Is extracorporeal membrane oxygenation effective for patients with severe acute respiratory distress syndrome?

This is a challenging study to interpret because of the large number of crossovers (which would tend to dilute or dampen the benefit of extracorporeal membrane oxygenation); the early stopping of the trial; and the clear benefit based on treatment failure, rather than mortality, as the outcome. It would be helpful to have longer term outcomes and outcomes regarding post-treatment exercise capacity and quality of life. This seems like a good opportunity for a meta-analysis, perhaps using patient-level data, which could help provide clarity given the modest sizes of all of the studies on this topic. (LOE = 1b-)

Combes A, Hajage D, Capellier G, et al, for the EOLIA Trial Group, REVA, and ECMONet. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med 2018;378(21):1965-1975.  [PMID:29791822]

Randomized controlled trial (nonblinded)

Government

Concealed

Inpatient (ICU only)

Extracorporeal membrane oxygenation (ECMO) is often used as a last-ditch treatment for adults with severe acute respiratory distress syndrome (ARDS), but its efficacy remains uncertain despite 3 previous trials. This study included adults with severe ARDS who had been using mechanical ventilation for less than 1 week and who were not responding despite maximal ventilator settings for at least 6 hours. The study was powered to require up to 331 patients to detect a difference of 40% vs 60% survival. Ultimately, the study was stopped after 249 patients were randomized to receive either ECMO or usual best care (including prone positioning and use of neuromuscular blocking agents) based on a prespecified definition of futility. Groups were balanced at the beginning of the study with a mean age of 53 years; 71% were men. The median time from intubation was 34 hours in both groups. During the study, there were a large number of crossovers from the usual care group to the ECMO group because of failing standard therapy (35/125 [28%]). The primary outcome was 60-day mortality, which did not differ significantly between groups (35% vs 46%; P = .09; relative risk 0.76; 95% CI 0.55 - 1.04). However, the secondary outcome of treatment failure (defined as death for patients in the ECMO group and death or crossover to ECMO in the control group) was significantly lower in the ECMO group (35% vs 58%; P < .001; number needed to treat = 4). The 90-day mortality was lower, as well, in the ECMO group, but did not achieve statistical significance (37% vs 47%). Harms of ECMO include a significantly longer hospital stay (36 vs 18 days), severe thrombocytopenia (27% vs 16%), and significant bleeding events (46% vs 28%). Rates of pneumothorax, dialysis, and hemorrhagic stroke were similar between groups. Finally, it is not stated whether the outcomes were assessed by investigators masked to treatment assignment.