Prone positioning during extracorporeal membrane oxygenation for severe acute respiratory distress syndrome.
PRO
Building solid evidence in intensive care medicine is challenging, mostly due to patient heterogeneity and concomitant treatment interactions. In the context of acute respiratory distress syndrome (ARDS), only two interventions have demonstrated a positive impact onpatient survival with high quality of evidence. The first isa low-tidal volume, low-plateau pressure ventilation [1],and the second is prone positioning (PP) of patients withmoderate-to-severe hypoxemia [2]. The latter recom-mendation was achieved after a long journey marked bytrials that failed for various reasons, including insufficient sample size, [3] relatively short duration of the PP cycles,[4, 5] patient selection criteria [4–6], and inadequate ven-tilation strategy. Guerin et al. [2] discovered the “optimalformula” for PP, emphasizing its early application com-bined with low-tidal volume ventilation, muscle relaxants, and prolonged sessions lasting at least 16 h. Mostimportantly, they implemented PP only in patients with more severe hypoxemia, defined as PaO2 to FiO 2 ratio below 150 mmHg after a stabilization period with stand-ardized ventilation settings. The most hypoxemic patients usually present a smaller end-expiratory lung volume(“baby lung”) due to a higher amount of collapse in thedependent lung regions [7]. Prone positioning is associated with a redistribution and homogenization of ventilation from ventral to dorsal lung regions, thus mitigating the risk of over distension of non dependent zones and facilitating the reopening of poorly aerated or collapsed dependent regions. These changes result in a reduction of intrapulmonary shunt fraction and in a better matching of ventilation and perfusion, which in turn translates into an improvement of oxygenation.
CON
Following the principles of evidence-based medicine,randomized controlled trials (RCTs) are the corner stone of clinical research, offering the highest level of medical evidence to evaluate the efficacy and safety of therapeutic interventions. RCTs play a crucial role in shaping newtherapeutic approaches by comparing the effectivenessof novel interventions against standard care. Theirstrengths, including unbiased con founder distribution,minimized systematic differences in treatment allocation,and, ideally, the application of blinded procedures,make RCTs an ideal study design in various intensive care research areas. Despite encouraging preliminary findings in observational studies within critical care,results from RCTs, have often contradicted them. The controlled nature of RCTs, such as the Prone PositioningDuring Extracorporeal Membrane Oxygenationin Patients with Severe Acute Respiratory DistressSyndrome (PRONECMO) trial [1], yielded, once again,contradictory results compared to previous observational studies [2, 3]. In the PRONECMO trial involving 170primarily patients affected by coronavirus disease 2019(COVID-19) on venovenous extracorporeal membraneoxygenation (VV-ECMO), prone positioning during ECMO did not yield significant benefits comparedto maintaining a supine position.
Indeed, successful ECMO weaning at day 60 occurred in 38 (44.2%)patients who received at least four prone positioning sessions on ECMO (prone ECMO group) and in 37(44%) patients who were maintained supine (supineECMO group) (subdistribution hazard ratio, 1.11; 95%confidence interval, 0.71–1.75; P = 0.64). Secondaryendpoints, including survival at days 60 and 90, ECMO and mechanical ventilation-free days, and intensive care unit and hospital length of stay, showed no significant differences between the prone and supine ECMO groups. Similarly, there was no statistically significant heterogeneity of treatment effect between the two study groups across any of the pre specified subgroups (Fig. 1).These results contrast with those obtained in more than13 unmatched and matched observational studies and meta-analyses suggesting accelerated ECMO weaningand lower mortality when prone position was applied during ECMO [2]. Furthermore, the PRONECMO results may deter the utilization of this procedure, which has experienced a growing trend in recent years.
NOT SURE
The benefit of prone positioning in moderate to severe acute respiratory distress syndrome (ARDS) has beenwell-established [1]. Improved outcomes have been attributed to reductions in ventilator-induced lung injury(VILI) by decreasing over distention of ventral lung segments and recruiting collapsed dorsal segments [2].Prone positioning may likewise improve hemodynamicsin a subset of patients [3]. Data from observational studies have suggested that similar benefits may be expected when prone positioning is applied during venovenous extracorporeal membrane oxygenation for ARDS [4–7].
In a meta-analysis of 13 studies, including ARDS both related and not-related to coronavirus disease 2019(COVID-19), 28-day survival was significantly higher in those who underwent prone positioning compared tousual supine positioning (risk ratio 1.31, 95% confidenceinterval 1.21–1.41, p < 0.0001) [7]. Similar results were observed when the meta-analysis was restricted tomatched pairs analyses in non-COVID-19-related ARDS.
More recently, Schmidt et al. published thePRONECMO study, a randomized controlled trial (RCT)of early prone positioning versus supine positioning during extracorporeal membrane oxygenation (ECMO)for ARDS [8]. Strengths of the trial include the high rateof enrollment, lack of loss to follow up, performance ofintention-to-treat analysis with limited use of prone positioning during ECMO in the control arm, clinically relevant outcomes that accounted for competing risks,and a highly protocolized weaning process to mitigate bias from the lack of blinding. Participating centers had substantial experience in both prone positioning and ECMO, initiated ECMO largely based on EOLIA trialcriteria [9], and exhibited high rates of adherence tostandard of care conventional management both beforeand during ECMO—nearly all patients received prone positioning prior to ECMO (supine 94.1%, prone 98.8%)and ventilator settings were similar to those used in theintervention arm of the EOLIA trial.