ARDS in the time of corona: context and perspective.

For more than 2 years, COVID-19 has been holding the world at awe with new waves of infections, novel mutants, and still limited (albeit improved) means to combat SARS-CoV-2-induced respiratory failure, the most common and fatal presentation of severe COVID-19. In the present perspective, we draw from the successes and-mostly-failures in previous acute respiratory distress syndrome (ARDS) work and the experiences from COVID-19 to define conceptual barriers that have so far hindered therapeutic breakthroughs in this deadly disease, and to open up new avenues of thinking and thus, ultimately of therapy.

INTRODUCTION

The SARS-CoV-2 virus has filled and brought to a breaking point the intensive care units all over the world. Most patients are dying a “lung death” (1); specifically, they die from acute respiratory failure, acute respiratory distress syndrome (ARDS) (2). The lung is built of delicately thin alveolar capillary membranes, and to function as a gas exchanger the organ must stay dry. ARDS constitutes a catastrophic breakdown of the integrity of barriers and a loss of homeostatic maintenance of lung fluid balance due to inadequate stress responses resulting in a cell- and protein-rich pulmonary edema that impairs pulmonary gas exchange. Accordingly, strategies to improve or replace the lung’s diffusive gas exchange capacity such as extracorporeal membrane oxygenation (ECMO) or high-flow oxygen therapy have proven successful in reducing COVID-19 mortality (3) or decreasing the time to clinical recovery (4), respectively.

It is historically important to note that the need to care for patients with respiratory failure and to manage the patient-ventilator “interface” stood at the beginning of respiratory care units and shaped the specialty of Pulmonary and Critical Care Medicine (5). COVID-19 has now turned intensive care units into combat zones, forced physicians (“First do no harm”) to apply triage algorithms, and painfully laid bare our state of ignorance of what ARDS is (6–8).

The purpose of this perspective is thus twofold: 1) to break down conceptual barriers that may have prevented progress as the mortality from ARDS has hardly improved over the years, and 2) to highlight areas of future research and encourage experimental treatments based on experiences where the COVID-19 pandemic either generated new insights (e.g., the use of steroids, vide infra) and opportunities (e.g., the potential to shift from treatment to prevention), triggered new areas of research (such as the extensive use of single cell transcriptomics), or notably failed to do so against initial expectations (e.g., with respect to ventilation settings).

As enormous as the COVID-19 challenge is, it is not insurmountable, and this challenge can become a driver, which motivates the community of emergency room- and ICU physician/scientists, in partnership with the pharmaceutical industry, to make investments in ARDS research, a field that—with some exceptions—has been largely neglected since the failure of a series of pharmacological trials in the late 1990s and early 2000s (9).

CONCEPTUAL BARRIERS

There are many, beginning with the misconception that ARDS “lives mostly in the airspaces,” neglecting the fact that in ARDS, as compartment barriers break down and the tissue is filled with myriads of inflammatory cells, the whole lung is sick, including the vasculature. Although this is not new information—pulmonary thrombotic obliterations and pulmonary hypertension have been recognized for decades as clinical associations and complications of ARDS—the description of the microvascular component of severe COVID-19-associated lung disease (10) has now stimulated the broader recognition and investigation of vascular barrier failure and intravascular inflammation. Specifically, in addition to infecting lung epithelial cells and activating a plethora of inflammatory cells, COVID-19 causes endothelial inflammation (10), barrier failure (11), and injury (12, 13). The end stage of COVID-19-induced lung tissue destruction has recently been described in a detailed autopsy study and compared with other etiologies of diffuse alveolar damage (14). In addition to extensive vasculopathic changes including thromboemboli, pulmonary infarcts, and perivascular inflammation, COVID-19 was specifically characterized by alveolar-septal congestion that was associated with a negative outcome, highlighting the relevance of the pulmonary vasculature in this disease.

Much of this response is neither novel nor unique to COVID-19; in fact, endothelial activation and leak have been recognized as hallmarks of experimental lung injury for decades (15, 16). What has been missing, however, is the translation of the extensive preclinical work in this area into transformative patient benefit. The COVID-19 pandemic has strikingly highlighted this “conceptual barrier,” in that clinical trials have focused almost exclusively on antiviral and anti-inflammatory/anti-immunothrombotic therapies, with little consideration for vascular or endothelial-targeted strategies in spite of excellent rationales (17, 18): 1) endothelial-targeted therapies may prevent barrier leak and hence respiratory failure, the main cause of fatality in ARDS; 2) endothelial-targeted therapies may attenuate vascular hallmarks of severe COVID-19, namely, vascular inflammation and thrombosis; 3) endothelial-targeted therapies may attenuate systemic dissemination of the disease; and 4) endothelial-targeted therapies are not pathogen-specific, thus provide strategies for future pandemics beyond COVID-19.

The systemic dissemination of an initially respiratory infection in COVID-19 disease also permits us to pose the hypothesis that the “sick lung circulation” releases cells, cell fragments, mediators of inflammation (19), free DNA, and extracellular vesicles (20), and generates an imbalanced oxidant/antioxidant milieu that impacts other organ circulations in the heart, the kidneys, and the brain (21–23). Importantly, this systemic dissemination is in most cases not caused by a spread of the virus throughout the body directly infecting systemic organs, but rather to the release of the above signals from the “sick lung circulation” (24, 25). In line with this view, initial reports of viral infection of the pulmonary or systemic endothelium could at large not be confirmed in the laboratory (26), identifying COVID-19 as a respiratory infection with systemic manifestations due to secondary effects rather than a systemic infection.

The ARDS lung then needs to be seen as the center of an integrated system, and as the motor of a pathological process rapidly spiraling out of (homeostatic) control. The injured lung’s “pleas for help” are answered by a host of cells released by the bone marrow. Specifically, single-cell transcriptomics have proven as an extremely powerful tool to decipher these immune cell responses. For example, studies by Grant et al. (27) identified a slowly unfolding, spatially restricted alveolitis in patients infected with SARS-CoV-2 in which alveolar macrophages containing the virus and T cells form a positive feedback loop that drives a persistent alveolar inflammation. Other work identified an upregulation of specific macrophage (28) or T cell subsets such as CD16+ T cells that directly promote lung microvascular endothelial cell injury (29).

Age, diabetes, hypertension, and obesity in patients infected with SARS-CoV-2 have been identified as predictors of bad outcome (30–32). It may be speculated that endothelial cell dysfunction and the low-grade inflammation of obesity prime the pulmonary endothelium for injury and failure in the COVID-19 scenario. However, partially due to the paucity of suitable animal models for COVID-19 research, the mechanistic basis of this link still remains to be elucidated and presently poses a considerable conceptual knowledge gap that should urgently be addressed to allow for a better understanding of the individual response to the virus and the role of preexisting comorbidities.

ARDS SUSCEPTIBILITY AND SUBPHENOTYPES

Provided identical insults to the lung tissue, like aspiration, pneumonia, or mass transfusion, not every patient develops ARDS. This was first observed at the University of Colorado where patients with known risk factors or triggers were followed to estimate the incidence of ARDS during a period of 1 year. Of 936 patients with one or more risk factors, 10% developed ARDS (33). The questions “Why only 10%?” and “What differentiates the one with ARDS from the nine who do not develop it?” have still not been answered. Several genetic traits have been identified to predispose for ARDS including variants or polymorphisms in the genes encoding for aquaporin-5, VEGF, angiotensin-converting enzyme (ACE), IL-8, or hypoxia-inducible factor (HIF)-1α, but can so far not sufficiently explain this clinical variability (34).

Armed with the tools of liquid biopsies, multiomics technology, and artificial intelligence (AI), it is anticipated that we will develop a better understanding of ARDS susceptibility, and this could lead to a treatment concept of intervening pharmacologically at the stage of “pre-ARDS” (35). As such, ARDS may be compared with an avalanche, as once the catastrophe has occurred it is too late for therapy. COVID-19 may be an almost ideal use case for such a pre-ARDS strategy, as—in contrast to classic ARDS which develops acutely—it is characterized by a gradual onset of symptoms that progress in a slow-crescendo from flu-like symptoms to pneumonia and ultimately, respiratory failure (36). Identifying and validating such a pre-ARDS stage would transcend efforts to identify subphenotypes of ARDS (37). Unbiased approaches by latent-class analyses have revealed the existence of such different subphenotypes, which are independent from underlying cause or the classic distinction between “direct” and “indirect” lung injury (38). Most notably, this has resulted in the recognition of a “hyper-inflammatory” and a “hypo-inflammatory” phenotype, which in retrospective analyses showed markedly different responses to pharmacological or physiological interventions (37, 38). Identification of these subphenotypes may hence present a powerful tool for improved patient stratification in future clinical ARDS trials.

VENTILATOR-INDUCED LUNG INJURY

When connecting a patient with respiratory failure to a mechanical ventilator, the hope is that this support system will be required only for a short term, in part because the machine-delivered pressure and volume trigger parenchymal stress, cellular injury, and mechano-induced endothelial barrier failure and inflammatory responses, thus further aggravating lung injury. Although many excellent recent reviews highlight this topic (39), only a small number of strategies to optimize mechanical ventilation and thus, reduce ventilator-induced lung injury (VILI) have proven successful in large clinical trials, namely, ventilation with lung protective low tidal volumes (40), prone positioning (41), and neuromuscular blockade (42). In essence, the same strategies improve outcome in COVID-19 ARDS and despite abundant initial speculations as to potential differences between severe COVID-19 and classic ARDS (43, 44), the pandemic did not yield any additional strategies that would improve outcome in ventilated patients.

Other respiratory maneuvers such as increased positive end-expiratory pressure (PEEP) (which is considered to prevent alveolar open-and-collapse phenomena) have—despite solid physiological concepts and abundant supportive preclinical data—failed to show clinical benefit in terms of mortality despite improved oxygenation in clinical trials (45). Possible reasons for these failures highlight the complexity of the business—for e.g., PEEP may in fact not alleviate alveolar opening-and-collapse but merely shift its vertical location (46), opening-and-collapse may in fact reflect cyclic alveolar flooding (47), PEEP may improve alveolar aeration yet impair ventilation-perfusion matching (48), increase mechanical power (49), and/or redistribute blood flow from alveolar capillaries to corner vessels (50), or PEEP may be beneficial in some yet detrimental in others as suggested by retrospective analysis of ARDS subphenotypes (36, 38). The spatial and temporal interplay between respiratory and cardiovascular mechanical forces that determines alveolar dynamics, capillary perfusion, gas exchange, and activation of mechanotransduction pathways is intricate to begin with, but it becomes infinitely more complex with increasing heterogeneity in ventilation and perfusion as characteristic for ARDS in both patients with COVID-19 and patients with ARDS of other causes.

PHARMACOLOGICAL TREATMENTS: “BECAUSE IT’S INFLAMMATION, STEROIDS SHOULD WORK”

Within a short time following the lung tissue damage-triggering event, there is a rapid influx of inflammatory cells into the tissue. Thus, chemotaxis and mediators of inflammation contribute to the avalanche-like disaster. From the first report of ARDS by Ashbaugh et al. (5), it has thus been postulated that steroids would be effective in this setting. Indeed, in the present COVID-19 pandemic, dexamethasone—in contrast to other drugs such as hydroxychloroquine or lopinavir-ritonavir tested in parallel in the UK’s RECOVERY trial (51)—proved beneficial, in that it cut deaths by a third among hospital patients who required ventilation and by a fifth among patients receiving oxygen only (52). However, important unresolved questions remain, as there was no evidence that dexamethasone provided any benefit among patients who were not receiving respiratory support at randomization; in fact, the results in this subgroup were consistent with possible harm. As such, the authors speculate that the beneficial effect of steroids in severe viral respiratory infections may depend on appropriate dosing, timing, and patient selection, yet criteria for such decisions remain to be resolved. Furthermore, the RECOVERY results contrast with the findings from previous ARDS steroid trials (53–55). This may in part be attributable to the fact that in contrast to other classical causes of ARDS such as pneumonia or sepsis, ridding the body of the infectious pathogen is not the core of the problem in COVID-19. Hence, while steroid therapy may aggravate pulmonary or systemic pathogen burden in bacterial or fungal infections, no similar increase in viral load has been reported for SARS-CoV-2, which generally is eliminated from the body within the first 1–2 wk after infection, whereas ARDS persists (56). Second, inflammatory cells are not only critically involved in lung injury, but also in lung repair (57). As such, steroids may—dependent on their timing—do more harm than good. Third, these disparate findings lead us to consider that there are steroid-resistant forms of inflammation, and also that this steroid resistance is poorly understood.

In the following, we also briefly highlight a few selected nonsteroidal anti-inflammatory agents that have been investigated or proposed for the treatment of ARDS in patients with COVID-19 (Fig. 1): The effects of imatinib treatment of hospitalized patients infected with SARS-CoV-2 have recently been investigated in a placebo-controlled multicenter clinical trial (58). Imatinib received orphan drug status for ARDS in Europe in 2015. The hypothesis was that imatinib would reduce the pulmonary capillary leak; this hypothesis was based on preclinical and clinical data, which support the concept that imatinib protects the integrity of the alveolar-capillary barrier in the setting of inflammation (59, 60). Imatinib inhibits the Abl-1 and Abl-2 tyrosine kinases, NF-κB-dependent production of inflammatory mediators, and also has an antiviral effect. The CounterCOVID study was a randomized, double-blind, placebo-controlled trial in hospitalized patients with COVID-19 requiring supplemental oxygen. The patients were randomly assigned to oral imatinib (800 mg loading dose, followed by 400 mg daily for 9 days) or placebo. Primary outcome of the study was time to liberation from ventilation and supplemental oxygen for more than 48 h while being alive during a 28-day period. Secondary outcomes included 28-day mortality and need for invasive mechanical ventilation. No difference was found in time to liberation from ventilation and supplemental oxygen between the two groups. However, significant reductions in mortality (8% in the imatinib group, compared with 14% in the placebo group) were observed, and in the median duration of invasive mechanical ventilation (58).

Figure 1.

Vascular-targeted therapeutic interventions in COVID-19. Strategies with direct or indirect action on the vasculature and demonstrated patient benefit in COVID-19 comprise (among others) the anti-inflammatory effect of steroids, antagonization of interleukin-6 (IL-6) receptors by tocilizumab, or endothelial barrier stabilization by imatinib. Experimental approaches illustrating novel strategies include the replacement of dead or diseased endothelial cells, or the inhibition of procalcitonin by dipeptidyl-peptidase 4 inhibitors such as sitagliptin. [Image created with BioRender and published with permission.]

Tocilizumab, a humanized monoclonal antibody against the interleukin-6 (IL-6) receptor that is clinically approved for the treatment of rheumatoid arthritis and other chronic inflammatory diseases in some countries, has been studied in hospitalized patients with severe COVID-19 (61, 62) with the assumption that neutralizing the action of IL-6 that is highly elevated in patients with a severe disease course would improve outcome. A meta-analysis of 22 clinical studies found that “the mortality pooled prevalence of tocilizumab-treated patients was moderately lower than the overall mortality of the severely sick patients with COVID-19” (61), and tocilizumab has been approved for the treatment of COVID-19.

CONCLUSIONS AND OVERALL PERSPECTIVE

At the time of this writing, the COVID-19 pandemic has taken more than one million lives in the United States. Annually, ∼200, 000 patients per year are diagnosed with ARDS in the United States; with a mortality rate of ∼30%, ARDS thus causes 60, 000 annual deaths in the United States alone. The mortality from COVID-19 “lung death” is now greater than 10-fold that in a “usual” year.

How can we affect this devastating ARDS? We have to ask ourselves whether our current supportive care—which has arrived at “intelligent” ventilation strategies and “chicken little” (not too much to flood the lung and not too little to starve the kidneys) fluid regimens in the setting of ARDS—is actually amounting more or less to “palliative care.”

We believe that the current practice goals need to be complemented by strategies that address the maintenance and/or reconstitution of barrier integrity and lung cell homeostasis. One of the lessons that COVID-19-induced lung injury may teach us is that the virus can destroy the lung repair program. Sequelae of COVID-19 lung disease may be permanent damage to the alveolar membranes and lung fibrosis—with no repair.

That is why we encourage the community of intensive care physicians and academic pulmonary and critical care physicians to participate in clinical trials, which can be designed to explore “out of the box” approaches to the treatment and prevention of COVID-19 lung disease and its long-term sequelae. One exemplary such approach has recently been described by Hisata et al. (63), who reported reversal of emphysema by replenishment of endothelial cells in a mouse model of elastase-induced lung injury. Admittedly, this study is still a long way from the clinic, as the authors used intravenous delivery of healthy endothelial cells in mice; however, the concept of supporting lung vascular endothelial cells in a clinical “pre-ARDS” stage is worth contemplating (Fig. 1). In a more translational example, Brabenec et al. (64) recently identified procalcitonin as an endothelial barrier-disruptive mediator. As procalcitonin requires activation by dipeptidyl-peptidase 4 (DPP4), DPP4 inhibitors such as the clinically approved drug sitagliptin are a means to interfere with this pathway. Notably, sitagliptin not only reduced capillary leak in a murine sepsis model, but when administered to patients before cardiac surgery improved sublingual microvascular perfusion while reducing fluid and vasopressor demand.

Fifty years after the description of the “Danang Lung” that first generated a wider appreciation of the problem of ARDS, these studies highlight the potential of novel approaches to preserve or restore lung barrier integrity and repair in the injured lung. These are great challenges that can only be met with a deeper understanding of the biology of the lung and experimental therapeutics. We hope that the COVID-19 pandemic and its medical challenges will provide a windfall for an improved treatment of all patients with ARDS; certainly, the pressure is on: learn from COVID-19 and apply to ARDS.

GRANTS

This work was supported by the German Research Foundation (Grants: SFB-TR84 A2 and C9, SFB 1449 B1, SFB 1470 A4, KU1218/9-1, KU1218/11-1, and KU1218/12-1), the German Ministry of Education and Research (Bundesministerium für Bildung und Forschung, BMBF) in the framework of SYMPATH (01ZX1906A) and PROVID (01KI20160A), and the German Centre for Cardiovascular Research (Deutsches Zentrum für Herz-Kreislaufforschung, DZHK).

DISCLOSURES

Wolfgang Kuebler is an editor of American Journal of Physiology-Lung Cellular and Molecular Physiology and was not involved and did not have access to information regarding the peer-review process or final disposition of this article. An alternate editor oversaw the peer-review and decision-making process for this article. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

N.F.V., H.J.B., and W.M.K. drafted manuscript; N.F.V., H.J.B., and W.M.K. edited and revised manuscript; N.F.V., H.J.B., and W.M.K. approved final version of manuscript.