Reply to Wengenmayer et al.: Extracorporeal Membrane Oxygenation for Critically Ill Patients with COVID-19–related Acute Respiratory Distress Syndrome: Worth the Effort!

From the Authors:

We read with interest the correspondence from Wengenmayer and colleagues. The authors suggested that we should have adjusted our ventilation strategy under extracorporeal membrane oxygenation (ECMO) to be more protective. As recommended in the Extracorporeal Life Support Organization (ELSO) guidelines (1), we maintained a high positive end-expiratory pressure (PEEP) and reduced Vt to maintain a plateau pressure (PP) under 25 cm H₂O, but we did not drastically reduce the respiratory rate and the driving pressure (ΔP). The measure of these two parameters are indeed associated with mortality at Day 1 of acute respiratory distress syndrome (ARDS) (2) but not the ΔP in patients with obesity (most of our patients) (3). Thus, reducing ΔP by decreasing Vt in patients with obesity could probably not be the main goal when PP remains acceptable. Indeed, the LUNG SAFE (Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure) study (2) did not show any superiority of the ΔP to predict mortality. Furthermore, even if the ΔP value at day 1 was associated with mortality, to date, optimizing this parameter during the following days is not correlated with survival. Knowing the specificity of coronavirus disease (COVID-19)–related ARDS (4) and the high rate of patients with obesity treated in our small cohort (58.8%), one could advance that our strategy might be more protective by preventing overdistension.

Our cannulation strategy is much more a matter of debate: the double-lumen cannulas are indeed not recommended in first intension by ELSO (1) because their positioning can be longer and require the use of an ultrasound system. Regarding oxygenation and decarboxylation, this type of cannula is as efficient as conventional cannulation (5). Our team is experienced in this type of cannulation, limiting the adverse events during cannulation. In view of the morphotype of our patients, a single jugular cannulation facilitated their half-seated position and nursing. Moreover, these cannulas have the advantage of encouraging patient mobilization (5) and potentially limiting the consumption of sedatives, which is not insignificant in the context of a period with work overload. Because this type of cannula is associated with more bleeding (6), we wondered if the high rate of bleeding in our series is facilitated by the cannula, anticoagulation, or the transfusion strategy. Our transfusion target is consistent with ELSO guidelines (1). Concerning the anticoagulation, neither of the two patients with serious hemorrhagic events were overanticoagulated, and the five other patients were transfused on minor bleedings or hemolysis without a negative impact on patient prognosis. On the other hand, we reported two oxygenator thrombosis and three thromboembolic events. Considering the high incidence of thrombotic events in patients with COVID-19 and the ELSO guidelines (1), our anticoagulation target seems to be reasonable.

In our series, two patients died of refractory ARDS with pulmonary fibrosis making the respiratory weaning impossible after decannulation. Two patients developed refractory septic shock with a predominance of vasoplegia, making conversion to venoaterial ECMO (VA-ECMO) ineffective. One patient died during cannulation of cardiac tamponade, and one was on VA-ECMO. Thus, optimizing the support during the time either by converting to VA-ECMO or adding a second cannula would not have modified the mortality of our case series. It is important to note that the context of pandemic-induced work overload and the patients’ management by interim intensivists who were not used to taking care of patients with ARDS with ECMO may explain some intensive care management difficulties and suboptimal ventilator settings.

In conclusion, in the context of the pandemic, we have chosen a mastered management of our patients. However, ECMO implantation in refractory ARDS related to COVID-19 allowed more protective ventilation parameters, improving patient status. Our results highlighted a preference for an adaptation of ventilator parameters on the PP and moderate PEEP in this specific series characterized by more obese patients and 65% survival in the ICU.