Results and Discussion:
Prior reports of ECMO use in Covid-19 patients with severe ARDS may need to be considered in the context of an unprecedented pandemic. Our experience demonstrates that ECMO can be successfully employed as a support modality for Covid-19 related ARDS in an appropriately resourced facility with judicious candidate selection. Health systems that were overwhelmed by Covid-19 patients were forced to abandon their normal practices in the presence of such unanticipated clinical demands. Standards of care change, like mass casualty events in the combat theater, as resources and staff are critically stretched. Many of these reports come from health systems operating at surge capacity (19). The case series described above demonstrates that extracorporeal support has had a beneficial role in carefully selected patients.
Of the fifteen Covid-19 ECMO patients, 11 have been successfully liberated from ECMO (Figure 6). Of the eleven that have been decannulated, 9 have been liberated from mechanical ventilation. At the time this article was written, all 11 of the patients removed from ECMO are following commands and have a chance at a good overall outcome (CPC score of 1-2) (Table 6) (20). Of the 9 patients liberated from MV, 4 have been sent home and 5 are recovering on the wards. Two of the patients still in the hospital are recovering from encephalopathy and critical illness polyneuropathy (Figure 7).
The potentially promising outcomes of the Covid-19 ECMO patients exist due to a multiplicity of reasons. The experiences of our colleagues around the world helped systems such as ours prepare for the Covid-19 pandemic and restrict selection criteria for the application of extracorporeal support. By selecting younger healthier patients, limited resources were used on patients more likely to survive with a good outcome. Instead of focusing on each individual as they presented, the focus moved to the collective group (21). Judicious use of ECMO may improve survival in carefully selected candidates (22).
Pre ECMO ventilator management that served to protect the lungs from additional injury may have been another supportive factor in the outcomes. A lung protective strategy of lower volumes and lower plateau pressures was maintained for most of our patients’ pre ECMO ventilator time. As soon as lung protection failed, namely plateau pressures higher than 30cmH2O, ECMO was initiated, thereby minimizing the period of additional lung injury (23). The patients whose lungs experienced higher pressures and volumes prior to ECMO had longer periods of ECMO support. Most patients also received NMB and trials of prone positioning prior to ECMO.
The higher levels of anticoagulation used during ECMO support may have favorably addressed the underlying thrombotic microangiopathy that attends Covid-19 infection (24). Early recognition of an impaired heparin response and fast transition to the alternative agent bivalirudin may have further treated the thrombotic microangiopathy. Early use of pre filter, high flow CRRT did appear to lower inflammatory markers and helped to carefully adjust intravascular volume in a fashion that did not compromise ECMO blood flow. Attempts to use post filter replacement fluid were met with prohibitive levels of filter thrombosis despite regional and systemic anticoagulation. The degree of cytokine removal from the PrismaFlex CRRT machine with the HF1000 filter (Baxter Healthcare Corporation, Deerfield, IL) remains under investigation (25).
Five of the ECMO patients have received compassionate emergency use of convalescent plasma. All tolerated the convalescent plasma infusions well. Three of the five have been liberated from both ECMO and MV, while two remain on full ECMO and MV support. Further investigation is required to assess the efficacy of the convalescent plasma.
The effectiveness of prone positioning in Covid-19 severe ARDS patients prior to ECMO support led to an increase in prone positioning while on ECMO support. If patients failed to show any improvement after 72 hours of ECMO support, a trial of prone positioning was implemented. Patients were also mobilized into a vertical position using the specialty bed. The use of a newly FDA approved subcutaneous EEG may have enabled earlier detection and treatment of epileptiform activity and nonconvulsive seizures. The EEG criteria used to define seizure activity with this new device are evolving, but treatment of the patients’ seizures improved their neurologic exams. Prolonged use of treatment dose anticoagulation after ECMO liberation may have mitigated some of the late thrombotic events described in prior Covid-19 populations.
The limitations of a small case series are clear. Any health systems outcomes are clearly influenced by preparation time and whether their beds are literally overflowing with critically ill patients. Our experience suggests a helpful role for extracorporeal support in select patients; however, the concerns of premature extrapolation of the data in a small series are readily appreciated.
Greater benefits are probable in a high-volume center with an experienced team that has time to prepare, and one that is not completely overwhelmed by a pandemic surge of patients. The rapid sharing of information from our colleagues in Asia, Europe, and the United States was extremely beneficial to those downstream in the pandemic, as care evolved dynamically as new experiences were reported.
Conclusion: In carefully selected patients at a well-resourced, high-volume center with an experienced team, ECMO may be beneficial in patients with Covid-19 severe respiratory failure. Efficacy of EMCO support is greater when applied as soon as optimal evidence-based interventions are unable to support the most critically ill Covid-19 patients.