ARDS in COVID-19 and beyond: Let’s keep our eyes on the goal instead of the straw man

Almost immediately after COVID-19 related acute respiratory failure began exponentially spreading across the globe, debate sparked regarding whether we were treating acute respiratory distress syndrome (ARDS) as we know it or some new variant unique to the SARS-CoV-2 virus. The debate has occurred in the medical literature amongst giants in ARDS research^1–3 and has spilled over into popular media, spreading predominantly on social media platforms. The crux of the debate is whether this new hypoxemic syndrome, presumably mediated by a combination of loss of autoregulatory mechanisms and microthrombosis precipitated by viral actions on the renin-angiogensin-aldosterone pathway and pulmonary vascular endothelium, respectively, conforms to our understanding of the “typical” ARDS with which we are already familiar. Some have even created new taxonomies, such as CARDS (COVID-19 patient with ARDS) ² and L-type ARDS, ⁴ to describe the severe hypoxemia with relatively normal lung compliance, with the implication that management strategies should accommodate these novel phenotypes. Using similar rationale, others have suggested that mechanical ventilation should be avoided, in favor of non-invasive positive pressure ventilation (NIPPV), high-flow nasal oxygen (HFNO), and even venovenous extracorporeal membrane oxygenation (VV-ECMO) to avoid ventilator-induced injury and prolonged mechanical ventilation.

However, we must ask ourselves whether the “CARDS” for which we are devising new supportive and ventilatory strategies is simply a straw man. Autopsy evidence does support an increased microvascular thrombotic burden in COVID-19 patients, particularly compared to patients with influenza. ⁵ However, autopsy studies also report that the predominant pulmonary finding is diffuse alveolar damage and hyaline membrane formation, ⁶ the same predominant findings as in “wild-type” ARDS. Evidence shows that ARDS has always occurred on a spectrum of compliance, and early evidence from COVID-19 patients with ARDS is consistent. ⁷ In essence, we think of ARDS as the patient with diffuse disease and stiff lungs (i.e. the “H-type” ARDS) because of availability heuristic ⁸ and the difficulty treating this dramatic syndrome.

While our understanding of lung protection is evolving, low tidal volume ventilation has been one of the only, if not the only, consistently durable interventions to reduce mortality in critically ill patients. ⁹ Whether low tidal volume, or a direct corollary like driving pressure, ¹⁰ is responsible for the reduced mortality is debatable. Unfortunately, a clinical trial would require patients that received low tidal volumes but high driving pressure compared to the converse, which is infeasible.

Recent physiological studies have revealed several variables for consideration in the pursuit of lung protection regardless of taxonomy, timing, and precipitating etiology. Excessive tidal volume, as we know, is injurious, particularly when approaching or exceeding the resting lung volume, which often decreases proportionally to the severity of the disease and is interrelated to driving pressure as a reflection of lung compliance. Lung strain (tidal volume/functional residual capacity, TV/FRC) and lung stress (transpulmonary pressure) become more injurious through regional amplification via stress risers in spontaneously breathing patients, ¹¹ contributing to patient-self-induced lung injury (P-SILI). ¹² This is a likely explanation, along with high tidal volumes, for why NIPPV failure is associated with increased mortality,^13–16 but HFNO failure may not be as detrimental. ¹⁷ While noninvasive strategies can be helpful at preventing intubation in some patients, failure can be consequential. Guided by the principles of regional injury through P-SILI, patients with high work of breathing despite noninvasive ventilation, are at risk of detrimental injury from further spontaneous breathing driven injury.^18,19 This progression of lung injury likely facilitates the evolution of “L-type” ARDS to the “H-type” that we have availability bias towards and the overwhelming patient volume has prodded our cognitive dissonance into explaining it away as a new phenotype.

Although every effort should be made to appropriately phenotype this disease, and ARDS in general, there are simply not sufficient data at this point to justify new taxonomy or management strategies. To the contrary, physiologic and pathologic data support the conclusion that this disease is on the same spectrum as ARDS of any etiology. We should stay the course using our current understanding of the principles of lung protective ventilation, which involve limiting tidal volume, limiting driving pressure, and reducing mechanical power. We should generally avoid practices that may cause further harm with injurious lung volumes and pressures. Similarly, ECMO may be an option in patients where lung protection and adequate gas exchange cannot be maintained simultaneously. There are, however, questions that require expedient answers as we face ARDS from a second COVID-19 wave and beyond. How can we best clinically monitor respiratory drive in patients on noninvasive oxygenation strategies? ¹⁹ What is the ideal duration of proning? What is the utility of the recruitment-to-inflation ratio to optimize ventilation/perfusion (V/Q) mismatch and reduce the proning duration? ²⁰ How can we best assess injurious respiratory drive in spontaneously breathing patients after the 48-hour quiescence from proning and paralysis using readily available information from the ventilator (e.g. P_0.1, P_occ)?^21–23 How can we best manage that drive to strike a balance between prevention of worsening lung injury and recurrent respiratory failure and avoidance of over-sedation, delirium, and deconditioning? What, if any, is the role of inhaled pulmonary vasodilator therapy in facilitating lung-protective ventilation, improving gas exchange, and reducing the fraction of inspired oxygen (FiO₂)? What is the utility for anticoagulation or thrombolytics, and what is the role of thromboelastography in guiding those decisions? These questions deserve rigorous scientific inquiry and debate so we can stay focused on the goal and not the straw man.