Lower Vt and Prone Position: Quo Vadis?

Abstract

The burden of the acute respiratory distress syndrome (ARDS) in ICUs (1) has increased dramatically during the coronavirus disease (COVID-19) pandemic. As the management of ARDS is basically supportive, accurate and adequate delivery of mechanical ventilation is of primary importance to prevent further lung damage that could compromise patient outcome. The basic tenet of mechanical ventilation in patients with ARDS is to protect the lung by using low VT and positive end-expiratory pressure (PEEP) (2, 3). In a previous trial (4) and subsequent meta-analyses, delivering lung protective ventilation in the prone position in moderate to severe ARDS was found to be beneficial to the patient outcome (5, 6). Despite its beneficial effects, however, prone positioning is not widely used in patients with ARDS (7). To close the loop between scientific evidence and clinical practice, rigorous assessment and implementation of mechanical ventilation–associated interventions are key. In this issue of the Journal, Sud and colleagues (pp. 1366–1377) show clinically relevant data (8). The authors performed a network meta-analysis of randomized trials to decipher the contribution of VT, PEEP, prone ventilation, highfrequency oscillation, and venovenous extracorporeal membrane oxygenation (VV-ECMO) on hospital mortality in moderate to severe ARDS not related to COVID-19. They enriched the network meta-analysis technique by adding an assessment of the certainty of evidence to its conventional metrics. The study included 34 trials totalizing 9,085 patients with mortality data available. The main finding was that the low VT–prone strategy was significantly better than the low VT strategy alone (relative risk, 0.74 [0.6–0.92]). This comparison also showed a high certainty of evidence. The low VT strategy, considered the standard care reference, was defined as a VT <8 ml/kg predicted body weight and a PEEP amount adjusted according to the ARMA trial (9). Among all the strategies and compared with standard care, low VT–prone was classified as the most effective approach and high VT as the least effective, both with high confidence. VVECMO was rated as possibly among the most effective strategies for very severe ARDS (PaO2/FIO2 ratio<75 mmHg), but the level of confidence was low as only two trials were analyzed. The strengths of the network meta-analysis by Sud and colleagues (8) are that it was conducted using a robust method and included 34 trials. Furthermore, there is no publication bias, inconsistency is minimal, and heterogeneity is low to moderate. The weaknesses are based on the fact that the analysis relied on trials and hence on selected patients. Patients with ECMO, for instance, had more severe hypoxemia at inclusion. Furthermore, the complications were not assessed. However, the rate of complications went down with prone positioning (7) and its benefit to risk ratio then increased. Finally, we do not know whether the results would be the same in a network meta-analysis selecting only patients with COVID-19–related ARDS. It remains under debate whether COVID-19–related ARDS is a true ARDS and whether COVID-19 and non–COVID-19 ARDS share similar pathophysiology (10–16). Whether or not they are similar, in two recent, large cohorts of patients with COVID-19 ARDS, the low VT–prone strategy was used in at least 70% of patients (11, 17). So where do we go from here? First, these recent findings indicate that the implementation of pronation can be increased as the pandemic of COVID-19 has—for unclear reasons—shown. Second, the “low VT” term could be made more precise by personalizing its size according to the driving airway pressure (18). Recent data show that respiratory system elastance (and thus airway driving pressure) predicts the benefit of treatment (in terms of lowering VT) on the risk of death (19). Such findings provide a solid ground to target airway driving pressure (i.e., below 15 cm H2O) rather than VT. Modestly larger VT would be allowed in patients with relatively low respiratory system elastance, whereas very low VT would be set in those with higher respiratory system elastance. Third, the issue of PEEP adjustment is still far from settled. Ideally, it should be individualized to achieve both maximum lung recruitment and minimal lung overdistension. A bedside monitoring tool would of course be needed for this purpose. Fourth, respiratory system elastance does not tell us the whole story. We need to individualize interventions according to the stress applied to the lung parenchyma (i.e., transpulmonary pressure) even though the data do not support how these interventions have been conducted (20). And finally, the use of prone position is or will be undergoing trials in specific circumstances of ARDS. These include the use of prone position in awake spontaneous breathing patients, the use of prone position in patients with ARDS with a PaO2/FIO2 ratio above 150 mmHg, and the simultaneous use of VV-ECMO and prone position in patients with ARDS. The good news is that the solid data we now have provided a good starting point to offer the best possible supportive care to patients with moderate to severe ARDS: a low VT–prone strategy. But regardless of how far we have come, we can still do better. It is our hope that in the near future we will be able to titrate VT and PEEP in prone position according to the unique pathophysiological conditions of individual ARDS lungs.