Pulmonary Manifestations of COVID-19

Epidemiology of Acute Respiratory Distress Syndrome

One of the largest pre-COVID ARDS studies, the LUNG-SAFE study, conducted in 2014 assessed an in-hospital mortality of 40% with a range of 35% to 46% based on severity of ARDS at the time of diagnosis. ^, The study estimated the prevalence of ARDS among patients hospitalized in intensive care units (ICUs) based on data from 50 countries to be 10.4% and also concluding that ARDS appeared to be underrecognized, undertreated, and underreported. In the United States the annual incidence of ARDS was estimated to be 190,000 cases, with a reported mortality of 38.5%. Several comorbidities, including alcohol abuse, cigarette smoking, polluted air, and hypoalbuminemia, have been recognized as contributing to the severity of the syndrome. There are also significant racial, ethnic, and sex disparities, as well as differences in susceptibility based on population density and socioeconomic status for ARDS morbidity and mortality, with higher rates among the Black and Hispanic populace, males, low-income groups, and people living in high-density population regions. ^{, ,}

Cause of Acute Respiratory Distress Syndrome

The causative risk factors for ARDS, which generally occurs in the setting of pneumonia can be broadly categorized as local and systemic. Respiratory bacterial, viral, fungal, and secondary infections that cause pneumonia, aspiration pneumonia, lung trauma resulting from blunt or penetrating injuries, and inhalation injuries are common local causes; systemic causes include nonpulmonary sepsis, acute pancreatitis, blood or blood product transfusion, drug overdose, major burns, cardiopulmonary bypass, graft-versus-host disease, and high-altitude pulmonary edema (HAPE). ^,

Pathophysiology of Acute Respiratory Distress Syndrome

Histopathology With Alveolar Focus

The characteristic pathological hallmark of ARDS is diffuse alveolar damage which compromises the basic lung physiology of ventilation and perfusion, leading to hypoxemia. ^{, ,} The temporal evolution of the alveolar damage occurs over weeks in three phases. The first or the exudative phase starts in the first week with alveolar epithelial injury affecting both type I and type II pneumocytes. This causes denudation of the alveolar membrane and increased permeability. Eosinophilic deposits result in hyaline membrane formation along the denuded alveolar basement membrane, a classic finding in diffuse alveolar damage along with white blood cell agglutination and formation of platelet-fibrin thrombi. Alveolar fluid reabsorption by ion transport, a normal function of alveolar cells, is affected resulting in accumulation of exudative fluid in the alveolar cavity. Macrophages residing in the alveolus secrete proinflammatory factors such as interleukin-8 (IL-8) and monocyte chemoattractant protein-1, which attract neutrophils and other macrophages into the lung, aggravating the lung injury. Concomitant endothelial damage to the capillaries lining the alveolus results in the disruption of endothelial barrier leading to interstitial flooding, with the excess fluid finding its way to the alveolar space and causing intraalveolar flooding and exacerbation of edema. The nature of insult to the endothelial cells is not well understood but apoptosis and pyroptosis have been proposed as plausible mechanisms. ^, Endothelial cell activation results in release of angiopoietin-2, leading to formation of neutrophil-platelet aggregates that pass through the weakened endothelial barrier and infiltrate the alveolar space. The neutrophil-platelet aggregates manifest thromboinflammatory mechanisms such as neutrophil extracellular traps (NETs), which are implicated in both capillary endothelial and alveolar epithelial barrier perturbations. ^, Other causative factors responsible for endothelial wall damage are pathogens and circulating toxins, alveolar macrophage factors, and proinflammatory cytokines and chemokines. ^, The cascade of events set in motion by epithelial and endothelial injury causing alveolar flooding with exudate results in surfactant dysfunction, which promotes atelectasis and biophysical injury to the lung.

The second phase, called the proliferative phase, sets in motion the repair mechanisms to tackle the intraalveolar and interstitial exudative edema of phase 1. Beginning in the second week, the proliferative phase lasts for about 2 weeks. The key events of this phase include interstitial fibroblast proliferation with secretion of epithelial growth factors, alveolar type II cell hyperplasia with differentiation into alveolar type I cells, phagocytosis of apoptotic neutrophils, and reestablishment of junctions between epithelial cells with restoration of the integrity of the epithelial barrier. Epithelial barrier repair enables reabsorption of edematous fluid into the interstitial space. Fibroblasts transform the exudate to cellular granulation tissue with matrix morphology, which eventually turns into fibrous tissue with the deposition of collagen. Likewise, in the lung capillaries endothelial cell proliferation sets in with repair of endothelial disruption and restoration of the endothelial barrier function. The pathogenetic mechanisms of the three phases of ARDS are enumerated in Figs. 5.1 and 5.2 , which illustrate the pathological process of the injured alveolus in the exudative phase and the repair mechanisms of the proliferative phase with the healthy alveolus as reference.

The third or fibrotic phase necessitates mechanical ventilation and is typically seen in patients surviving beyond 3 to 4 weeks from the onset of ARDS. This phase is marked by significant interstitial and intraalveolar fibrosis with lung remodeling by thick collagenous connective tissue. ^, The process of fibrosis acts as a nonspecific response to the lung injury and resembles the lung tissue reorganization that occurs in typical organizing pneumonia.

Additional pathological features of ARDS are summarized in Table 5.6 and salient histopathological changes observed in the temporal evolution of diffuse alveolar damage are shown in Table 5.7 . After exploration of the pathophysiology of ARDS, the chapter discusses alterations in respiratory mechanics. To formulate a rational approach to the management of ARDS it is necessary to understand the changes in respiratory mechanics in ARDS.

Table 5.6

Acute Respiratory Distress Syndrome (ARDS) Salient Pathophysiological Mechanisms

ARDS
Pathophysiological Characteristics	Description
Diffuse alveolar damage (DAD)	The predominant pathological condition is epithelial injury that involves types I and II pneumocytes and dissociation of junctions between these alveolar cells resulting in increased permeability. Neutrophil migration resulting from endothelial injury leads to alveolar epithelial injury by the perturbation of intercellular junctions and causing enhanced epithelial permeability. See Table 5.7 for a list of additional histopathological details of DAD and its three phases.
Impaired alveolar fluid clearance	Alveolar fluid reabsorption is performed by types I and II alveolar pneumocytes by active ion transport into the interstitium from which clearance occurs by lymphatic channels. Because of alveolar cell damage, permeability is lost, affecting ion transport resulting in accumulation of exudative fluid in alveolar spaces.
Ventilation-perfusion abnormalities (gas exchange anomalies)	Type II alveolar cells secrete surfactant, which plays a key role in gas exchange in the lungs by keeping the alveoli open based on its capacity to lower surface tension. The loss of surfactant causes ventilation-perfusion mismatch and impairs carbon dioxide elimination.
Reduction in lung compliance	Increased permeability, loss of surfactant resulting from damage to type II pneumocytes and accumulation of fluid in the alveolus leads to reduction in lung volume available for gas exchange. This in turn causes increased dead space in the lungs and lowering of lung compliance.
Increased pulmonary vascular resistance	Impaired carbon dioxide elimination causes hypercapnia leading to pulmonary vasoconstriction. This, along with thromboembolism and interstitial edema, results in increased pulmonary vascular resistance and pulmonary hypertension.
Endothelial injury	Endothelial cell damage results in leakage of fluid across the capillary barrier into the interstitial space. This results in an edematous interstitium with fluid moving into the alveoli through the damaged epithelial barrier.
Thromboembolism	Postmortem studies of lungs from patients dying of ARDS have consistently demonstrated thromboembolic lesions. Macrothrombi and microthrombi have been extensively observed in arterioles and capillaries of the lung. Two types of microthrombi have been reported: (1) composed of hyaline platelet-fibrin seen in capillaries and arterioles and (2) laminated fibrin clots visualized in preacinar and intraacinar arteries.
Ventilatory support injury	See Table 5.8

Data from Matthay MA, Zemans RL, Zimmerman GA, et al. Acute respiratory distress syndrome. Nat Rev Dis Primers. 2019;5(1):1-22; Tomashefski JF Jr. Pulmonary pathology of acute respiratory distress syndrome. Clin Chest Med. 2000;21(3):435-466; Swenson KE, Swenson ER. Pathophysiology of ARDS and COVID-19 lung injury. Crit Care Clin. 2021;37(4):749-776; and Huppert LA, Matthay MA, Ware LB. Pathogenesis of acute respiratory distress syndrome. Semin Respir Crit Care Med. 2019;40(1):31-39.

Table 5.7

Chronology of Acute Respiratory Distress Syndrome (ARDS) Diffuse Alveolar Damage (DAD)

ARDS DAD
Exudative Phase (wk 1)	Proliferative Phase (wk 2 and 3)	Fibrotic Phase (>3 wk)
Interstitial edema Alveolar edema Leukocyte infiltration and agglutination Pneumocyte type I necrosis Endothelial cell necrosis Alveolar basement membrane denudation Hyaline membrane formation Platelet-fibrin thrombi	Interstitial fibroblast proliferation Pneumocyte type II proliferation and differentiation into pneumocyte type I Reestablishment of tight junctions and adherent junctions Restoration of epithelial barrier function Parenchymal necrosis Phagocytosis of apoptotic neutrophils Endothelial cell proliferation Restoration of endothelial barrier function Obliterative endarteritis Macrothrombi	Prolonged proliferation of fibroblasts Collagenous fibrosis Microcystic honeycombing Traction bronchiectasis Obliteration of microcapillaries Arterial tortuosity with mural fibrosis and medial hypertrophy

Modified from Tomashefski JF Jr. Pulmonary pathology of acute respiratory distress syndrome. Clin Chest Med. 2000;21(3):435-466. See also references , , and . wk: week.

Lung Injury From Mechanical Ventilation

Ventilator-induced lung injury (VILI) has garnered a lot of attention over the last decade ^, and continues to do so with the advent of the COVID-19 pandemic. In this section we will introduce the salient types of VILI and examine how mechanical ventilation exacerbates evolving lung injury of ARDS. At the end of inspiration when there is no air flowing into the lungs, lung inflation is maintained by what is referred to as the transpulmonary pressure, which is the difference between alveolar and pleural pressures and is closely linked with VILI. Details of transpulmonary pressure and how it can be estimated in the clinical setting are illustrated in Fig. 5.3 . Lung volume and transpulmonary pressure are interrelated, and the type of injury resulting from mechanical ventilation is based on whether ventilation occurs at high or low lung volumes. ^, Four types of VILI have been described—barotrauma, volutrauma, atelectrauma, and biotrauma. VILI is also related to lung injury that results from the efforts to breathe spontaneously on the part of the patient during mechanical ventilation and is described as patient self-inflicted lung injury (P-SILI). ^, Both VILI and P-SILI are “ventilation”-induced lung injury. Table 5.8 and Fig. 5.4 summarize the types of VILI, the mechanisms underlying the injury, and their clinical manifestations.

Table 5.8

Ventilator-Induced Lung Injury (VILI)

Mechanisms, Types, and Manifestations ^a

Mechanism of Injury	Type of Injury	Manifestations
Mechanical ventilation at high lung volumes Overdistension of the alveoli Repetitive strain	Barotrauma (air leaks) Volutrauma	Pneumothorax Pneumomediastinum Air embolism Subcutaneous emphysema Pulmonary edema
MV at low lung volumes Lung injury from opening and closing of alveoli Parenchymal shear injury Aggravated by heterogeneity	Atelectrauma
Patient self-inflicted lung injury (P-SILI), that is, spontaneous breathing during MV causing lung injury High tidal volumes from effort High transpulmonary pressure from vigorous efforts (pendelluft effect) Negative pleural pressure leading to increased transvascular pressure resulting in alveoli filling up with edematous fluid Asynchrony between spontaneous effort and MV Forced expiration causes diaphragmatic shift leading to lower end-expiratory lung volume and hypoxemia	Barotrauma	Pneumothorax
Downstream biological effects of all types of lung injury High lung volume Low lung volume Molecular and cellular events of lung parenchymal stress Increased alveolar-capillary permeability	Biotrauma	Pulmonary edema Pulmonary fibrosis (long-term) Multiple-organ failure from proinflammatory mediators entering systemic circulation

Data from Swenson KE, Swenson ER. Pathophysiology of ARDS and COVID-19 lung injury. Crit Care Clin. 2021;37(4):749-776; Slutsky AS, Ranieri VM. Ventilator-induced lung injury. N Engl J Med. 2013;369(22):2126-2136; Curley GF, Laffey JG, Zhang H, Slutsky AS. Biotrauma and ventilator-induced lung injury: clinical implications. Chest. 2016, 150(5):1109-1117; and Yoshida T, Fujino Y, Amato MB, Kavanagh BP. Fifty years of research in ARDS: spontaneous breathing during mechanical ventilation—risks, mechanisms, and management. Am J Respir Crit Care Med. 2017;195(8):985–992.

a The table also includes patient self-inflicted lung injury (P-SILI). Both VILI and P-SILI are “ventilation”-induced lung injury.

Following up on the observation that VILI originates in isolated and scattered regions of the lung parenchyma based on heterogeneity, Gaver III et al. proposed a permeability-originated obstruction response framework with four features: (1) enhanced endothelial permeability, (2) alveolar surfactant deactivation, (3) alveolar and interstitial edema, and (4) repetitive recruitment/derecruitment resulting in decreased compliance leading to atelectrauma, which has implications for mechanical ventilation management. Unless effective interventions are promptly undertaken, the proponents of the model warn that a repetitive cycle of these four steps will induce a vicious cycle resulting in increasing VILI.

Clinical Features and Diagnosis of Acute Respiratory Distress Syndrome

ARDS is characterized by the triad of rapid-onset hypoxemia, bilateral lung infiltrates, and pulmonary edema not attributable to a pathological cardiac condition. As the syndromic nomenclature based on clinical and pathophysiological criteria signifies, there is considerable heterogeneity in terms of cause, comorbidities, pathophysiology, and biology of ARDS. There is no specific biomarker to diagnose ARDS; moreover, no set of findings by themselves—molecular, clinical, radiological, or pathological—can be considered a sufficient set of markers for the diagnosis of ARDS. Thus there is a need for consensus diagnostic criteria for uniformity in approach to the diagnosis of ARDS.

ARDS diagnosis follows the Berlin definition criteria presented earlier in Table 5.1 . To recapitulate, the clinical recognition is based on (1) the onset of acute respiratory failure within 1 week of clinical insult, (2) noncardiac origin of respiratory failure, (3) bilateral opacities on chest radiograph or chest CT, and (4) hypoxemia. Although the diffuse alveolar damage feature is implicitly incorporated into the definition, autopsy studies of the lung identified diffuse alveolar damage features in only half of the patients dying from the syndrome. ^, Likewise, lung biopsy samples of patients hospitalized with ARDS demonstrated histological evidence of diffuse alveolar damage in similar proportions. ^{, ,}

Clinically, ARDS can be broadly categorized into two clinical phenotypes based on local or systemic etiological factors. See Fig. 5.5 for an illustration of these clinical phenotypes and their risk/causative factors. It has been shown that the risk for developing ARDS increases when the insult is local when compared with systemic factors.

Morphological phenotypes have been postulated based on radiological patterns observed in chest CT of patients with ARDS. Two distinct subgroups have been identified: (1) focal ARDS evidenced by consolidations in the lower lobes and dependent dorsal regions and (2) nonfocal ARDS with diffuse patchy opacities; the two groups had significant differences in lung physiological processes, with the lungs of the nonfocal ARDS group more recruitable than the focal ARDS group. ^{, ,} Researchers have developed a noninvasive method to estimate the extent of pulmonary edema based on chest x-rays. The radiographic assessment of lung edema (RALE) score, which evaluates the extent and density of alveolar opacities on chest radiographs, was found to be highly correlated with the severity of ARDS and increased mortality from ARDS. However, the relationship of the RALE score to the focal and nonfocal ARDS groups based on chest CT findings has not been established.

Investigators have also explored protein biomarkers to stratify and discover ARDS phenotypes. In particular, Forel et al. identified a specific protein biomarker, procollagen peptide (PCP-III), with the best cutoff value from bronchoalveolar lavage fluid in persistent ARDS to identify patients with lung fibroproliferation who could likely benefit from the administration of corticosteroids. Hamon et al. used the threshold level of PCP-III greater than 9 µG/L, indicating lung fibroproliferation, and compared chest scans of 228 patients with ARDS who had a bronchoalveolar lavage fluid PCP-III level and a chest CT scan; the study demonstrated significant association between high levels of PCP-III and CT scan fibrosis scores.

Using data from two randomized controlled trials (RCTs), the ARMA and ALVEOLI, and applying a latent class analytics approach, Calfee et al. identified two subphenotypes: the hyperinflammatory, characterized by severe inflammation, shock, and metabolic acidosis, leading to worse outcomes and labeled subphenotype 2, and the less inflammatory type, denoted subphenotype 1; the distribution was 30% for subphenotype 2 and 70% for subphenotype 1. The hyperinflammatory subphenotype 2 had higher plasma levels of inflammatory biomarkers, greater vasopressor use, lower sodium bicarbonate levels, and a higher prevalence of sepsis compared with subphenotype 1; subphenotype 2 also had a differential response to positive end-expiratory pressure (PEEP).

Using a different ARDS cohort from another trial, the Fluid and Catheter Treatment Trial (FACTT), Famous et al. identified the two subphenotypes, hyperinflammatory subphenotype 2 and the less inflammatory subphenotype 1, based on levels of IL-8, bicarbonate, and tumor necrosis factor receptor-1, and reported that fluid management protocols had significantly different effects on 90-day mortality outcomes in the two subphenotypes. A fluid conservative strategy resulted in lower 90-day mortality in subphenotype 1, whereas the same approach caused an increase in 90-day mortality in subphenotype 2.

There is currently no clinically valid biological marker for detecting ARDS with high sensitivity and specificity. However, markers of epithelial and endothelial injuries have been shown to be elevated in ARDS and they could be useful in distinguishing ARDS from local and systemic insults. Plasma biomarkers of epithelial injury such as surfactant protein-D and soluble receptor for advanced glycation end products are found to be elevated in ARDS from local causes. Likewise, plasma biomarkers of endothelial injury, for example, angiopoietin-2 and von Willebrand factor antigen, are higher in ARDS from systemic causes such as nonpulmonary sepsis.

Because bacterial and viral infections are common causes, a typical ARDS workup involves blood culture, sputum or tracheal aspirate Gram staining, and culture. However, identification of the causative pathogen may not be straightforward. Nucleic acid detection methods might help improve detection and can help with the identification of both bacterial and viral pathogens. ^,

Conditions that can mimic ARDS include congestive heart failure, interstitial lung diseases, connective tissue disorders such as polymyositis, vasculitis, lymphomas, and endobronchial tuberculosis; these are commonly referred to as ARDS mimics and would require additional investigations for a precise diagnosis.

Treatment and Prevention

The treatment for ARDS is evolving, but it is still mostly preventive and supportive. Once the patient has transitioned from lung injury to the stage of ARDS, effective treatments are still in experimental stages. Hence the focus of investigators has broadened to include identification and treatment of patients who are at risk for developing ARDS and show early evidence of lung injury. Based on this approach, three categories have been developed: (1) patients with risk factors for lung injury such as sepsis, shock, or pneumonia; (2) patients in the early stage of acute lung injury; and (3) patients meeting ARDS criteria. These categories and modified approaches for treating them are presented in Fig. 5.6 . Table 5.9 provides the protective ventilation and intervention strategies and their rationale. ^,

To summarize, the various strategies to reduce VILI include (1) low tidal volume ventilation to limit overdistention, (2) moderate to high PEEP to prevent derecruitment and injury from low lung volume, and (3) recruitment maneuvers with application of greater than 35 cm H ₂ O airway pressure to open collapsed lung segments.

There has been a recent trend toward finding subphenotypes of ARDS with the goal of personalization of treatment based on cause, various types of biomarkers, or gene expression data. ^{, ,} Sinha et al. proposed a method of clinical stratification of ARDS for subphenotype identification based on a machine learning method called latent class analysis using clinically available data. Once the model is validated using clinical data from different centers, it is possible that the machine learning model could be useful at the bedside for phenotype identification. Wick et al. discuss the promises and challenges of developing a personalized medicine approach for the management of ARDS. The review specifically identified the heterogeneity in the presentation and manifestation of ARDS and the differences in the cause of ARDS in developed and developing countries as two significant challenges for developing a patient-specific paradigm of treatment.

As Kotas and Thompson have nicely put it, ARDS is a syndrome and not a disease. To maximize desirable outcomes, it is important to identify the cause of ARDS in a patient-specific manner while following the general principles and guidelines for the management of ARDS.

Long-Term Outcomes

Patients recovering from ARDS suffer from significant long-term health consequences. Muscular weakness, anxiety, depression, and posttraumatic stress disorder (PTSD) have been reported in survivors.

With this basic layout of the different facets of ARDS, which hopefully will serve as a foundation for understanding the lung pathological processes and management of COVID-19 respiratory manifestations, the chapter will proceed to discussion of the pathophysiology, management, and complications of lung injury resulting from SARS-CoV-2 infection. The focus will be on the pathophysiology of COVID-19 lung injury, similarities and differences between COVID-19 ARDS and classic ARDS, pulmonary complications of COVID-19, management protocols, and long-term sequelae after SARS-CoV-2 infection.

Epidemiology of COVID-19 Acute Respiratory Distress Syndrome

The epidemiology of SARS-CoV-2 and COVID-19 has been covered in detail in an earlier chapter. As mentioned earlier, the annual incidence of ARDS in the United States before the advent of the COVID-19 pandemic was 190,000. Based on the mortality figure of 650,000 in the United States from March 2020 to September 2021 and assuming that 80% had severe pneumonia/ARDS and attributing a mortality of 20% from COVID-19 ARDS, Wick et al. estimate an incidence of 2.5 million cases of COVID-19–associated ARDS over 18 months or about 1.7 million over a period of 12 months.

Pathophysiology of COVID-19 Acute Respiratory Distress Syndrome

In the early months of the SARS-CoV-2 pandemic, a controversy emerged on whether the pathophysiological mechanisms of COVID-19 ARDS and classic ARDS were significantly different to warrant an alternative approach to the management of the condition. However, a consensus later emerged signifying that there are more similarities than differences between COVID-19 ARDS and ARDS from other causes and that COVID-19–related lung injury should be managed based on the lung-protective strategies recommended for classic ARDS. ^,