If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address for reprints: Muhammad Aftab, MD, Division of Cardiothoracic Surgery, Department of Surgery, University of Colorado School of Medicine, Anschutz Medical Campus, 12631 E 17th Ave, C-310, Room 6602, Aurora, CO 80045.
The COVID-19 pandemic presents a high mortality rate amongst patients who develop severe acute respiratory distress syndrome (ARDS). The purpose of this study was to evaluate the outcomes of venovenous extracorporeal membrane oxygenation (VV-ECMO) in COVID–19-related ARDS and identify the patients who benefit the most from this procedure.
Methods
Adult patients with COVID-19 and severe ARDS requiring VV-ECMO support at 4 academic institutions between March and October 2020 were included. Data were collected through retrospective chart reviews. Bivariate and multivariable analyses were performed with the primary outcome of in-hospital mortality.
Results
Fifty-one consecutive patients underwent VV-ECMO with a mean age of 50.4 years; 64.7% were men. Survival to hospital discharge was 62.8%. Median intensive care unit and hospitalization duration were 27.4 days (interquartile range [IQR], 17-37 days) and 34.5 days (IQR, 23-43 days), respectively. Survivors and nonsurvivors had a median ECMO cannulation time of 11 days (IQR, 8-18) and 17 days (IQR, 12-25 days). The average postdecannulation length of stay was 17.5 days (IQR, 12.4-25 days) for survivors and 0 days for nonsurvivors (IQR, 0-6 days). Only 1 nonsurvivor was able to be decannulated. Clinical characteristics associated with mortality between nonsurviors and survivors included increasing age (P = .0048), hemorrhagic stroke (P = .0014), and postoperative dialysis (P = .0013) were associated with mortality in a bivariate model and retained statistical significance in a multivariable model.
Conclusions
This multicenter study confirms the effectiveness of VV-ECMO in selected critically ill patients with COVID–19-related severe ARDS. The survival of these patients is comparable to non-COVID–19-related ARDS.
Venovenous extracorporeal membrane oxygenation in acute respiratory distress syndrome. The outcomes of patients with COVID-19 patients undergoing VV-ECMO are favorable, and survival is comparable to patients without COVID-19 with ARDS. Clinical markers associated with mortality may help guide patient selection for VV-ECMO cannulation and prognostication.
The outcomes of patients with COVID-19 undergoing VV-ECMO are favorable during the first year of pandemic. Their survival is comparable to patients without COVID-19 with ARDS. The utilization of VV-ECMO support should be considered in selected patients. Clinical markers associated with mortality may help guide patient selection for VV-ECMO cannulation and prognostication.
Infection with SARS-CoV-2 causes COVID-19 and results in 15% to 20% of patients developing severe acute respiratory distress syndrome (ARDS).
Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study.
Poor survival with extracorporeal membrane oxygenation in acute respiratory distress syndrome (ARDS) due to coronavirus disease 2019 (COVID-19): pooled analysis of early reports.
in early reports from 2020. Venovenous extracorporeal membrane oxygenation (VV-ECMO) therapy allows carbon dioxide removal and blood oxygenation in patients with severe pulmonary compromise. This temporary extracorporeal circuit serves as a bridge to gradual lung recovery, and possible lung transplantation. It has been beneficial in treating patients with ARDS without COVID-19.
Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome and posterior probability of mortality benefit in a post hoc bayesian analysis of a randomized clinical trial.
Understandably, there has been significant interest in utilizing this therapy in patients with COVID-19. Early in the pandemic, data from China demonstrated poor outcomes with the implementation of VV-ECMO.
demonstrated better survival than initially reported, with a mortality rate of 38%. Given the high resource utilization of VV-ECMO in patients with COVID-19, the purpose of this multicenter study was to evaluate the outcomes of this modality with a focus on identifying the risk factors associated with in-hospital mortality. We hypothesize that mortality after VV-ECMO support in patients with COVID-19 would be comparable to mortality in patients without COVID-19 with severe ARDS (Video Abstract).
Methods
This study involved collaboration of 4 ECMO referral centers to develop a large prospective, observational database analyzing the outcomes of adult patients with COVID-19 with severe ARDS who underwent VV-ECMO support. VV-ECMO cannulation at each institution followed the international ELSO guidelines. Each institution contributed to the ELSO Registry. Contribution to this study was not equal between institutions with the majority of patients coming from 2 of the centers. The institutional review board at each participating institution approved the study protocol (protocol No. 20-1298; February 1, 2019). Given the observational nature of the study, informed consent was waived. Clinical data were collected through comprehensive retrospective reviews of electronic medical records. Elements of the past medical history were abstracted from the admission history and physical note. Vasopressors were defined as norepinephrine and vasopressin, whereas iotropes were defined as dobutamine, epinephrine, and milrinone. Right heart failure was diagnosed by echocardiography. Between March and October 2020, 51 consecutive adult patients with COVID-19 with ARDS were enrolled and placed on VV-ECMO.
Statistical Analysis
We performed a bivariate analysis on 272 pre-ECMO and during-ECMO clinical variables regarding their association with the primary outcome of in-hospital mortality. The χ2 and Fisher exact tests were utilized for evaluating categorical variables, whereas t test or Wilcoxon rank-sum tests were used for continuous variables. Using the 24 statistically significant variables from the bivariate analysis, we then performed a multivariable analysis utilizing logistic regression and forward stepwise selection. All statistical tests were considered significant at a 2-sided P < .05. All analyses were performed using SAS software version 9.4 (SAS Institute Inc).
Results
Patient Baseline Characteristics
VV-ECMO support was performed on 51 consecutive critically ill patients with COVID–19-related ARDS. The mean age was 50.4 years, and 64.7% were men. Most patients were obese (60.8%) with the average body mass index of 33.2 ± 8.6. In-hospital mortality was 37.2%, whereas 62.8% survived to hospital discharge. Significant differences existed in the baseline patient characteristics, medical history, and comorbidities among the nonsurvivors and survivors (Table 1). Patient characteristics associated with in-hospital mortality included increasing age (56.6 vs 46.7 years; P = .0048), pre-ECMO immunosuppression (42.1% vs 9.4%; P = .02), history of central nervous system dysfunction (ie, neurotrauma, stroke, encephalopathy, or seizure disorder) (21% vs 0%; P = .02), and essential hypertension (57.9% vs 28.1%; P = .03).
Table 1Patient characteristics and demographic characteristics of 51 consecutive critically ill patients with COVID-19 and severe acute respiratory distress syndrome who underwent veno-venous extracorporeal membrane oxygenation cannulation, stratified by survivors and nonsurvivors
Clinical and Laboratory Characteristics of Patients Immediately Before VV-ECMO Cannulation
Nonsurvivors also had significant differences in their immediate pre-ECMO cannulation characteristics compared with survivors (Table 2). Inotrope requirement (52.6% vs 15.6%; P = .01) and steroid treatment (52.6% vs 15.6%; P = .0014) in the 24 hours immediately before cannulation were associated with mortality. Regarding pre-ECMO laboratory values, patients who experienced in-hospital death had higher white blood cell counts (18.9 vs 13.0; P = .0096). A lower minimum hemoglobin level was protective for mortality (14.3 vs 11.3; P = .0138). This study also evaluated several markers of inflammation such as C-reactive protein, procalcitonin, interleukin 6, and ferritin; however, only a higher pre-ECMO maximum ferritin was associated with mortality (2774 vs 1266; P = .026).
Table 2Clinical and laboratory data of 51 patients with COVID-19 immediately before initiating venovenous extracorporeal membrane oxygenation (VV-ECMO) for severe acute respiratory distress syndrome
Characteristics
Discharge alive (n = 32)
In-hospital death (n = 19)
P value
Pre-ECMO characteristics
Indication for VV-ECMO
.0629
Hypoxia Pao2/Fio2 <75
5 (15.6)
5 (26.3)
Hypoxia Pao2/Fio2 <100
10 (31.3)
4 (21.1)
Hypoxia Pao2/Fio2 100-150
8 (25)
0 (0)
Hypercapnea pH <7.25, Pco2 >60
0 (0)
2 (10.5)
Hypoxia (Pao2/Fio2) and hypercapnia
2 (6.3)
3 (15.8)
RESP score
.3519
I
2 (6.3)
0 (0)
II
10 (31.3)
6 (31.6)
III
13 (40.6)
5 (26.3)
IV
2 (6.3)
5 (26.3)
V
1 (3.1)
1 (5.3)
Transported on VV-ECMO
2 (6.3)
2 (10.5)
.3530
Cannulation location
.2774
Intensive care unit bedside
31 (96.9)
17 (89.5)
Operating room
1 (3.1)
2 (10.5)
Inotrope <24 h before ECMO
5 (15.6)
10 (52.6)
.01
Vasopressors <24 h before ECMO
17 (53.1)
11 (57.9)
.44
Hyperventilation <24 h before ECMO
12 (37.5)
10 (52.63)
.4012
CPR before VV-ECMO
0 (0)
2 (10.5)
.1587
Bicarbonate infusion
2 (6.3)
1 (5.3)
.8849
Nitric oxide use
3 (9.4)
1 (5.3)
.63
Plasmapheresis <24 h before ECMO
1 (3.1)
1 (5.3)
.0603
Prone positioning <24 h before ECMO
28 (87.5)
15 (78.9)
.42
Anticoagulation before cannulation
.7294
Heparin infusion
17 (53.1)
9 (47.4)
Bivalirudin infusion
1 (3.1)
0 (0)
Neuromuscular blockade
29 (90.6)
17 (89.5)
.89
Nonpulmonary infection
8 (25)
2 (10.5)
.09
Steroid <24 h before ECMO
5 (15.6)
10 (52.6)
.0014
Ventilator days before ECMO
6.2 ± 4.5
3.9 ± 3.1
.08
Epoprostenol use <24 h before ECMO
22 (68.8)
12 (63.2)
.84
Pre-ECMO echocardiography
.1384
Right ventricular dysfunction
3 (9.4)
2 (10.5)
Left ventricular dysfunction
0 (0)
1 (5.3)
Acute kidney injury before ECMO
13 (40.6)
6 (31.6)
.7817
Dialysis before ECMO
1 (3.1)
0 (0)
.7278
Bacterial pneumonia
2 (6.3)
3 (15.8)
.27
Septic shock
1 (3.1)
2 (10.5)
.28
Covid-19 treatments
Convalescent plasma
18 (56.3)
13 (68.4)
.39
Hydroxychloroquine/chloroquine
15 (46.9)
3 (15.8)
.025
Remdesivir
15 (46.9)
12 (63.2)
.26
Janus kinase inhibitor
1 (3.1)
0 (0)
.43
Cytokine blocker
7 (21.9)
4 (21.1)
.95
Steroids
6 (18.8)
13 (68.4)
.0004
Pre-ECMO laboratory values
Absolute neutrophil count
1254.9
2028.2
.5673
Creatinine
0.94 ± 0.26
1.20 ± 1.34
.3302
Lowest pH <24 h before ECMO
7.27 ± 0.087
7.26 ± 0.12
.6004
Highest Paco2 <24 h before ECMO
61.48 ± 23.2
64.64 ± 19.7
.6379
Lowest Pao2 <24 h before ECMO
65.84 ± 18.03
64.29 ± 14.81
.7658
Lowest bicarbonate <24 h before ECMO
27.83 ± 7.2
26.87 ± 5.8
.6456
Lowest Sao2 <24 h before ECMO
85.32 ± 13.6
81.47 ± 23.4
.4883
Creatinine, maximum <24 h before ECMO
2.42 ± 5.1
1.53 ± 1.51
.4933
Lactate, maximum (mmol/L)
24.2 ± 105.3
3.69 ± 4.4
.4897
Total bilirubin, maximum
1.38 ± 1.37
0.88 ± 0.62
.1487
Platelets, minimum <24 h before ECMO
73,806 ± 180,151
13,454 ± 51,053
.2129
White blood cells, maximum <24h before ECMO
13.0 ± 6.9
18.9 ± 7.5
.0096
Hemoglobin, minimum, <24h before ECMO
11.3 ± 2.0
14.3 ± 5.6
.0138
proBNP, maximum
161.3 ± 208.4
5779.6 ± 14,888.9
.2016
Pre-ECMO markers of inflammation
C-Reactive protein, maximum
245.6 ± 154.6
224.6 ± 140.6
.6554
Procalcitonin, maximum
4.65 ± 9.22
7.55 ± 18.1
.5836
Interleukin 6, maximum
99.67 ± 91.2
17.9 ± 11.1
.5192
Ferritin, maximum
1266.9 ± 1085.9
2774 ± 2757.6
.0260
Pre-ECMO ventilator settings
Respiratory rate
29.6 (11.0)
28.9 (8.0)
.83
Tidal volume, maximum
382.8 (136.1)
425.7 (99.6)
.35
Mean airway pressure, maximum
24.3 (3.9)
25.5 (3.8)
.49
Positive end expiratory pressure, maximum
16.1 (3.0)
14.7 (2.8)
.15
Plateau pressure, maximum
30.7 (6.4)
32.0
.85
Values are presented as n (%) or mean ± SD. Bold values indicate P value <.05. Pao2, partial pressure of oxygen; Fio2, inspired oxygen fraction; RESP, respiratory extracorporeal membrane oxygenation survival prediction; CPR, cardiopulmonary resuscitation; Paco2 , partial pressure of carbon dioxide; Sao2, oxygen saturation; proBNP, B-type natriuretic peptide.
Use of steroids <24 hours before ECMO (15.6% vs 52.6%; P = .00140) and steroid treatment in general (18.8% vs 68.4%; P = .0004) were associated with increased mortality. Hydroxychloroquine/chloroquine treatment was associated with survival (46.9% vs 15.8%; P = .025).
ECMO Cannulation Details and Post-ECMO Patient Characteristics
Patients with in-hospital mortality had a significantly higher frequency of cannulation site bleeding requiring transfusion (31.6% vs 9.4%; P = .04) and ECMO oxygenator failure (15.8% vs 0%; P = .02). Higher ECMO sweep at 4 hours postcannulation was also associated with in-hospital mortality (4.7 vs 3.28; P = .02). Furthermore, the levels of biomarkers collected 24-hours after ECMO initiation demonstrated significant differences between nonsurvivor and survivor cohorts. These included elevated C-reactive protein (290.8 vs 196.5; P = .023), D-dimer (15,724 vs 5349; P = .05), and B-type natriuretic peptide (proBNP) (16,411 vs 1185; P = .039) (Table 3).
Table 3Extracorporeal membrane oxygenation (ECMO) cannulation details and post-ECMO characteristics of 51 patients who underwent venovenous ECMO for severe COVID–19-associated acute respiratory distress syndrome
P values are either χ2 or Fisher exact for categorical variables and t test for continuous variables.
During-ECMO characteristics
Initial access cannulation site
.42
Internal jugular, right
18 (56.3)
13 (68.4)
Superior vena cava
0 (0)
1 (5.3)
Common femoral vein, right
9 (28.1)
3 (15.8)
Common femoral vein, left
3 (9.4)
2 (10.5)
Initial return cannulation site
.84
Internal jugular, right
19 (59.4)
12 (63.2)
Superior vena cava
1 (3.1)
0 (0)
Common femoral vein, right
8 (25)
6 (31.6)
Common femoral vein, left
3 (6.3)
2 (10.5)
Access cannulation site conversion
.32
Internal jugular, right
0 (0)
1 (5.3)
Common femoral vein, left
1 (3.1)
0 (0)
Return cannulation site conversion
.39
Internal jugular, right
1 (3.1)
1 (5.3)
Common femoral vein, right
0 (0)
1 (5.3)
Prone positioning
9 (28.1)
6 (31.6)
.89
Anticoagulation
.24
Heparin
29 (90.6)
15 (78.9)
Bivalirudin
3 (9.4)
4 (21.1)
Laboratory values 24 h after cannulation
pH
7.39 ± 0.06
7.39 ± 0.06
.76
Paco2
47.5 ± 7.5
48.7 ± 48.7
.68
Pao2
92.7 ± 36.4
92.9 ± 42.9
.98
Bicarbonate
28.8 ± 5.6
28.6 ± 4.9
.90
Sao2
94.5 ± 3.0
95.7 ± 2.1
.14
Svo2
67.6 ± 12.6
67.3 ± 7.4
.97
Fio2
55.6 ± 23.1
56.5 ± 16.9
.90
Creatinine
1.63 ± 1.5
1.62 ± 1.5
.98
Bilirubin, total
1.59 ± 1.7
1.7 ± 2.2
.84
Sodium, serum
142.1 ± 4.7
138.8 ± 23.7
.58
Platelets
225 ± 83.7
182.6 ± 77.4
.09
White blood cell count
10.7 ± 4.3
13.7 ± 6.8
.10
Hemoglobin
9.86 ± 1.5
9.76 ± 1.3
.63
ProBNP
592.8 ± 934.2
3338.4 ± 5412
.11
ProBNP, maximum during ECMO
1185 ± 1825
16,411 ± 25,059
.039
Troponin, maximum
0.09 ± 0.21
0.29 ± 0.53
.14
During-ECMO markers of inflammation
C-Reactive protein, 24 h postcannulation
154.4 ± 123.6
170.7 ± 120.7
.68
C-Reactive protein, maximum
196.5 ± 120.6
290.8 ± 135.0
.023
Procalcitonin, 24 h postcannulation
3.03 ± 8.6
4.08 ± 5.7
.84
Procalcitonin, maximum
3.14 ± 8.3
6.76 ± 11.5
.47
D-Dimer, 24 h postcannulation
5349.7 ± 7604.7
15,724 ± 2147
.05
D-Dimer, maximum
26,831 ± 27,599
43,929 ± 31,272
.07
Ferritin, 24 h postcannulation
1207 ± 1016.7
1493 ± 824
.40
Ferritin, maximum
1561 ± 1529
12,224 ± 31,747
.12
ECMO settings
Flow at 4 h
4.15 ± 1.03
4.24 ± 1.98
.84
Flow at 24 h
4.35 ± 0.98
4.25 ± 1.07
.74
Sweep at 4 h
3.28 ± 1.55
4.7 ± 2.60
.02
Sweep at 24 h
4.04 ± 2.12
5.44 ± 3.12
.06
Ventilator settings 24 h after cannulation
Respiratory rate
15.1 (4.5)
15.2 (5.7)
.94
Tidal volume
273.9 (98.7)
293.6 (128.7)
.64
Peak pressure
27.5 (3.9)
30.9 (8.9)
.13
Mean airway pressure
17.0 (3.8)
16.8 (2.3)
.86
Positive end expiratory pressure
12.6 (3.2)
11.4 (2.5)
.21
Plateau pressure
27.6 (4.6)
26 (4.6)
.65
Pao2/Fio2
174.3 (103.9)
133.4 (59.6)
.056
Procedures performed during ECMO
Chest tube
3 (9.4)
2 (10.5)
.89
Reintubation
1 (3.1)
1 (3.1)
.70
Therapeutic bronchoscopy
6 (18.8)
6 (31.6)
.30
Blood product use during ECMO
Red blood cell units
7.95 (8.7)
8.93 (10.1)
.75
Platelet
1.93 (2.8)
1.3 (1.5)
.49
Fresh frozen plasma
1.88 (3.6)
1.25 (1.8)
.58
Values are presented as n (row %) or mean ± SD. Bold values indicate P value <.05. Paco2, partial pressure of carbon dioxide; Pao2, partial pressure of oxygen; Sao2, oxygen saturation; Svo2, mixed venous oxygen saturation; Fio2, inspired oxygen fraction; proBNP, B-type natriuretic peptide.
∗ P values are either χ2 or Fisher exact for categorical variables and t test for continuous variables.
Nonsurvivors had significantly higher rates of end-organ dysfunction such as hemorrhagic stroke (36.8% vs 3.1%; P = .0014), right heart failure (15.8% vs 0%; P = .02) and renal failure needing dialysis (63.2% vs 18.8%; P = .0013). Additionally, localized infections resulting in culture-proven infection of a body cavity (positive cultures taken from drained pleural or abdominal fluid) were also associated with in-hospital mortality (31.6% vs 6.3%; P = .014).
Patient Outcomes: Causes of Death and Complications
Among the nonsurvivors, 73.6% died from end-stage respiratory failure, 21.1% died from multisystem organ failure, and 5% died from intracranial hemorrhage. The median total ECMO cannulation time for survivors was 11 days (interquartile range [IQR], 8-18 days) and 17 days (IQR, 12-25 days) for nonsurvivors. The median postdecannulation length of stay for survivors was 17.5 days (IQR, 12.4-25 days) and 0 days (IQR, 0-6 days) for nonsurvivors. The median pre-ECMO ventilator days for survivors was 6 days (IQR, 3-8.6 days) and 4.5 days (IQR, 1-6 days) for nonsurvivors. The median hospital length of stay at the time for ECMO cannulation was 7 days (IQR, 1-9 days) for survivors, and 5 days (IQR, 0-7 days) for nonsurvivors. The median total hospitalization days for survivors were 37 days (IQR, 28-47 days) and 23 days (IQR, 16-37 days) for nonsurvivors. Median intensive care unit days for survivors was 26 days (IQR, 19-38 days) and 22 days (IQR, 16-37 days) for nonsurvivors (Table 4). Among survivors, 14 were discharged home, 8 were discharged to long-term acute care, and 9 were discharged to an acute rehabilitation facility (Table 4). Utilizing a multivariable model, increasing age (odds ratio [OR], 1.156; 95% CI, 1.028-1.3; P = .0157), postoperative dialysis requirement (OR, 20.015; 95% CI, 2.837-141.17; P = .0026) and postoperative hemorrhagic stroke (OR, 58.265; 95% CI, 3.809-891.47; P = .0035) were predictive of in-hospital mortality.
Table 4Patient outcomes and cause of death of 51 patients who underwent venovenous extracorporeal membrane oxygenation (VV-ECMO) for COVID–19-associated acute respiratory distress syndrome
ECMO outcome
Discharge alive (n = 32)
In-hospital death (n = 19)
P value
ECMO cannulation time
11 (8-18)
17 (12-25)
Pre-ECMO ventilator days
6 (3-8.6)
4.5 (1-6)
Pre-ECMO hospital days
7 (1-9)
5 (0-7)
Postdecannulation length of stay (days)
17.5 (12.4-25)
0 (0-6)
Total intensive care unit length of stay (days)
26 (19-38)
22 (16-37)
Total hospitalization days
37 (28-47)
23 (16-37)
Cause of death
End stage respiratory failure
14
Multisystem organ failure
4
Intracranial hemorrhage
1
Hospital discharge location
Long-term acute care
8
Home
14
Acute rehab unit
9
Other
1
Complications during ECMO
Conversion to veno-arterio-venous ECMO
0 (0)
1 (5.3)
.35
Upper extremity deep vein thrombosis
11 (34.4)
2 (10.5)
.16
Lower extremity deep vein thrombosis
4 (12.5)
1 (5.3)
.62
Acute kidney injury
10 (31.3)
7 (36.8)
.69
Renal failure requiring dialysis
6 (18.8)
12 (63.2)
.0013
Pneumothorax
4 (12.5)
4 (21.05)
.42
Hemothorax
0 (0)
1 (5.3)
.19
Secondary bacterial pneumonia
10 (31.3)
7 (36.8)
.68
Positive blood culture
11 (34.4)
6 (31.6)
.86
Positive urine culture
5 (15.6)
2 (10.5)
.78
Positive body cavity fluid culture
2 (6.3)
6 (31.6)
.014
Right heart failure
0 (0)
3 (15.8)
.02
New inotrope requirement
1 (3.1)
3 (15.8)
.11
New vasodilator/antihypertensive agent requirement
6 (18.8)
2 (10.5)
.44
New vasopressor requirement
4 (12.5)
4 (21.1)
.42
Atrial fibrillation
1 (3.1)
1 (5.3)
.70
Ventricular tachycardia
2 (6.3)
0 (0)
.27
Cardiopulmonary resuscitation during ECMO
1 (3.1)
2 (10.5)
.28
Hemorrhagic stroke
1 (3.1)
7 (36.8)
.0014
Seizures
1 (3.1)
0 (0)
.44
Oxygenator failure
0 (0)
3 (15.8)
.02
Cannulation site bleeding
3 (9.4)
6 (31.6)
.044
Gastrointestinal bleeding
8 (25)
3 (15.8)
.44
Retroperitoneal bleed
1 (3.1)
0 (0)
.44
Hematuria
2 (6.3)
0 (0)
.27
Values are presented as median (interquartile range), n, or n (%). Bold values indicate P value <.05.
This study reports the experience of VV-ECMO support for COVID–19-related ARDS at 4 major academic centers in the United States during the first year of pandemic. Fifty-one consecutive patients received VV-ECMO support. Among them, 32 patients were weaned from VV-ECMO and discharged from the hospital alive, whereas 19 patients died. The in-hospital mortality rate of patients who received ECMO therapy during this study was 37.2%.
Early data during the pandemic reported dismal outcomes after VV-ECMO implementation in patients with COVID-19,
Poor survival with extracorporeal membrane oxygenation in acute respiratory distress syndrome (ARDS) due to coronavirus disease 2019 (COVID-19): pooled analysis of early reports.
As the pandemic progressed, the medical community regained confidence in the utility of VV-ECMO. The recent publication of the international ELSO Registry cited a 38% in-hospital mortality consistent with the mortality rate reported in our study and with previously published reports of VV-ECMO use in patients without COVID-19 with ARDS.
Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial.
In our study, essential hypertension was a significant risk factor for in-hospital mortality in patients receiving VV-ECMO. Essential hypertension has been recognized as a risk factor for worsened severity of COVID-19 infection.
hypertension was the most common comorbidity among patients with COVID-19 who developed severe complications and required intubation. In a meta-analysis based on 6 studies, patients with COVID-19 with severe respiratory complications were 2-fold more likely to have primary hypertension.
Multiple explanations for the association between hypertension and COVID-19 severity have been proposed, with the most common being accumulated end-organ damage caused by long-standing hypertension.
Unfortunately, primary hypertension is a complex variable to quantify. In many of these retrospective studies, including our own, there are significant uncertainties about the severity of hypertension, the timing of hypertension diagnoses, and antihypertensive medication adherence.
In this report, we also noted that the use of inotropic drugs and steroids within 24 hours before the cannulation is associated with a higher risk for in-hospital mortality. Hemodynamic instability needing inotropic support is indicative of escalating heart failure and ultimately cardiogenic shock. Consistent with our data, the ELSO VV-ECMO study demonstrated that patients in severe cardiogenic shock requiring veno-arterial ECMO support had a significant association with in-hospital mortality.
Our study found pro-BNP elevation to be associated with mortality. As a marker for heart failure, elevated pro-BNP is congruent with our findings, indicating adverse outcomes in VV-ECMO patients with cardiac dysfunction. This association between elevated pro-BNP and mortality has also been supported by previous COVID-19 reports.
Elevated N-terminal pro-brain natriuretic peptide is associated with increased mortality in patients with COVID-19: systematic review and meta-analysis.
The use of steroids for ARDS has been focus of numerous clinical trials. Recent data from the Dexamethasone Treatment for the Acute Respiratory Distress Syndrome trial (DEXA-ARDS) demonstrated a lower 60-day all-cause mortality (21% vs 36%; P = .005) and increased ventilator-free days (12 vs 7; P < .001).
Accordingly, corticosteroids emerged as an early treatment option for patients with COVID-19, with early observational data from China demonstrating a mortality benefit.
The UK-based Dexamethasone in Hospitalized Patients with Covid-19 trial (RECOVERY) recruited 6425 patients with COVID-19 and showed a significant reduction in mortality (29.3% vs 41.4%) in patients receiving mechanical ventilation who received dexamethasone treatment.
However, this mortality benefit did not extend to patients not receiving ventilation. The Randomized, Embedded, Multifactorial Adaptive Platform Trial for Community- Acquired Pneumonia (REMAP-CAP) trial randomized 403 patients with severe COVID-19 into fixed low-dose hydrocortisone, a shock-dependent hydrocortisone dose, and no steroids groups. The results from this trial's Bayesian analysis found that steroid treatment was likely superior to nontreatment. However, the trial lacked the sample size and granularity to determine the optimal steroid treatment regimen. The Effect of Dexamethasone on Days Alive and Ventilator-Free in Patients With Moderate or Severe Acute Respiratory Distress Syndrome and COVID-19 (CoDEX) trial took place in Brazil and randomized 299 patients with severe ARDS with COVID-19 to high-dose dexamethasone versus usual care alone and found an increase in ventilator-free days with steroid use. The use of steroids in the patients in our study, due to their deteriorating clinical condition, may have a temporal association with their cannulation and indicates severe inflammatory response.
Patients with severe COVID-19 infection can manifest an inflammatory cytokine storm that results in the elevation of several acute phase reactants.
Our study found pre-ECMO elevation of the inflammatory marker ferritin to be associated with mortality. Similarly, a few prior studies have corroborated this finding. A recent meta-analysis of 18 COVID-19 trials found that ferritin levels were significantly higher in patients who eventually required intubation and in those who did not survive hospitalization.
with micro- and macrovascular thromboses being the predominating phenotype. Regardless of etiology, a defining feature of ARDS is airspace fibrin deposition resulting in fibrin-platelet conglomeration and ultimately microthrombi in the pulmonary vasculature.
With increased pulmonary vascular resistance, patients with ARDS can often exhibit significant right heart failure. Our results support this finding of COVID-19 induced hypercoagulability with elevated D-dimer, a by-product of clot dissolution, and right heart strain on echocardiography associated with in-hospital mortality.
Another key finding is the similarity in mortality rates between our study and ECMO to Rescue Lung Injury in Severe ARDS (EOLIA) trial. The EOLIA trial found improved 60-day mortality (41% vs 57%) with the institution of VV-ECMO support in patients with severe ARDS.
The average Pao2 to fraction of inspired oxygen ratio was 74 mm Hg in our study, and the mean Pao2 to fraction of inspired oxygen ratio in the EOLIA trial was 73 mm Hg. Although our study did not strictly measure 60-day survival, our 38.3% mortality is in line with the expected mortality in patients receiving VV-ECMO for ARDS who were not infected with COVID-19.
have noted findings that are discordant with our study. These include the association of elevated body mass index with mortality and the finding that early ECMO support being associated with improved survival. Additionally, these studies have noted a correlation between respiratory ECMO survival prediction score and successful VV-ECMO implementation in patients with COVID-19, but our report did not demonstrate a linear correlation between increasing respiratory ECMO survival prediction score and mortality.
Our study has significant limitations. First, it describes the outcomes of selected 4 academic centers located in the Midwest and Rocky Mountain West with established ECMO programs and a significant cumulative experience. These centers have the resources to efficiently place patients on this therapy and collect and submit patient data during a pandemic. Second, our study does not incorporate long-term outcomes for patients after index hospitalization needing VV-ECMO support. Indeed, many of the ARDS studies
Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial.
have longer follow-up times, between 60 days and 6 months, which limits our ability to directly compare mortality rates. Given the recently described coagulopathies arising in patients with COVID-19,
it is possible that our study underreports the prevalence of many long-term complications arising after COVID-19 infection. Finally, the observational nature of this study and lack of randomization limits our ability to draw definitive conclusions about the comparative efficacy of VV-ECMO as a therapy in patients with COVID-19.
Conclusions
The results of our multicenter registry confirm the potential utility of VV-ECMO in the treatment of critically ill patients with COVID-19 with severe ARDS (Figure 1). Advanced age, immunosuppression, severe inflammatory response with increasing biomarkers, end-organ dysfunction, stroke, heart failure, renal failure, superimposed infection while on ECMO support, and ECMO-related complications are associated with higher mortality. The overall survival of patients with COVID-19 undergoing VV-ECMO is comparable to patients without COVID-19 with ARDS undergoing VV-ECMO. Further prospective studies are needed to investigate a mortality benefit in COVID-19 patients undergoing VV-ECMO cannulation.
Figure 1The outcomes of patients with COVID-19 undergoing veno-venous extracorporeal membrane oxygenation (VV-ECMO) are favorable, and survival is comparable to patients without COVID-19 with acute respiratory distress syndrome (ARDS). Clinical markers associated with mortality may help guide patient selection for VV-ECMO cannulation and prognostication. COVID-19, Coronavirus disease 2019.
The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.
Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study.
Poor survival with extracorporeal membrane oxygenation in acute respiratory distress syndrome (ARDS) due to coronavirus disease 2019 (COVID-19): pooled analysis of early reports.
Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome and posterior probability of mortality benefit in a post hoc bayesian analysis of a randomized clinical trial.
Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial.
Elevated N-terminal pro-brain natriuretic peptide is associated with increased mortality in patients with COVID-19: systematic review and meta-analysis.
00344-8&id=fx1.jpg){kind=link}
00344-8&id=fx2.jpg){kind=link}
00344-8&id=gr1.jpg){kind=link}