Acute Lung Injury: A Clinical and Molecular Review

The first description of ALI/ARDS, albeit not termed as such, was probably from a physician serving with the Canadian Forces in 1915⁵ in describing a soldier with “shock lung” due to exposure to poison gas.⁶  In 1967, Ashbaugh et al¹  coined the now well-known term acute respiratory distress syndrome to describe a clinical syndrome characterized by “an acute onset of tachypnea, hypoxaemia, and loss of compliance after a variety of stimuli.” In 1994, the American-European Consensus Conferences (AECC) on ARDS published a statement on definitions, mechanisms, relevant outcomes, and clinical trial coordination⁷  that attempted to delineate and guide treatment; however, there remained some confusion due to overlapping criteria applied to the definition of ALI and ARDS, specifically in relation to hypoxia levels and imaging interpretation. In 2012, the Berlin definition of ARDS was published, emphasizing 3 categories of ARDS—mild, moderate, and severe—based on the degree of hypoxemia.⁸  It provided superior predictive validity for mortality in comparison to the AECC definitions.⁸  According to the Berlin definition, ARDS is defined by an acute hypoxemia, a ratio of partial pressure of atrial oxygen to the fraction of inspired oxygen less than or equal to 300 mm Hg on positive end-expiratory pressure greater than or equal to 5 cm H₂O, together with bilateral infiltration on radiology that is not otherwise explained fully by fluid overload or cardiac failure.⁸  In 2015, Riviello et al,⁹  concerned that the Berlin definition underestimates ARDS incidence in low-income countries, suggested a further definition, termed the Kigali modification, attempting to define ARDS without access to imaging or advanced testing modalities. ARDS incidence may indeed be underestimated; it varies widely depending on the definition used and the strictness of adherence to the current Berlin criteria.

The causes of ALI/ARDS are legion, and include, but are not limited to, infection, collagen vascular diseases, drug effects, ingestants, inhalants, shock, acute eosinophilic pneumonia, immunologically mediated pulmonary hemorrhage and vasculitis, and radiation pneumonitis. The term acute interstitial pneumonia, also termed Hamman-Rich syndrome, describes cases of diffuse alveolar damage (DAD) that are idiopathic. ARDS mortality has remained at approximately 40% for the past 2 decades.¹⁰ 

ARDS is typically a diagnosis based on clinical and radiologic features. Radiologically, the early exudative phase shows bilateral and patchy ground-glass densities, corresponding to interstitial edema and hyaline membranes. The geographic distribution of the patchy ground-glass densities, together with areas of lobular sparing and lower lobe consolidation, serve as radiologic hallmarks.¹¹  Radiologic features of disease advancement to the proliferative and fibrotic phases are characterized by traction bronchiolectasis or bronchiectasis within areas of increased attenuation on high-resolution computerized tomography scans.¹² 

Because most ARDS cases are diagnosed clinicoradiographically, biopsies are uncommonly required for diagnosis. Biopsies typically are performed in cases where presentation is not straightforward, specific infection is being considered, or therapeutic response is disappointing.

Diffuse alveolar damage, the histologic counterpart of ALI/ARDS, occurs as the culmination of the cascading process of ALI, developing from epithelial barrier dysfunction, endothelial dysfunction, and resultant pulmonary edema. As described by Katzenstein,¹³  DAD can be divided into 3 phases: acute/early or exudative, organizing or proliferative, and late-resolving or fibrotic phase. The acute phase is characterized by distinctive hyaline membranes lining alveolar spaces (Figure 1). Edema is frequently identified and acute alveolar hemorrhage may be present. Endothelial cells and pneumocytes undergo necrosis. The hyaline membranes begin to organize as DAD continues into the organizing phase, and granulation tissue develops in the alveolar spaces. Type II pneumocytes demonstrate marked reactivity and become hyperplastic near the end of the early phase. These features can continue through the proliferative phase (Figure 2). Squamous metaplasia, occasionally exuberant enough to mimic carcinoma, may arise (Figure 3). As the organizing phase progresses, granulation tissue is incorporated into the alveolar septa, leading to organizing fibrosis (Figure 4). In the late-resolving or fibrotic phase there is dense collagen fibrosis and hyalinization of the alveolar walls.^13–15  The stages occur in a continuum rather than in a strict chronologic fashion. Continuing ALI may occur, with the biopsy specimen showing acute and organizing phases simultaneously. For patients who survive ARDS, many cases resolve with minimal lung damage; however, patients may develop varying degrees of end-stage lung change.

An additional histologic pattern of ALI/ARDS, acute fibrinous and organizing pneumonia (AFOP), is characterized by patchy areas of eosinophilic fibrin aggregates or balls within intraalveolar spaces. AFOP can be seen as a predominant pattern or a component of a DAD. Diagnosis should be based on the predominant feature. Cases demonstrating a primarily AFOP pattern with minimal or no hyaline membranes have shown a worse prognosis.¹⁶ 

Much research is being performed in an attempt to elucidate the intricacies of the process of ALI development and progression of ALI to ARDS.¹⁷  One of the emerging concepts in immunology as applied to the development of ARDS is pattern recognition receptors (PRRs)—critical components of the innate immune system that serve as a “first line of defense.” Pattern recognition receptors recognize 2 categories of ligands: nonendogenous pathogen-associated molecular patterns (PAMPs), and endogenous danger (or damage)–associated molecular patterns (DAMPs). Pattern recognition receptors can initiate inflammatory signaling cascades and the release of proinflammatory cytokines such as tumor necrosis factor α (TNF-α), interleukin (IL)–1β, and IL-8; stimulate autophagy or apoptosis; and induce production of antibacterial molecules.¹⁸  Ten functional Toll-like receptors (TLRs), transmembrane PRRs that are highly conserved in vertebrates, have been identified in human beings.^19,20  Nucleotide-binding oligomerization domain–like receptors (NLRs) are cytosolic PRRs.²¹  The NLRs are expressed on both white blood cells and epithelial cells, both of which are in contact with invading microorganisms. Examples of DAMPs include histones and high-mobility group box 1, as contrasted with PAMPs, which include lipopolysaccharides and lipoteichoic acid. In murine studies, TLR signaling pathways have been shown to be involved not only in ARDS development, but also in its resolution.²²  For example, degradation products of hyaluronan, an extracellular matrix glucosaminoglycan produced after tissue injury, interact with TLR4 and TLR2 to induce inflammatory responses leading the ALI.²²  It has also been shown that overexpression of high-molecular mass hyaluronan in lung epithelial cells is protective against lung injury and apoptosis.²²  Other important products of cellular injury are mitochondrial DAMPs, which include cardiolipin, formyl peptides, and mitochondrial DNA.^23,24  Mitochondrial DAMPs have been shown to be involved with ARDS in several ways, the foremost being their ability to activate neutrophils by inducing production of the powerful neutrophil attractor and activator IL-8.²⁵  Interleukin 8 forms complexes with anti–IL-8 neutralizing autoantibodies, which in turn interact with FcγRII receptors that then affect neutrophil apoptosis.²⁵  In addition, increased plasma levels of mitochondrial DAMPs are associated with higher mortality rates.²⁵  Studies²⁶  have shown that circulating mitochondrial DAMPs can cause neutrophil-mediated tissue damage.

ALI/ARDS severity and possibly development is greatly influenced by neutrophil migration into the lungs in response to activated alveolar macrophages.²⁷  In the lung, activated neutrophils produce numerous cytotoxic substances, including granular enzymes, reactive oxygen species, bioactive lipids, various proinflammatory cytokines, and neutrophil extracellular traps (NETs), which trap pathogens in the extracellular space through NETosis.²⁸  NETs are cleared slowly from the lungs owing to the lower levels of surfactant proteins A and D, which are required for clearing NETs, in patients with ARDS.²⁸  Regardless of the inciting molecule or molecules, complement activation ensues and is critical for ALI/ARDS development.²⁹ 

The 2012 Berlin definition of ALI/ARDS did not recommend any biomarkers, and currently there are few markers actively used in clinical practice.³⁰  Brain natriuretic peptide is used to distinguish hydrostatic pulmonary edema from edema due to ARDS, however, its use is controversial.^30,31  Procalcitonin has been suggested for use in differentiating bacterial pneumonia from ARDS, but its relatively low sensitivity (approximately 70%) for bacterial pneumonia—itself an etiology of ARDS—makes its use problematic.³²  There are no currently clinically useful molecular biomarkers for predicting ARDS severity; however, the PaO₂/FiO₂ ratio (partial pressure of alveolar oxygen/fraction of inspired oxygen) is used clinically to predict severity.⁸

Several potential biomarkers exist in plasma or bronchoalveolar lavage (BAL) specimens, many of which are currently being actively researched in animal studies; however, these have generally failed to show significant results in early human clinical trials. Among the proinflammatory cytokines, TNF-α, IL-1β, IL-6, IL-8, and IL-18 are the most promising potential molecular biomarkers for predicting morbidity and mortality (Table).

Interleukin 18 has promise as a molecular biomarker for prognosis prediction. Dolinay et al,³³  studying 217 patients with trauma- or sepsis-induced ARDS, found that increased IL-18 levels correlated with increased in-hospital mortality. Interleukin 18 and IL-1β are activated by caspase-1, which cleaves them from their proforms. Caspase-1, a proinflammatory cytokine belonging to the IL-1 cytokine family, is itself activated by an inflammasome, an intracellular macromolecular complex.^34,35  Inflammasomes, of which many are now recognized, respond to TLRs, bacterial- and viral-derived molecules, particle irritants, and reactive oxygen species from neutrophils or mitochondria, gram-negative bacteria, and DNA.^36–44  The caspase-1 pathway is integral in the acute inflammatory response.

Interleukin 8 in particular deserves note, especially in regard to its association with neutrophils. Interleukin 8 concentrations in BAL fluid are higher in patients with ARDS and, as previously mentioned, attract and activate neutrophils.^45,46  Interleukin 8 forms complexes—demonstrated in BAL fluids of patients with ARDS—with anti–IL-8 autoantibodies, which in turn affect interactions of IL-8 with neutrophils.⁴⁶  These complexes have been shown to activate and chemoattract neutrophils as well as prolong neutrophil life by inhibiting apoptosis through engagement of FcγRIIa.^45,46  Also, clinical disease activity is correlated with anti–IL-8:IL-8 concentrations in BAL fluids.⁴⁷  There is a significant correlation between the onset of ARDS and the concentration of anti–IL-8:IL-8 complexes.^46,47  The concentration may also serve as a marker of mortality, as complex concentrations were found to be higher on day 1 of ARDS onset in patients who died of disease, as compared with those who survived.^48,49  It is likely that the high levels of these complexes in the lung overwhelm removal mechanisms, and further contribute to neutrophil concentrations and activation, thus contributing to the ongoing damage occurring in ALI/ARDS.

A recent meta-analysis by Terpstra et al⁵⁰  (reviewing 54 studies comprising 3753 patients) of the performance of 20 potential plasma biomarkers found that IL-4, IL-2, angiopoietin-2, Krebs von den Lungen-6, and protein C (decreased levels) showed the strongest associations with mortality in patients with ARDS.

Vascular endothelial growth factor (VEGF) concentrations have been studied in lung endothelial lining fluid in a small cohort of patients and are inversely correlated with lung injury; it is hypothesized that upregulation of VEGF production contributes to decreasing inflammation and is correlated with improved outcomes.⁵¹  A study of 32 patients with ARDS measured the levels of alveolar type II epithelial cell growth factors, hepatocyte growth factor, and keratinocyte growth factor, and found that keratinocyte growth factor in BAL fluids of patients with ARDS correlated with a poorer prognosis.⁵²  However, hepatocyte growth factor, while elevated in patients who did poorly, was identified in most BAL specimens, including control samples.⁵² 

Treatment is generally supportive, with emphasis on treating the underlying cause of disease. One clinical intervention to have shown benefit was observed in a phase III trial, demonstrating a modest but significant decrease in mortality with lower tidal volumes (initially 6 mL/kg of predicted body weight with a plateau pressure of 30 cm H₂O or less).²  The lower volumes also served to increase the number of ventilator-free days.² 

Current presumed molecular pathways leading to ALI/ARDS initially involve complement activation potentially induced by a variety of inflammatory insults (such as sepsis, pneumonia, traumatic injury, blood transfusions, or mitochondrial DAMPs). Alveolar macrophages are activated via TLR and NLR signaling pathways that lead to further macrophage and circulating neutrophil recruitment. Neutrophils accumulate in the lungs and release proinflammatory cytokines and NETs. The lung epithelium, specifically type II alveolar cells, is damaged by these cells and their products, with resultant disruption of the alveolar-capillary interface and increased pulmonary microvascular permeability. The resulting pulmonary edema impairs gas exchange, which in some cases may lead to respiratory failure. And ARDS is not an isolated pulmonary process; the release of oxidants and proteases from the activated neutrophils and macrophages can also cause distant tissue and organ damage and dysfunction.⁵³ 

While a significant amount of research has been done in an attempt to understand ALI/ARDS pathophysiology, the development of viable predictive molecular biomarkers and molecular-based therapies remains elusive. Yet, a significant amount of progress has been made in the elucidation of ARDS pathophysiology and in predicting patient response. As Kress⁵⁴  states, we still measure mortality as our main outcome in clinical trials in ARDS, but now it is time to begin focusing on function and quality of life in survivors of ARDS.