Reply to Camporota et al.: “Established” Respiratory Treatment in Acute Respiratory Distress Syndrome: Scientific Rigor or a Square Peg in a Round Hole?

From the Editorialist:

Acute respiratory distress syndrome (ARDS) is a heterogeneous syndrome. Given this heterogeneity, there have been many efforts to define subgroups of patients with ARDS: by etiology, by respiratory mechanics, by the distribution of radiographic abnormalities, by the severity of gas exchange abnormality, or by biomarkers. In their response to the editorial “Novel Phenotypes in Respiratory Failure: Same as It Ever Was” (1), Dr. Camporota, Dr. Gattinoni, and Dr. Marini articulate the plausible and widely held hypothesis that such subgroups would benefit from distinct treatment strategies. Indeed, this hypothesis has animated much of the recent research literature on ARDS. To date, however, it has proven surprisingly difficult to prospectively demonstrate a mortality benefit from any particular tailored approach. Prone ventilation has proven beneficial for patients with a Pa_O2/Fi_O2 (P/F) ratio < 150, though many clinicians continue to harbor suspicion that those are not the only patients who benefit. In contrast, trials that enrolled patients regardless of subtype have shown robust mortality benefits. Although studies are urgently needed to examine the hypothesis that Camporota and colleagues articulate, plausibility and internal consistency of a hypothesis are not the standards on which we should base a change in practice. In the particular case of coronavirus disease (COVID-19), we have witnessed the widespread adoption of plausible therapies that subsequently proved to be of no benefit (2). For the present, the best evidence indicates that patients with acute onset of respiratory failure that is not fully explained by fluid overload, bilateral infiltrates, and P/F ratio < 300 on positive-end expiratory pressure (PEEP) of at least 5 cm H₂O benefit from low Vt ventilation, regardless of etiology.

I agree wholeheartedly with the letter writers that lung strain—increment from resting volume—is likely the key mechanical determinant of lung injury in ventilated patients with ARDS. I further agree that Vt alone is a poor proxy for lung strain. There have been signals of this in the literature, notably the retrospective study demonstrating a stronger association of driving pressure with mortality than of Vt (3). It is therefore plausible that there exists a better way to determine what strain is tolerable in a given patient. Absent of evidence, however, we must not become overly persuaded by the plausibility of any given hypothesis. After all, it is very plausible that there exists a way to titrate PEEP to an individual patient’s mechanics, and yet, thus far, trials have consistently failed to demonstrate a benefit to individualized PEEP titration.

With respect to their assertion that ARDS secondary to COVID-19 presents with a distinct distribution of respiratory system compliance, the authors are on less solid ground. In support, they cite an editorial by Dr. Marini, which was published in association with an electrical impedance tomography study of 10 patients in a single center. In contrast, the now multiple large case series of patients with COVID-19 (4, 5) report distributions of respiratory system compliance (and P/F ratio) that are remarkably consistent with large case series before the COVID-19 pandemic, such as LUNG-SAFE (Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure) (6). The literature to date simply does not support the idea that severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes a mechanically distinct form of respiratory failure.

While we await further investigation and potential identification of individualized therapies that improve outcomes, clinicians should feel very comfortable with the continued application of the evidence-based therapies developed in the era before COVID-19.