Introduction
Venoarterial extracorporeal membrane oxygenation (VA-ECMO) has been used as a salvage intervention to provide hemodynamic support in patients with cardiogenic shock (CS).1,2 Despite the beneficial effects of VA-ECMO, ECMO-related complications and mortality remain significant. In CS patients under VA-ECMO support, left ventricular (LV) distension with pulmonary oedema commonly occurs may be due to severely decreased LV contractility and increased LV afterload caused by the retrograde ECMO blood flow. Furthermore, this complication promotes myocardial ischemia and ventricular arrhythmia, renders ventricular recovery.3 Various LV unloading strategies have been proposed to reduce the LV distension. However, previous reports have shown heterogeneous results on the effect of LV unloading in regard to survival, according to the causes of CS, the timing of LV unloading, the strategy of LV unloading, and short-term, or long-term mortality.4–6 In particular, the effect of unloading in acute myocardial infarction (AMI)-induced CS patients with VA-ECMO remains controversial. A systematic review reported that the concomitant use of IABP with VA-ECMO was not associated with improved survival in patients with CS caused by AMI.7 However, another systematic review and meta-analysis reported that the concomitant IABP with VA-ECMO was associated with lower mortality in patients with AMI-induced CS.8 Also although the reduction in 30-day mortality from LV unloading by Impella® was consistently observed in both AMI and non-ischemic subgroup,9 the LV unloading effect by IABP or percutaneous ventricular assist device was relatively small in AMI subgroup.10 Based on variable results from previous studies, we hypothesized that AMI-induced CS, which has unique pathology and clinical features, might have different LV unloading effects compared with non-AMI-induced CS. Therefore, we aimed to evaluate whether the effect of LV unloading is different between AMI-induced and non-AMI-induced CS in patients receiving VA-ECMO.
Methods
Study population and data collection
This is a single-centre retrospective cohort study including 153 patients who underwent VA-ECMO between January 2011 and March 2019 (Figure 1). Post-cardiotomy patients, patients who deceased within 24 h after implantation of VA-ECMO, and patients diagnosed as other types of shock were excluded. Finally, we included 128 patients who underwent VA-ECMO for CS. The CS was defined as follows: (1) systolic blood pressure (SBP) < 90 mmHg or vasoactive agents were required to maintain systolic blood pressure >90 mmHg, and (2) accompanying end-organ hypoperfusion was represented by serum lactate levels ≥2.0 mmol/L or clinical pulmonary congestion according to clinical criteria.11 As for the cause of CS, diagnosis of AMI was based on clinical manifestations, electrocardiography, and serum level of cardiac biomarkers (CK-MB and troponin I), and was classified as ST-segment elevation myocardial infarction (STEMI) or non-ST-segment elevation myocardial infarction (NSTEMI) according to the guideline.12,13 CS caused by AMI, including STEMI or NSTEMI was defined as AMI-induced CS and CS with a non-ischemic cause was defined as non-AMI- induced CS. We collected patients' data including demographic, hemodynamic, biochemical, and echocardiographic data from the electronic medical health records at baseline and during the follow-up from the time of VA-ECMO implantation by two specialized cardiologists (J. K. and K. S. L.). If indicated, the prEdictioN of Cardiogenic shock OUtcome foR AMI patients salvaGed by VA-ECMO (ENCOURAGE) score and Simplified Acute Physiology Score (SAPS) II were calculated using published descriptions.14,15 Written informed consent was waived by the Institutional Review Board (IRB) because of the retrospective nature of the study, and the ethical approval on this study was done by the IRB of Seoul National University Hospital. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki.
[IMAGE OMITTED. SEE PDF]
Venoaterial extracorporeal membrane oxygenation procedures and management
All cases judged by the physician as VA-ECMO indications underwent VA-ECMO implantation. All ECMO procedures were performed with percutaneous access and peripheral cannulation. In the case of conventional cardiopulmonary resuscitation (CPR), ECMO was performed while maintaining resuscitation, and ultrasound guidance was provided if necessary. Extracorporeal cardiopulmonary resuscitation (ECPR) was defined as the application of rapid-deployment VA-ECMO to provide circulatory support in patients in whom conventional CPR is unsuccessful in achieving sustained return of spontaneous circulation. All ECPR patients who met the criteria received targeted temperature management according to the institutional protocol. The procedure and management of VA-ECMO followed the extracorporeal life support organization (ELSO) guidelines.1,3
Left ventricular unloading strategies/protocols
First, in order to prevent LV distension with pulmonary oedema, we tried to reduce the VA-ECMO flow while maintaining optimal end-organ perfusion, and administered intravascular diuretics, inotropes, or vasodilators. Furthermore, the need for LV unloading was determined according to guidelines1,3 and using the following clinical and/or imaging factors; arterial pulsatility in the right radial artery, central venous blood oxygen saturation (ScvO2) level, the presence and degree of aortic valve opening, changes in LV dimension on echocardiography, and worsening pulmonary oedema on a chest X-ray. As an LV unloading strategy, IABP or direct left atrial venting was determined according to the preferences of the physician.
Intra-aortic balloon pump (IABP) support
IABP (Arrow AutoCAT2 wave, Teleflex Incorporated, USA) was inserted through a femoral sheath and with the tip located in the 2nd to 3rd intercostal space with a 30 or 40 mL IABP balloon, which was judged according to the patient's height. The support was initiated at a 1:1 inflation–deflation to cardiac cycle ratio, triggered by the R wave of the electrocardiogram. The weaning criteria of IABP were the systolic blood pressure above 90 mmHg or mean arterial pressure of 65 mmHg or above without an inotropic agent after removal of ECMO. The support was decreased at a 1:4 or 1:8 inflation–deflation to cardiac cycle ratio when the weaning program was initiated, and the patients were weaned off of IABP if the hemodynamic condition was stable.
Left atrial venting
The transseptal puncture was guided using transesophageal echocardiography or intracardiac echocardiography and fluoroscopy. The ECMO venous catheter was placed in the left atrium after puncturing from the right atrial side.
Study outcomes
The primary outcome was 90-day all-cause mortality after VA-ECMO implantation. Secondary outcomes were 1-year mortality, duration of VA-ECMO support, length of stay in the intensive care unit (ICU), or rate of heart transplantation. The safety outcomes included VA-ECMO complications such as major bleeding as Bleeding Academic Research Consortium (BARC) type 3 to 5,16 limb ischemia requiring surgical treatment or intervention, acute kidney injury needed renal replacement therapy, and stroke.
Statistical analysis
Categorical data were compared with the chi-square or Fisher exact test as required. Continuous data are presented as mean ± standard deviation or medians (25th–75th percentiles) and their group difference were compared using the student's t-test or the Mann–Whitney test. Cumulative event or mortality rates were estimated using the Kaplan–Meier method and compared using the log-rank test. The Cox proportional hazard model was used to estimate the hazard ratios (HR) and 95% confidence intervals (CI) for 90-day mortality. As a first step, a univariate Cox regression analysis was performed to identify factors affecting mortality in patients with CS receiving VA-ECMO support. Considering that the number of subjects in this study was relatively small, only some variables with a P value less than 0.3 were selected in the univariate Cox regression analysis. Next, multivariable Cox proportional hazard regression analysis was performed to evaluate the LV unloading on mortality while controlling for the effect of some of the variables that appeared to affect mortality as a result of univariate analysis. The effect of LV unloading on mortality was presented with an adjusted hazard ratio (HR) and its 95% confidence interval (CI). The VA-ECMO alone group was used as a reference. All P values <0.05 were considered statistically significant. SPSS version 25.0 (IBM SPSS Statistics, Chicago, Illinois, USA) and MedCalc (Version 20.106) were used for statistical analysis.
Results
Baseline characteristics
The median age of the study population was 66 years, and 67.2% of the patients were male (Table 1). The aetiology of CS was AMI in 55.5% and non-AMI in 45.5% (acute decompensation of chronic HF in 33.6%, acute myocarditis in 7.8%, and others in 3.1%) (Table 1 and Figure 2). LV unloading during VA-ECMO support was performed in 55 (43.0%) patients according to ELSO guideline,1 of which intra-aortic balloon pump (IABP) was used in 48 (87.3%) patients, transseptal venting in three (5.5%) patients, and both of IABP and transseptal venting in four (7.3%) patients. The VA-ECMO with LV unloading group was younger and had a higher platelet count and C-reactive protein level than the VA-ECMO alone group. However, the proportion of ECPR was higher and the median CPR time was longer in the VA-ECMO alone group than VA-ECMO with LV unloading group. Table S1 showed a comparison of baseline characteristics between the AMI-induced and non-AMI-induced CS groups. Compared with non-AMI-induced CS, patients with AMI-induced CS were older, had a higher body mass index, higher prevalence of hypertension, dyslipidemia, and coronary artery disease, and were more likely to have received prior revascularization. The rate of ECPR was higher and the median CRP time was longer in the AMI-induced CS group compared with those of the non-AMI-induced CS. On the contrary, the non-AMI-induced CS group had a higher prevalence of atrial fibrillation, valvular heart disease, and chronic HF. The median LVEF was significantly lower in patients with non-AMI-induced CS than those with AMI-induced CS. The ENCOURAGE mortality risk score and simplified acute physiology score III (SAPS III) are higher in the AMI-induced CS group than in the non-AMI-induced CS group.
Table 1 Baseline characteristics of patients undergoing VA-ECMO for CS
| All patients ( |
ECMO with LV unloading ( |
ECMO alone ( |
||
| Median age (IQR), years | 66 (57–74) | 62 (50–72) | 68 (59–75) | 0.036 |
| Median BMI (IQR), kg/m2 | 23 (22–26) | 23 (21–26) | 24 (22–26) | 0.328 |
| Male sex, n (%) | 86 (67.2) | 39 (70.9) | 47 (64.4) | 0.436 |
| Diabetes mellitus, n (%) | 49 (38.3) | 23 (41.8) | 26 (35.6) | 0.475 |
| Hypertension, n (%) | 60 (46.9) | 26 (47.3) | 34 (46.6) | 0.938 |
| Dyslipidaemia, n (%) | 53 (41.4) | 23 (41.8) | 30 (41.1) | 0.935 |
| Chronic kidney disease, n (%) | 33 (25.8) | 12 (21.8) | 21 (28.8) | 0.374 |
| Stroke, n (%) | 16 (12.5) | 7 (12.7) | 9 (12.3) | 0.946 |
| Revascularization, n (%) | 39 (30.5) | 14 (25.5) | 25 (34.2) | 0.285 |
| CAD, n (%) | 80 (62.5) | 17 (30.9) | 31 (42.5) | 0.181 |
| Prior MI, n (%) | 27 (21.1) | 10 (18.2) | 17 (23.3) | 0.483 |
| Atrial fibrillation, n (%) | 25 (19.5) | 8 (14.5) | 17 (23.3) | 0.217 |
| Valvular disease, n (%) | 21 (16.4) | 9 (16.4) | 12 (16.4) | 0.991 |
| Congestive HF, n (%) | 49 (38.3) | 19 (34.5) | 30 (41.7) | 0.414 |
| Causes of CS, n (%) | 0.860 | |||
| AMI | 71 (55.5) | 31 (56.4) | 40 (54.8) | |
| Non-AMI | 57 (44.5) | 24 (43.6) | 33 (45.2) | |
| Prior CPR, n (%) | 73 (57.0) | 27 (49.1) | 46 (63.0) | 0.115 |
| ECPR, n (%) | 55 (43.0) | 16 (29.1) | 39 (53.4) | 0.006 |
| Median CPR time (IQR)-min | 3.0 (0.0–24.0) | 1.0 (0.0–15.0) | 6.0 (0.0–29.0) | 0.031 |
| Median ENCOURAGE score (IQR) | 19.0 (13.0–24.0) | 19.0 (11.0–25.0) | 20.0 (14.0–25.0) | 0.420 |
| Mean SAPSII score (±SD) | 53.4 ± 23.0 | 50.3 ± 22.4 | 55.9 ± 23.5 | 0.179 |
| Median GCS score (IQR) | 9.0 (2.0–15.0) | 9.0 (2.0–15.0) | 10.0 (2.0–15.0) | 0.596 |
| Median LVEF (IQR)-% | 27.0 (21.0–39.0) | 25.0 (21.0–34.0) | 30.0 (20.0–40.0) | 0.154 |
| Laboratory findingsa | ||||
| WBC, 103/μL | 12.0 (9.0–15.0) | 12.6 (9.2–15.9) | 11.6 (9.0–14.0) | 0.761 |
| Hb, g/dL | 11.1 ± 2.6 | 11.5 ± 2.4 | 10.8 ± 2.7 | 0.135 |
| Platelet, 103/μL | 153.4 ± 74.0 | 169.3 ± 73.6 | 140.9 ± 72.3 | 0.033 |
| AST, IU/L | 218.0 (49.0–560.0) | 204.5 (39.3–840.3) | 240.5 (51.3–529.3) | 0.653 |
| ALT, IU/L | 99.0 (31.0–355.8) | 99.0 (32.8–348.8) | 118.0 (27.5–381.8) | 0.800 |
| BUN, mg/dL | 29.0 (19.0–41.5) | 30.0 (19.0–36.0) | 27.0 (19.0–42.3) | 0.786 |
| Creatinine, mg/dL | 1.5 (1.0–2.0) | 1.6 (1.0–2.0) | 1.3 (1.0–2.0) | 0.821 |
| hs-CRP, mg/dL | 2.0 (0.2–7.0) | 3.0 (1.0–8.4) | 1.0 (0.0–4.9) | 0.032 |
| PT, INR | 1.4 (1.1–1.9) | 1.4 (1.1–2.1) | 1.4 (1.1–1.9) | 0.884 |
| Fibrinogen, mg/dL | 270.9 ± 128.4 | 296.1 ± 141.3 | 251.0 ± 114.4 | 0.056 |
| Lactates, mmol/L | 7.7 (4.0–11.9) | 6.0 (4.1–11.2) | 8.0 (3.9–12.4) | 0.259 |
| LV unloading strategies, n (%) | ||||
| IABP | - | 48 (87.3) | - | - |
| Transseptal venting | - | 3 (5.5) | - | - |
| IABP and transseptal venting | - | 4 (7.3) | - | - |
[IMAGE OMITTED. SEE PDF]
Clinical outcomes
The 90-day mortality rate of the total population was 53.1% in our cohort. There was no significant difference in 90-day mortality between VA-ECMO with LV unloading and VA-ECMO alone (50.9% vs. 54.9%; adjusted HR 0.85, 95% CI 0.49–1.46, P = 0.550; Table 2, Figure 3, Tables S2 and S3). Furthermore, there was no significant difference between the two groups regarding secondary outcomes, including 1-year mortality, duration of VA-ECMO support, and length of ICU stay. However, the VA-ECMO with LV unloading group received more heart transplantation than those with VA-ECMO alone (22.8% vs. 9.6%, P = 0.078) (Table 2).
Table 2 Clinical outcomes according to LV unloading in the total population, non-AMI-induced, and AMI-induced CS
| Outcomes | Total population ( |
Non-AMI-induced CS ( |
AMI-induced CS ( |
|||||||||
| VA-ECMO alone (n = 73) | VA-ECMO + LV unloading (n = 55) | Adjusted HR (95% CI) | VA-ECMO alone (n = 33) | VA-ECMO + LV unloading (n = 24) | Adjusted HR (95% CI) | VA-ECMO alone (n = 40) | VA-ECMO + LV unloading (n = 31) | Adjusted HR (95% CI) | ||||
| Primary outcome | ||||||||||||
| 90-day mortality, n (%) | 40 (54.9) | 28 (50.9) | 0.85 (0.49–1.46) | 0.550 | 20 (60.8) | 8 (33.3) | 0.37 (0.14–0.96) | 0.041 | 20 (50.0) | 20 (64.5) | 1.96 (0.90–4.27) | 0.089 |
| Secondary outcomes | ||||||||||||
| 1-year mortality, n (%) | 41 (56.4) | 29 (52.7) | 0.86 (0.50–1.47) | 0.572 | 21 (64.6) | 8 (33.3%) | 0.37 (0.14–0.98) | 0.046 | 20 (50.0) | 21 (66.7) | 2.11 (0.98–4.56) | 0.060 |
| HTx, n (%) | 7 (9.6) | 12 (22.8) | - | 0.078 | 6 (18.2) | 11 (45.8) | - | 0.039 | 1 (2.5) | 1 (3.2) | - | 1.000 |
| Median ICU LOS (IQR), days | 10 (5–19) | 13 (7–21) | - | 0.568 | 15 (9–29) | 19 (8–25) | - | 0.771 | 7 (4–14) | 8 (4–15) | - | 0.299 |
| Median duration of ECMO (IQR), days | 4 (3–8) | 6 (3–9) | - | 0.073 | 6 (3–9) | 8 (5–12) | - | 0.080 | 3 (2–7) | 4 (3–7) | - | 0.508 |
| Safety outcomes | ||||||||||||
| Major bleeding, n (%)a | 13 (17.8) | 6 (10.9) | - | 0.277 | 6 (18.2) | 2 (8.3) | - | 0.291 | 7 (17.5) | 4 (12.9) | - | 0.595 |
| RRT, n (%) | 36 (49.3) | 25 (45.5) | 0.665 | 19 (57.6) | 12 (50.0) | - | 0.571 | 17 (42.5) | 13 (41.9) | - | 0.962 | |
| Limb ischaemia, n (%) | 9 (12.3) | 8 (14.5) | 0.715 | 6 (18.2) | 3 (12.5) | - | 0.561 | 3 (7.5) | 5 (16.1) | - | 0.254 | |
| Stroke, n (%) | 4 (5.5) | 4 (7.3) | 0.678 | 1 (3.0) | 3 (12.5) | - | 0.167 | 3 (7.5) | 1 (3.2) | - | 0.439 |
[IMAGE OMITTED. SEE PDF]
When patients with CS were divided into AMI-induced CS and non-AMI-induced CS, the risk of 90-day mortality was numerically higher in AMI-induced CS patients who underwent VA-ECMO with LV unloading than those without LV unloading (64.5% vs. 50.0%; adjusted HR 1.96, 95% CI 0.90–4.27, P = 0.089). However, LV unloading significantly reduced 90-day mortality in patients with non-AMI-induced CS compared with VA-ECMO alone (33.3% vs. 60.8%; adjusted HR 0.37, 95% CI 0.14–0.96, P = 0.041; Table 2, Figure 3, Tables S4 and S5). There was a significant interaction between the LV unloading treatment arm and the cause of CS (P for interaction = 0.029). Furthermore, LV unloading combined with VA-ECMO also reduced 1-year mortality only in patients with CS caused by non-AMI (33.3% vs. 64.6%; adjusted HR 0.37, 95% CI 0.14–0.98, P = 0.046). However, there was no statistically significant difference in other secondary outcomes except for the rate of heart transplantation according to LV unloading in the AMI-induced CS and non-AMI-induced CS groups. The non-AMI-induced CS group received more heart transplantation than those with AMI-induced CS ([17/57] 29.8% vs. [2/71] 2.8%, P ≤ 0.001). Patients with VA-ECMO and LV unloading have a significantly higher rate of heart transplantation than those with VA-ECMO alone ([11/24] 45.8% vs. [6/33] 18.2%, P = 0.039) in non-AMI-induced CS group. Furthermore, one ADHF-induced CS patient received LVAD implantation. Safety outcomes, including major bleeding events requiring transfusion and intervention, renal replacement therapy, limb ischemia, and stroke were not significantly different according to LV unloading (Table 2).
In subgroup analysis, there was no significant interaction between LV unloading and various subgroups (age, sex, diabetes status, cardiac arrest, LV ejection fraction, total CPR time, and presence of ECPR) in terms of 90-day mortality (Figure 4).
[IMAGE OMITTED. SEE PDF]
The aetiologies of mortality
To determine how mortality differs between the two groups (AMI-induced CS vs. non-AMI-induced CS) and with or without LV-unloading, we analysed the etiologies of death in detail according to follow-up time. For convenience, we divided the time of death into acute (1–7 days), subacute (8-30 days), and late (>30 days) phases. In our study, the all-cause mortality was higher in the AMI-induced CS group than in the non-AMI-induced CS group. Furthermore, deaths mainly occurred in acute or subacute phases in both groups. The main causes of death at acute and subacute phase in AMI-induced CS group were systemic infection and ventricular arrhythmia (VT/VF) or refractory CS. In particular, despite the use of inotropes or vasopressors with mechanical circulatory support, the AMI-induced CS group had a higher mortality rate due to failure to recover from CS or ventricular arrhythmia (VT/VF) compared with the non-AMI-induced CS group (42.3% vs. 17.5%, P = 0.004). There was no significant difference in the mortality rate between VA-ECMO with LV unloading and VA-ECMO alone in the AMI-induced CS group (64.5% vs. 50.0%, P = 0.221). In contrary, the causes of death in the non-AMI-induced CS group at acute and subacute phase were diverse including bleeding, systemic infection, withdrawal of care, and ventricular arrhythmia or refractory CS. Unlike in the AMI-induced CS group, LV unloading is associated with lower mortality in the non-AMI-induced CS group (33.3% vs. 60.6%, P = 0.042). The mortality rates due to systemic infection (8.3% vs. 18.2%, P = 0.291) and due to refractory CS or ventricular arrhythmia (VT/VF) (8.3% vs. 24.2%, P = 0.119) were numerically lower compared with those supporting only VA-ECMO for CS (Table 3, Tables S6 and S7).
Table 3 Comparison of aetiologies of mortality according to the presence or absence of LV unloading
| Causes of death-no. (%) | CS population with VA-ECMO ( |
Non-AMI-induced CS ( |
AMI-induced CS ( |
||||||
| Non-AMI-induced CS ( |
AMI-induced CS ( |
VA-ECMO with LV unloading ( |
VA-ECMO alone ( |
VA-ECMO with LV unloading ( |
VA-ECMO alone ( |
||||
| Bleeding | 5 (8.8) | 1 (1.4) | 0.088 | 2 (8.3) | 3 (9.1) | 0.920 | 0 (0.0) | 1 (2.5) | 0.375 |
| Systemic infection | 8 (14.0) | 8 (8.5) | 0.397 | 2 (8.3) | 6 (18.2) | 0.291 | 5 (16.1) | 1 (2.5) | 0.079 |
| Withdrawal of care | 5 (8.8) | 2 (2.8) | 0.241 | 2 (8.3) | 3 (9.1) | 0.920 | 1 (3.2) | 1 (2.5) | 0.855 |
| Refractory CS or ventricular arrhythmia | 10 (17.5) | 30 (42.3) | 0.004 | 2 (8.3) | 8 (24.2) | 0.119 | 14 (45.2) | 16 (40.0) | 0.809 |
| Procedure-related complication | 0 (0.0) | 1 (1.4) | 1.000 | 0 (0.0) | 0 (0.0) | NA | 0 (0.0) | 1 (2.5) | 0.375 |
| Total death | 28 (49.1) | 40 (56.4) | 0.477 | 8 (33.3) | 20 (60.6) | 0.042 | 20 (64.5) | 20 (50.0) | 0.221 |
Discussion
In this study, we retrospectively compared the effects of LV unloading during VA-ECMO support between AMI-induced CS and non-AMI-induced CS. Several previous studies have demonstrated the effect of LV unloading in patients receiving VA-ECMO.17–20 Attention is being paid to the emerging concept of ECMO with LV unloading; however, unsolved problems remain. In particular, the optimal timing and strategy need to be determined, and identifying patients who may benefit from LV unloading remains uncertain. In our study, VA-ECMO with LV unloading significantly reduced 90-day all-cause mortality in patients with non-AMI-induced CS, but not in patients with AMI-induced CS. Several studies reported that the concurrent use of extracorporeal life support and IABP cannot improve in-hospital survival in patients with AMI complicated by CS.7,21 Consistent with prior reports, our study showed that LV unloading during VA-ECMO in patients with AMI-induced CS did not significantly reduce 90-day all-cause mortality as compared with the support of VA-ECMO alone.
In the present study, the main causes of CS were AMI (55.5%) and although causes of non-AMI-induced CS were heterogeneous, acute decompensated heart failure (ADHF) (33.6%) was the main cause of non-AMI-induced CS. Their common feature is the fact that the onset of CS is caused by a significant decrease in LV contractility. This leads to a decrease in cardiac output and is accompanied by an increase in LV end-diastolic pressure, pulmonary venous pressure, and systemic vascular resistance. However, there are several important differences regarding the pathophysiological features and underlying hemodynamics between these 2 clinical scenarios.22 Based on this, several possible reasons can be explained why the impact of LV unloading on mortality may differ between AMI-induced CS and non-AMI-induced CS during VA-ECMO support.
First, AMI-induced CS is characterized by severe hemodynamic compromise status due to a rapid decrease of ventricular contractility in previously normal heart function. On the contrary, patients with chronic severe HF develop cardiogenic shock due to progressive ventricular dysfunction. These patients are chronically adapted to reduce cardiac output (CO),23 but CO is very sensitive to changes in afterload and systemic vascular resistance is remarkably higher.22,24 IABP was mainly used as a strategy for LV unloading in this study. IABP counterpulsation improves coronary perfusion via diastolic flow augmentation and reduces LV afterload via end-diastolic pressure-lowering effects. For IABP to work optimally, the pulsatile activity must be preserved to ensure adequate stroke volume to fill the displaced volume in the descending aorta. In addition, LV dilatation in case of chronic HF may ensure a greater stroke volume for any given ejection fraction.24 Malick et al. reported differences in hemodynamic response to IABP application between AMI-induced and ADHF-induced CS. In ADHF-induced CS patients, IABP insertion resulted in a significantly greater change in cardiac output (0.58 ± 0.79 L/min vs. 0.12 ± 1.00 L/min, P = 0.0009) and systemic vascular resistance further decreased (−253.1 ± 493.0 dyn/s/cm−5 vs. 21.3 ± 843.0 dyn/s/cm−5, P = 0.01) compared with in those with AMI-induced CS.25 These findings suggest that IABP may have a positive role regarding LV unloading during VA-ECMO support in patients with non-AMI-induced CS (especially in the case of CS due to exacerbation of chronic HF). However, in the AMI-induced CS with acute hemodynamic compromise caused by severe LV dysfunction, the LV unloading with an IABP during systolic phase may be insufficient. In the current study, the AMI-induced CS group had numerically higher baseline ENCOURAGE mortality risk score and simplified acute physiology score III (SAPS III) than the non-AMI-induced CS group. This finding supports that the AMI-induced CS group has higher CS severity than the non-AMI-induced CS group and this hemodynamic feature implicates that the role of LV unloading of IABP is not optimal in this group. Second, when considering the LV pressure-volume loops for these 2 clinical scenarios, there are important differences in the underlying haemodynamics including the initial and final ejection fractions and degree of compensatory LV dilatation (Figure S1).22 These differences are important, particularly when it comes to considering LV sizes and the need for triggering the introduction of an LV unloading strategy during VA-ECMO. For this reason, even if the indication of LV unloading was the same between the two groups, the timing of LV unloading would be different, which may have affected the results. Third, the ideal IABP balloon volume to achieve adequate diastolic augmentation has to equal the volume of blood in the aorta. The balloon is limited by the volume of blood contained within the aorta just prior to inflation.26 This aortic volume is related to mean arterial pressure, and which is lower in the AMI-induced CS than in the ADHF-induced CS. Therefore, the effect of counterpulsation is probably less in the AMI-induced CS setting. Fourth, other reason for the blunted impact of LV unloading on mortality in AMI-induced CS is the clinical characteristics of patients with AMI-induced CS. AMI-induced CS patients in our study population were relatively older than those with non-AMI-induced CS and had multivessel coronary artery disease including left main coronary artery disease. Despite that complete revascularization is an important prognostic factor in patients with AMI-induced CS, the complete revascularization rate was only 59.2%. These patients are characterized by rapid deterioration of their condition if primary or complete revascularization is not immediately performed. For this reason, the rate of ECPR was significantly higher in patients with AMI-induced CS than those with non-AMI-induced CS in our study. Fifth, the specific LV unloading strategy could have influenced the results. Impella® has not yet been approved in Korea. Among various techniques used for LV unloading in patients with VA-ECMO, IABP remains the most used LV unloading strategy in Korea because of its ubiquitous availability, ease of insertion, and relatively small arteriostomy. However, the effect of IABP with VA-ECMO on CS is still debated. Although the use of IABP counterpulsation in VA-ECMO is thought to enhance systolic unloading and improve the myocardial oxygen supply–demand ratio, IABP only provides a modest LV unloading effect compared with other strategies including the Impella pump, right upper pulmonary/trans-septal catheters, and LV surgical vents.27
In our study, the 90-day mortality rate was 53.1% (68/128). More patients with AMI-induced CS died than those with non-AMI-induced CS. The causes of death were diverse including bleeding, systemic infection, withdrawal of care, procedure-related complications, and refractory CS or ventricular arrhythmia. Unlike in the AMI-induced CS group, LV unloading is associated with lower mortality in the non-AMI-induced CS group. The benefit of LV unloading with IABP or direct LA venting may be driven by the numerical reduction of death from systemic infection (8.3% vs. 18.2%, P = 0.291) and from refractory CS or ventricular arrhythmia (VT/VF) (8.3% vs. 24.2%, P = 0.119) compared with those supporting only VA-ECMO for CS (Table 3, Tables S6 and S7). Furthermore, the rate of heart transplantation was significantly higher in the non-AMI-induced CS group with VA-ECMO support and LV unloading than in those with VA-ECMO only support, but not in the AMI-induced CS group. These findings implicated that even if IABP has a different mechanism from other LV unloading strategies and the effect of LV unloading is modest, IABP may reduce mortality in non-AMI-induced CS during VA-ECMO support by reduced the refractory CS, ventricular arrhythmia and by increasing the chances of the heart transplantation. However, considering that our study is a retrospective observational study and the sample size is small, further studies to verify our results will be needed.
This study has several limitations. First, our analysis was based on a single-centre, retrospective observational study which has not been externally validated. Therefore, the results should be generalized with caution. Although adequate adjustments were made for baseline characteristics in the multivariable models, there was a possibility of confounders and selection bias in a highly selected group of patients. For example, there are significant baseline differences between the AMI-induced CS group and non-AMI-induced CS group. The AMI-induced CS group was older with a higher rate of cardiac co-morbidities and received a higher rate of ECPR with longer CPR times than those with non-AMI-induced CS, and the rate of complete revascularization was not relatively high in this group. These factors known to be associated with high mortality may limit the ability of either ECMO or ECMO with LV unloading to demonstrate any mortality benefit. For these reasons, our result may be only applicable in CS populations without cardiac arrest or ECPR. Second, the sample size was relatively small, and the indications for ECMO and the aetiology of CS were diverse. Third, because the majority of the LV unloading strategy used in our study was IABP, our results cannot be generalizable to CS populations with other LV unloading strategies such as Impella® device. Lastly, this study was conducted in an Asian population. Therefore, we should be cautious in extrapolating the current results to non-Asian CS patients who may have a different profile of CS.
Conclusions
In patients with non-AMI-induced CS, the utilization of LV unloading during VA-ECMO support may reduce 90-day mortality by reducing the refractory CS or ventricular arrhythmia, and by increasing the opportunity for the heart transplantation compared with VA-ECMO alone, but not in those with severe hemodynamically compromised AMI-induced CS.
Funding
Not applicable.
Conflict of interests
The authors declare that they have no conflicts of interest.
Lorusso R, Shekar K, MacLaren G, Schmidt M, Pellegrino V, Meyns B, et al. ELSO interim guidelines for venoarterial extracorporeal membrane oxygenation in adult cardiac patients. ASAIO J 2021;67:827–844. doi:
Guglin M, Zucker MJ, Bazan VM, Bozkurt B, El Banayosy A, Estep JD, et al. Venoarterial ECMO for adults: JACC scientific expert panel. J Am Coll Cardiol 2019;73:698–716. doi:
Assmann A, Beckmann A, Schmid C, Werdan K, Michels G, Miera O, et al. Use of extracorporeal circulation (ECLS/ECMO) for cardiac and circulatory failure–a clinical practice guideline level 3. ESC Heart Fail 2022;9:506–518. doi:
Kowalewski M, Malvindi PG, Zielinski K, Martucci G, Slomka A, Suwalski P, et al. Left ventricle unloading with veno‐arterial extracorporeal membrane oxygenation for cardiogenic shock. Systematic review and meta‐analysis. J Clin Med 2020;9: [eLocator: 1039]. doi:
Al‐Fares A, Pettenuzzo T, Del Sorbo L. Extracorporeal life support and systemic inflammation. Intensive Care Med Exp 2019;7:46. doi:
Char S, Fried J, Melehy A, Mehta S, Ning Y, Kurlansky P, et al. Clinical efficacy of direct or indirect left ventricular unloading during venoarterial extracorporeal membrane oxygenation for primary cardiogenic shock. J Thorac Cardiovasc Surg 2021;40:S406. doi:
Wang D, Chao V, Yap KH, Tan TE. Does concurrent use of intra‐aortic balloon pumps improve survival in patients with cardiogenic shock requiring venoarterial extracorporeal membrane oxygenation? Interact Cardiovasc Thorac Surg 2020;30:312–315. doi:
Vallabhajosyula S, O'Horo JC, Antharam P, Ananthaneni S, Vallabhajosyula S, Stulak JM, et al. Concomitant intra‐aortic balloon pump use in cardiogenic shock requiring veno‐arterial extracorporeal membrane oxygenation. Circ Cardiovasc Interv 2018;11: [eLocator: e006930]. doi:
Schrage B, Becher PM, Bernhardt A, Bezerra H, Blankenberg S, Brunner S, et al. Left ventricular unloading is associated with lower mortality in patients with cardiogenic shock treated with venoarterial extracorporeal membrane oxygenation. Circulation 2020;142:2095–2106. doi:
Grandin EW, Nunez JI, Willar B, Kennedy K, Rycus P, Tonna JE, et al. Mechanical left ventricular unloading in patients undergoing venoarterial extracorporeal membrane oxygenation. J Am Coll Cardiol 2022;79:1239–1250. doi:
Thiele H, Zeymer U, Neumann FJ, Ferenc M, Olbrich HG, Hausleiter J, et al. Intraaortic balloon support for myocardial infarction with cardiogenic shock. N Engl J Med 2012;367:1287–1296. doi:
Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, et al. Fourth universal definition of myocardial infarction (2018). Circulation 2018;138:e618–e651. doi:
Collet JP, Thiele H, Barbato E, Barthelemy O, Bauersachs J, Bhatt DL, et al. 2020 ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST‐segment elevation. Rev Esp Cardiol (Engl Ed) 2021;74:544. doi:
Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993;270:2957–2963. doi:
Muller G, Flecher E, Lebreton G, Luyt CE, Trouillet JL, Brechot N, et al. The ENCOURAGE mortality risk score and analysis of long‐term outcomes after VA‐ECMO for acute myocardial infarction with cardiogenic shock. Intensive Care Med 2016;42:370–378. doi:
Mehran R, Rao SV, Bhatt DL, Gibson CM, Caixeta A, Eikelboom J, et al. Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the bleeding academic research consortium. Circulation 2011;123:2736–2747. doi:
Russo JJ, Aleksova N, Pitcher I, Couture E, Parlow S, Faraz M, et al. Left ventricular unloading during extracorporeal membrane oxygenation in patients with cardiogenic shock. J Am Coll Cardiol 2019;73:654–662. doi:
Cheng A, Swartz MF, Massey HT. Impella to unload the left ventricle during peripheral extracorporeal membrane oxygenation. ASAIO J 2013;59:533–536. doi:
Koeckert MS, Jorde UP, Naka Y, Moses JW, Takayama H. Impella LP 2.5 for left ventricular unloading during venoarterial extracorporeal membrane oxygenation support. J Card Surg 2011;26:666–668. doi:
Barbone A, Malvindi PG, Ferrara P, Tarelli G. Left ventricle unloading by percutaneous pigtail during extracorporeal membrane oxygenation. Interact Cardiovasc Thorac Surg 2011;13:293–295. doi:
Park TK, Yang JH, Choi SH, Song YB, Hahn JY, Choi JH, et al. Clinical impact of intra‐aortic balloon pump during extracorporeal life support in patients with acute myocardial infarction complicated by cardiogenic shock. BMC Anesthesiol 2014;14: [eLocator: 27]. doi:
Rao P, Khalpey Z, Smith R, Burkhoff D, Kociol RD. Venoarterial extracorporeal membrane oxygenation for cardiogenic shock and cardiac arrest. Circ Heart Fail 2018;11: [eLocator: e004905]. doi:
Lim HS, Howell N. Cardiogenic shock due to end‐stage heart failure and acute myocardial infarction: characteristics and outcome of temporary mechanical circulatory support. Shock 2018;50:167–172. doi:
Baldetti L, Pagnesi M, Gramegna M, Belletti A, Beneduce A, Pazzanese V, et al. Intra‐aortic balloon pumping in acute decompensated heart failure with hypoperfusion: from pathophysiology to clinical practice. Circ Heart Fail 2021;14: [eLocator: e008527]. doi:
Malick W, Fried JA, Masoumi A, Nair A, Zuver A, Huang A, et al. Comparison of the hemodynamic response to intra‐aortic balloon counterpulsation in patients with cardiogenic shock resulting from acute myocardial infarction versus acute decompensated heart failure. Am J Cardiol 2019;124:1947–1953. doi:
Parissis H, Graham V, Lampridis S, Lau M, Hooks G, Mhandu PC. IABP: history‐evolution‐pathophysiology‐indications: what we need to know. J Cardiothorac Surg 2016;11:122. doi:
Baldetti L, Gramegna M, Beneduce A, Melillo F, Moroni F, Calvo F, et al. Strategies of left ventricular unloading during VA‐ECMO support: a network meta‐analysis. Int J Cardiol 2020;312:16–21. doi:
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2024. This work is published under http://creativecommons.org/licenses/by-nc/4.0/ (the "License"). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
Aims
Evidence for the effectiveness of left ventricular (LV) unloading in patients who received venoaterial extracorporeal membrane oxygenation (VA‐ECMO) for acute myocardial infarction (AMI) or non‐AMI induced cardiogenic shock (CS) is limited. The aim of the present study was to compare the effect of LV unloading in AMI‐induced and non‐AMI‐induced CS.
Methods and results
This is a single‐centre retrospective observational study of patients with CS undergoing VA‐ECMO from January 2011 to March 2019. Patients were classified as AMI‐induced and non‐AMI‐induced CS. The association of LV unloading with 90‐day mortality in both groups was analysed using Cox proportional hazard regression analysis.
Results
Of the 128 CS patients, 71 (55.5%) patients received VA‐ECMO due to AMI‐induced CS, and the remaining 57 (44.5%) received VA‐ECMO due to non‐AMI‐induced CS. The modality of LV unloading was predominantly with IABP (94.5%). In the AMI‐induced CS group, LV unloading did not reduce 90‐day mortality (adjusted hazard ratio 1.96, 95% confidence interval 0.90–4.27, P = 0.089). However, in the non‐AMI‐induced CS group, LV unloading combined with VA‐ECMO significantly reduced 90‐day mortality (adjusted hazard ratio 0.37, 95% confidence interval 0.14–0.96, P = 0.041; P for interaction = 0.029) as compared with those who received VA‐ECMO alone.
Conclusions
LV unloading with VA‐ECMO may reduce 90‐day mortality compared with VA‐ECMO alone in patients with non‐AMI‐induced CS, but not in AMI‐induced CS.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Details
1 Department of Internal Medicine and Cardiovascular Center, Seoul National University Hospital and Seoul National University College of Medicine, Seoul, Republic of Korea
2 Department of Internal Medicine and Cardiovascular Center, Eulji University Hospital and Eulji University School of Medicine, Daejeon, Republic of Korea
3 Medical AI Co., Ltd, Seoul, Republic of Korea