Clinical Characteristics of COVID-19 Patients Receiving ECMO Treatment for Severe Acute Respiratory Distress Syndrome

The novel coronavirus disease 2019 (COVID-19) caused by
severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)...


Introduction
The novel coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has become a global pandemic at the beginning of 2020 and affected almost all of the human beings on this planet directly or indirectly.
Our health systems, especially in the epicenters, such as Wuhan in China and New York City in the US, faced unprecedented challenges due to the enormous number of cases rising exponentially [1,2].
A report including approximately 72,314 cases from the Chinese Center for Disease Control and Prevention demonstrated that over 5% of COVID-19 patients developed severe pneumonia and had a high risk of acute respiratory distress syndrome (ARDS) [1,3].
Patients with ARDS having high mortality rate need Extracorporeal Membrane Oxygenation (ECMO) to support gas exchange, if their clinical condition cannot be improved by conventional mechanical ventilation. A recent study showed that only half of the COVID-19 patients receiving ECMO support were survival [4]. However, it remains unclear how to identify the patients who may benefit more before the ECMO therapy. In this study, we aimed to summary the factors related to the outcomes of COVID-19 patients receiving ECMO therapy, which may help the health workers in intensive care unit (ICU) make decisions on choosing the most appropriate treatment for a particular COVID-19 patient with ARDS.

Study Design and Participants
This is a two-center, retrospective, observational study

ECMO Therapy Procedure
In this study, COVID-19 was diagnosed based on chest scan and nucleic acid assay according to the World Health Organization Interim Guidelines [5]. ARDS was confirmed using the Berlin definition [3]. The patients received venovenous ECMO support when either partial pressure of oxygen (Pao 2 )/fractional of inspired oxygen (Fio 2 ) <100 mm Hg or power of hydrogen (PH) <7.25 and partial pressure of carbon dioxide (PaCO 2 ) >60 mm Hg over 6 hours. During ECMO therapy, the blood flow and oxygen flow were adjusted to maintain peripheral capillary oxygen saturation >90% and mixed venous oxygen saturation >70%. Pressure controlled ventilation strategy was adopted during ECMO therapy with settings of pressure < 25 cm H2O, positive end-expiratory pressure (PEEP) 10-15 cm H2O, respiratory rate 4-10 breaths per minute, and Fio 2 less than 50%. All the patients accepted heparin continuous IV infusion to maintain activated clotting time (ACT) at 160-200 seconds and activated partial thromboplastin time (APTT) no more than two times of the upper limit of normal.

Statistical Analysis
Continuous variables were presented as mean with standard deviation (SD). Categorical variables were described as frequency with percentage. Fisher's exact test for categorical variables and Wilcoxon-Mann-Whitney U test for continuous variables were used to compare the differences between survivors and non-survivors.
Statistical analyses were performed using SPSS (version 25.0). P value of less than 0.05 was regarded statistically significant.

Characteristics of the Patients with ECMO Therapy
The general characteristics of the 9 COVID-19 patients who received ECMO therapy are described in (Table 1). Of the nine patients, four were survival to hospital discharge, mean age was

Imaging Features
Ground

Dynamic Changes of Variables around ECMO Initiation (±48 hours)
To identify the factors which may be associated with prognosis of COVID-19 patients treated with ECMO, we depicted the dynamic changes of Pao 2 /Fio 2 ratio, lactic acid, lymphocyte count, neutrophilto-lymphocyte ratio (NLR), CRP, D-Dimer, and procalcitonin around ECMO treatment (Figure 1). Expect for Pao 2 /Fio 2 ratio, the measurements 48-hour before ECMO and 48-hour after ECMO were normalized as percent changes of measurements at ECMO day 0. As shown in (Figure 1A), the nine patients had comparable Pao 2 /Fio 2 ratio before ECMO therapy, however, the Pao 2 /Fio 2 ratios of survivors recovered to higher than 150 mm Hg and the Pao 2 /Fio 2 ratios of non-survivors were still lower than 100 mm Hg (P=0.008, ( Table 2). The plasma lactic acid level of the four survivors decreased after 48-hour ECMO support, while most of the non-survivors (4/5) had increased lactic acid after ECMO therapy ( Figure 1B); P=0.048, ( Table 2). All of the four survivors had elevated lymphocyte count after 48-hour ECMO therapy, in contrast, 4/5 of the non-survivors had decreased lymphocyte ( Figure 1C); P=0.048, ( Table 2). The

NLR of the four survivors decreased after 48-hour ECMO support,
however, all of the five non-survivors had increased NLR after ECMO therapy ( Figure 1D); P=0.008, ( Table 2). In addition, the survivors received ECMO support when CRP was in rising period (Day 0 > Before ECMO), however, most of the non-survivors (4/5) received ECMO therapy when CRP was in declining period (Day 0 < Before ECMO) ( Figure 1E); P=0.048, ( Table 2).   As for D-dimer and procalcitonin, no obvious differences were found between the survivors and the non-survivors around ECMO support (Figures 1F & 1G). The absolute values of these measurements are shown in an additional file [see Additional file 1]. According to the results in (Table 2), the combination of ABO blood group, Pao2/Fio2, lactic acid, lymphocyte, and CRP may be related to the outcome of ECMO therapy.

Discussion
In this study, we described the characteristics of COVD-19 patients with ARDS receiving ECMO support from two hospitals and found that 44.4% (4/9) were survival to discharge. Furthermore, our results demonstrated that ABO blood group and patterns of changes of Pao 2 /Fio 2 , lactic acid, lymphocyte count, NLR, and CRP around ECMO initiation (±48h) may be related to the outcome of ECMO therapy in COVID-19 patients. ECMO is considered as the last rescue treatment for COVID-19 patients with ARDS, although the effect of ECMO therapy on the management of COVID-19 remains unclear at the current stage [6,7]. The case fatality rate (CFR) of COVID-19 has been reported to be 4.3% [5]. This number can be as high as 60-70% in critically ill COVID-19 patients with ARDS as the main symptom [8,9]. A recent pooled analysis of 331 COVID-19 patients receiving ECMO therapy suggested a CFR of 46% (95%CI: 34%-59%) [4]. The CFR found in our study was 55.6% (5/9), however, which may be biased by the small sample size. As a comparison, a previous study showed that ECMO support can lower CFR to 21% in patients with 2009 Influenza A (H1N1) ARDS [10].
As the last life-saving rescue strategy, ECMO is resourceintensive, highly specialized, high-priced, and extremely finite compared to the tremendous cases. Therefore, it is urgently needed to identify the potential factors that were associated with prognosis of an individual patient and facilitate the frontline health workers to optimize the use of limited medical resources. product of anaerobic glycolysis and has been used widely as a marker of altered tissue perfusion in critically ill patients [16]. Elevated lactic acid level may reflect inadequate oxygen delivery [17] and is associated with higher mortality rate in critically ill patients [18]. Inconsistent with the above-mentioned findings on Pao 2 /Fio 2 , decreased plasma lactic acid level in the first 48 hours of ECMO therapy mean that the patients may have a higher probability to be survival to discharge eventually.
It has been reported that SARS-Cov may act on T lymphocytes and exacerbate a patient's immune function [19]. COVID-19 is generally accompanied with a high incidence of lymphopenia.
Lymphocyte has been thought as a potential indicator for critical illness of COVID-19 [20]. In this study, we found that increased lymphocyte count and decreased NLR in the first 48 hours of receiving ECMO support, which means that immune function was improved, was associated with favorable prognosis in COVID-19 patients with ARDS. CRP is a non-specific acute-phase biomarker of inflammation, infection, and tissue damage [21]. CRP facilitates clearance of pathogenic microorganisms invading the body through complement activation and enhanced phagocytosis [22].

Conclusion
We reported a proportion of 44.4% (4/9) COVID-19 patients with ARDS receiving ECMO therapy were survival to discharge.
Moreover, we found five factors which may facilitate the doctors in ICU to optimize the very finite ECMO resources: blood group A and a decreasing CRP level before ECMO initiation may be associated with mortality outcome; Pao 2 /Fio 2 rising to higher than 150 mm Hg, increased lymphocyte count, and decreased lactic acid and NLR in the first 48 hours of ECMO support may be associated with survival outcome.

Supplementary Information
Supplementary information providing additional Table S1 in one Microsoft Word file. Additional file 1: Table S1. Changes of clinical characteristics of ECMO support in COVID-19 patients with ARDS.