access to both the distal lung epithelium and the pulmonary endothelium for stem cell therapies (80). Moreover, MSCs therapeutic potential for ALI/ARDS has been strongly supported (81). However, MSC are still in the early phases of clinical use in the form of various trials. Some trials showed positive effects and favored clinical use (82). These preliminary studies show potential, but larger trials are required for definitive answers.

Conclusions

To clearly understand the complex interplay between HCA and the NF-κB pathway might allow us to develop more effective strategies to minimize lung injury in ARDS. Generating new insights into the therapeutic potential of MSCs might transform traditional approaches to ARDS management by combining lung repair tactics with attenuation of injury.