Recent advances in the understanding and management of ARDS

Initial ventilator settings

With the lack of effective pharmacologic interventions, initial management has focused on ventilator strategies that stabilize gas exchange while avoiding further injury to lung tissue (ventilator-associated lung injury). Despite many studies on ventilator strategies and guidelines published by various professional societies to promote evidence-based management^18,19, there remains a diversity of opinion on optimal management. A foundational study of ARDS demonstrated that mortality was reduced in ARDS when patients received lower tidal volumes (initial tidal volume of 6 mL/kg) compared with higher tidal volumes (initial tidal volume 12 of mL/kg)²⁰. This is theorized to be due to a reduction in overdistention of ventilated lung with high regional volumes and pressures—so-called volutrauma and barotrauma—and has become the standard of care. Subsequent studies have shown that the use of an “open lung” ventilation strategy improves lung mechanics, oxygenation, and inflammatory markers²¹. This strategy relies on the use of optimal lung recruitment to increase the fraction of lung that is aerated with the goal of delivering the set tidal volume to the largest functional lung. Different strategies of titrating PEEP have been studied extensively. PEEP is thought to benefit patients with ARDS by preventing derecruitment of collapsible lung units at the end of exhalation, thus preventing the cyclic opening and closing of alveoli that can promote lung injury via shear stress. However, the ALVEOLI trial²² previously demonstrated no benefit of high PEEP over low PEEP in ARDS when low tidal volumes and limitations of plateau pressure (PPlat) were used. More recent studies may help clarify the effect of various open lung ventilation and PEEP titration methods on outcomes.

Driving pressure and positive end-expiratory pressure titration

Given the prior association of high airway pressures with poor outcomes in ARDS, conventional management focused on minimizing the PPlat. However, a recent retrospective analysis of patients enrolled in ARDS trials showed that the physiologic parameter best associated with outcomes was, in fact, driving pressure (PPlat – PEEP)²³. This variable is determined by both the compliance of the respiratory system and the delivered tidal volume and may be better associated with outcomes because in some patients a high PEEP is required to optimize recruitment (for example, in obesity), which in turn results in a high PPlat (>30 cm H₂O) but these patients are not experiencing significant regional overdistention. Conversely, patients may be on very low PEEP and have an “acceptable” PPlat but have a very high distending pressure, reflecting a poorly compliant respiratory system and high regional strain. Data on driving pressure remain limited; it has not yet been the target of a randomized clinical trial. Additionally, in the recent ART trial²⁴ (discussed below), the intervention group had a lower mean driving pressure and higher mortality, suggesting that its predictive value is limited. However, as we await further clinical trial data, the physiologic rationale of optimizing compliance through recruitment and judicious tidal volumes to avoid overdistention, thereby minimizing driving pressure, remains strong.

The ART trial²⁴ assessed the effect of a lung recruitment maneuver and PEEP titration to the best respiratory system compliance (versus an empirical PEEP strategy) on all-cause mortality. In that study, patients in the intervention group underwent a “maximum alveolar recruitment maneuver” followed by a decremental PEEP trial to determine the level at which respiratory system compliance was optimized. Of note, the recruitment maneuver protocol was modified partway through the study after reported serious adverse outcomes, including cardiac arrest. The study authors found a significantly increased 28-day mortality in the intervention group compared with control (55.3% versus 49.3%), raising questions about the safety of such recruitment maneuvers as part of an open lung ventilation strategy. Additionally, the patients in the study cohort did not appear to be very recruitable; although driving pressure was improved in the intervention group, it was not different enough to be expected to significantly affect mortality. It remains unknown whether recruitment maneuvers with lower (and possibly safer) airway pressures in a more recruitable cohort of patients may provide the benefit of alveolar recruitment with less risk. It also remains unknown whether a decremental, physiologically targeted PEEP titration method may have benefit when not combined with such aggressive recruitment.

The placement of an esophageal catheter to measure esophageal pressure (a surrogate for pleural pressure) has also been used to guide PEEP titration. This is based on the concept that excessive transpulmonary pressure (airway pressure minus pleural pressure, the distending pressure of the lung parenchyma) is a more accurate indicator of ventilator-induced lung injury than use of airway pressure alone. However, the recent EPVent-2 study²⁵ showed no significant improvement in ARDS outcomes with use of esophageal manometry to target a combination of transpulmonary pressure and FiO₂ compared with PEEP titration based on an empirical high PEEP-FiO₂ table. Significantly, there was no difference in the primary outcome measure, a composite score of death and days free from mechanical ventilation through day 28. There was also no significant difference in other physiologic variables—including PEEP, driving pressure, and PPlat—between the two groups. There are therefore no data to suggest that routine use of esophageal manometry, in combination with oxygen targets, is necessary to optimize ventilator settings and minimize ventilator-associated lung injury in patients with ARDS. However, clinically, not all physicians use a combination of transpulmonary pressure and oxygen targets in titrating PEEP with an esophageal balloon, which means that this study may not reflect common practice. Additionally, esophageal balloons allow distinction of the two mechanical elements of the respiratory system (chest wall and lung parenchyma), which remains useful in complex clinical situations.

Prone positioning

Placing patients with ARDS in the prone position has multiple physiologic effects that are potentially beneficial. It has been known for quite some time that the prone position improves oxygenation; this occurs because of both more homogeneous distribution of perfusion to the lung and recruitment of collapsed lung units, which reduces that amount of perfusion that is functional shunt. Additionally, recruitment alone may be beneficial since it increases the functional size of the lung, thereby reducing the risk of volutrauma/barotrauma, and may decrease the risk of atelectrauma. However, clinical trial data on prone positioning in ARDS have been conflicting, and there have been ongoing concerns about potential adverse events, including dislodgement of indwelling lines and the endotracheal tube and pressure ulcers^26–28. A large randomized clinical trial, the PROSEVA trial²⁹, did show an impressive mortality benefit (32.8% versus 16.0% 28-day mortality) in patients who are placed in the prone position early in illness (<48 hours after ARDS onset) and who are maintained in the prone position for most of the day and until gas exchange is significantly improved. Additionally, this study did not show any increased risk for adverse events in patents placed in the prone position despite being conducted across many trial sites. These data should strongly encourage clinicians to consider prone positioning in patients with moderate to severe ARDS, and prone positioning should be considered as an upfront therapy and not a rescue therapy. However, the comparison group in the PROSEVA trial received relatively low levels of PEEP, which may have biased toward a benefit with prone position since PEEP can provide some of the same physiologic benefits as prone positioning. Therefore, a second large study comparing prone position with higher levels of—or personalized—PEEP would be useful in assessing the necessity of prone positioning in patients with ARDS.

Pharmacologic therapy

Neuromuscular blockade (NMB) has been a common intervention for patients with moderate to severe ARDS after a randomized trial demonstrated a mortality benefit to NMB for 48 hours in patients with a P:F ratio of less than 150 early in their illness³⁰. There are multiple plausible mechanisms for the benefit of NMB in ARDS, including elimination of patient-ventilator asynchrony, which reduces the risk of volutrauma and barotrauma, and a reduction in end-expiratory derecruitment due to expiratory effort, which thereby reduces atelectrauma. These potential benefits are mirrored by a reduction in circulating inflammatory mediators in patients receiving NMB, which may also be due in part to a direct effect of NMB. However, a recent randomized trial that attempted to replicate the prior positive result showed no benefit to NMB in a similar patient cohort³¹. Additionally, the group of patients receiving NMB were less mobile and had a greater incidence of serious adverse cardiovascular events. One criticism of that study has been that many patients were excluded because they were already receiving NMB and this may have biased against finding a benefit. Given the current body of evidence, it is reasonable to use NMB in patients in whom ventilator synchrony cannot otherwise be achieved, but there are no data to support routine use in patients with moderate to severe ARDS.

Many other pharmacologic therapies have been studied and have failed to demonstrate a clinical benefit. These include surfactant³², N-acetylcysteine³³, and sivelestat³⁴ (a neutrophil elastase inhibitor). Even inhaled nitric oxide, which can provide transient improvement in oxygenation, has been repeatedly demonstrated to offer no outcome benefit and to increase the incidence of adverse renal outcomes^35,36. Despite the lack of effective drug interventions thus far, there remains significant interest in identifying medications that may improve ARDS outcomes either by modifying the inflammatory process or by promoting the re-establishment of functional lung tissue.

The previously referenced HARP-2 trial³⁷ evaluated high-dose simvastatin (80 mg) compared with placebo but found no difference in either ventilator-free days or mortality at 28 days. Subgroup analyses did not identify any separate group that may benefit, including patients with sepsis or with elevated C-reactive protein levels. However, as noted above, a more recent post-hoc re-evaluation of the HARP-2 dataset has used latent class analysis to identify a subgroup characterized by a more inflammatory phenotype of ARDS that may benefit from statin therapy.

Inhaled beta-agonist therapy has been of interest as a treatment for ARDS after the observation in animal models that beta-agonist therapy enhanced alveolar fluid clearance. The ALTA trial³⁸ randomly assigned patients to receive aerosolized albuterol versus placebo but was stopped early for futility with no observed benefit in ventilator-free days at 28 days or mortality at days 60 and 90. In an attempt to gauge whether beta-agonist therapy may reduce inflammation, serum IL-6 and IL-8 levels were measured; however, there was no significant difference found between the two groups in the serum concentration of either of these cytokines. Intravenous delivery of beta-agonist therapy showed some promise of reduced pulmonary edema in early studies. However, the BALTI-2 trial³⁹, which evaluated intravenous salbutamol versus placebo, was terminated early after data showed increased 28-day mortality in the intervention group. There are therefore no data to support the use of either inhaled or intravenous beta-agonist therapy in patients with ARDS.

Perhaps the most studied and discussed pharmacologic intervention for ARDS is steroids. Multiple prior studies have shown mixed results. For example, the LaSRS trial⁴⁰ found that patients with ARDS for at least 7 days had a similar rate of mortality when given methylprednisolone compared with those receiving placebo. However, those patients who received methylprednisolone at least 14 days after the onset of ARDS had an increased rate of mortality compared with the placebo group, suggesting a detrimental effect of steroids in late ARDS. A more recent meta-analysis⁴¹ that included the LaSRS trial found that patients receiving steroids (either methylprednisolone or hydrocortisone) had a higher rate of achieving unassisted breathing by day 28 (80% versus 50%) and had a lower rate of hospital mortality (20% versus 33%) compared with those receiving placebo. A 2019 Cochrane Review on this topic concludes that steroids may reduce 3-month mortality, but the certainty of the evidence is low⁴². While it is possible (and perhaps likely) that a subgroup of patients with ARDS will benefit from steroids, we do not yet know which patients these are. Additionally, given the heterogeneity of the underlying pathology in patients with diagnosed ARDS, it remains possible that those patients who respond, in fact, have an undiagnosed inflammatory lung disease such as organizing pneumonia. At present, steroids are not recommended for routine use in ARDS and further study is needed.

Rescue therapies: extracorporeal membrane oxygenation

Since its development decades ago, extracorporeal membrane oxygenation (ECMO) has been an appealing rescue therapy for patients with severe hypoxemia. Though historically more commonly employed in neonates, the use of veno-venous ECMO in adults with hypoxemic respiratory failure has been on the rise⁴³. This is due in part to the lack of other specific therapies and in part to very promising case reports demonstrating fairly low mortality in patients with severe ARDS who had not responded to other interventions⁴⁴. However, few randomized trials have evaluated the use of ECMO in the management of ARDS and the ones that have done so do not demonstrate a convincing benefit. The CESAR trial⁴⁵ randomly assigned patients with severe ARDS to either standard of care at the institution to which they presented or transfer to a specialty center for ECMO therapy. Although there was a significant difference in the primary outcome, a composite endpoint of survival without significant disability, not all patients who were randomly assigned to ECMO actually received the therapy and the patients in the “usual care” arm did not uniformly receive what is now considered standard of care, including low tidal volume ventilation. Therefore, the degree to which the improved outcomes in the intervention arm were due to ECMO alone is unclear. In order to address these ongoing questions, a large randomized trial that compared true standard of care with early ECMO was recently completed. The EOLIA trial⁴⁶ did not find a significant difference in mortality between the two groups and was stopped early after crossing a futility boundary. Notably, the patients in the conventional management group were allowed to “cross over” and receive ECMO if they failed standard interventions, including prone positioning, paralysis, and inhaled pulmonary vasodilators. It seems unlikely that another large randomized trial of ECMO will be undertaken soon, and although early use of ECMO for all patients with severe ARDS is not supported by the data, it remains an important rescue therapy for patients who are declining despite other interventions. Patients who are likely to have the best outcomes on ECMO are those who have few other organ failures and who are early in their illness (<7 days). Local experience in ECMO cannulation and management is also likely to be an important factor in patient outcomes, and early referral to an ECMO-capable ARDS treatment center is recommended for patients with severe ARDS.