Vasoactive-Inotropic Score as a Determinant of Timely Initiation of Venoarterial Extracorporeal Membrane Oxygenation in Patients With Cardiogenic Shock

Abstract

Background: The predictive role of the vasoactive-inotropic score (VIS) for clinical outcomes after venoarterial extracorporeal membrane oxygenation (VA-ECMO) in patients with cardiogenic shock is not well known. This study investigated the predictive value of VIS on in-hospital outcomes and the determination of optimal timing for the initiation of VA-ECMO.

Methods and Results: Overall, 160 patients with cardiogenic shock requiring VA-ECMO who were treated between December 2012 and August 2018 were analyzed. The in-hospital outcomes according to VIS were compared. Pre-ECMO VIS had an area under the receiver-operating characteristic curve (AUC) of 0.60 (P=0.03) for the prediction of in-hospital death. When the patients were divided into the high (≥32) and low (<32) VIS groups, the high VIS group had a higher rate of in-hospital death (P=0.002) and a lower rate of ECMO weaning (P=0.004). The difference in in-hospital death according to VIS was significant only in patients with a cardiogenic shock of non-ischemic etiology (P=0.01). Extracorporeal cardiopulmonary resuscitation (hazard ratio [HR], 1.99), age (HR, 1.02), pre-ECMO lactate (HR, 1.06), and VIS ≥32 (HR, 2.46) were independently predictive of in-hospital death.

Conclusions: Among patients with cardiogenic shock requiring VA-ECMO, the initiation of VA-ECMO before reaching high VIS (≥32) showed better in-hospital outcomes, suggesting that VIS may be a potential marker for determining the initiation of hemodynamic support with VA-ECMO.

Cardiogenic shock is a devastating condition that leads to major morbidity or death; however, the management and clinical outcome of cardiogenic shocks remains a challenging issue. Patients with cardiogenic shock often need vasopressor or inotropic agent support to maintain adequate levels of blood pressure and tissue oxygenation. However, inotropic support is commonly insufficient for adequate perfusion, thus leading to the use of mechanical circulatory support (MCS). Extracorporeal membrane oxygenation (ECMO) is one of the most commonly used MCS devices for the management of cases with refractory cardiogenic shock sustaining tissue hypoperfusion despite high doses of vasoactive or inotropic agents. Furthermore, adequate hemodynamic support with venoarterial ECMO (VA-ECMO) is reported to improve the clinical outcomes of patients suffering from cardiogenic shock.^1,2

Editorial p 695

However, the optimal timing of the initiation of VA-ECMO support is difficult to determine in real-world practice. Several studies reported that the initiation of VA-ECMO before the development of organ dysfunction led to favorable outcomes.^3–5 Importantly, the clinical outcome of patients suffering from cardiogenic shock is directly related to the amount of vasoactive or inotropic agents used, and a high dose of catecholamine may cause vasoconstriction and subsequent aggravation of end-organ perfusion and multiorgan dysfunction, which further worsen clinical outcomes.^6–8 However, most prognosis prediction models including the amount of vasoactive drug use are calculated after the initiation of VA-ECMO; furthermore, they are often too complicated to be applied in real-world settings and cannot help determining the adequate time-point for the initiation of VA-ECMO support.^3,4

The vasoactive-inotropic score (VIS) represents the amount of vasopressors or inotropic agent use and has been reported to predict adverse outcomes in various clinical settings including pediatric, adult post-cardiotomy shock, heart transplantation, and adult cardiogenic shock.^9–13 As such, the VIS may be useful as a clinical marker for determining the timing of the initiation of VA-ECMO before the development of multiorgan failure. Therefore, we investigated the predictive role of VIS in patients with medically intractable cardiogenic shock on the clinical outcomes after the initiation of VA-ECMO.

Methods

Study Design and Outcomes

We retrospectively gathered the data of patients who were aged >18 years and had a cardiogenic shock requiring VA-ECMO between December 2012 and August 2018 at Asan Medical Center (Seoul, South Korea). Those who were diagnosed with a shock of an etiology other than cardiogenic or received venovenous ECMO were excluded. Cardiogenic shock was defined as: (1) presence of a clinical (e.g., altered mental status or dizziness, diaphoresis, cold skin or extremities, systolic blood pressure <90 mmHg for ≥30 min, or requirement of infusion of catecholamines to maintain a systolic blood pressure >90 mmHg, urine output <30 mL/h) and biochemical (e.g., serum lactate >2.0 mmol/L) evidence of tissue hypoperfusion; (2) presence of primary cardiac dysfunction documented by 2-dimensional transthoracic echocardiography; and (3) exclusion of non-cardiac etiology. Medical records including baseline demographics, etiology of the cardiogenic shock, biochemical results, and VIS before and after the application of VA-ECMO were reviewed. Clinical presentation of cardiogenic shock was divided into 3 clinical scenarios: (1) acute setting including acute myocardial infarction, myocarditis, stress-induced cardiomyopathy, cardiac tamponade, acute right ventricular failure, and acute valvular dysfunction; (2) acute decompensation of a previously diagnosed heart failure; and (3) post-cardiotomy shock.

VIS was calculated as follows: dobutamine dose (μg/kg/min) + dopamine dose (μg/kg/min) + norepinephrine dose (μg/kg/min) × 100 + epinephrine dose (μg/kg/min) × 100 milrinone dose (μg/kg/min) × 10 + vasopressin dose (unit/kg/min) × 10,000.¹⁰ In order to determine the optimal timing of VA-ECMO application, we focused on the pre-ECMO VIS (immediately before the application of VA-ECMO), in terms of how we should use a high degree of a vasoactive or inotropic agent before applying VA-ECMO.

The main outcome of the study was in-hospital death. The secondary outcomes were the rate of ECMO weaning, duration of hospital and intensive care unit (ICU) stay, and predictive factors for in-hospital death. This study was approved by the ethical review board of the Asan Medical Center, Seoul, Korea.

Statistical Analysis

Continuous variables were checked with the Kolmogorov-Smirnov normality test and analyzed using the Student’s t-test or Mann-Whitney U-test, as appropriate, and are presented as mean±standard deviation. Categorical variables were analyzed using the Chi-squared test or Fisher’s exact test, as appropriate, and are presented as numbers with percentages. Receiver-operating characteristics (ROC) analysis was used to determine the optimal cut-off point of VIS for predicting in-hospital death. The areas under the ROC curve (AUC) are presented with 95% confidential intervals (CI). The event rates of in-hospital death were estimated using the Kaplan-Meier method, and the differences were compared using the log-rank test. Cox proportional hazard analyses were performed to calculate the hazard ratios (HRs) of in-hospital death, with adjustment for variables that had clinical relevance or statistical significance (defined as P<0.10 on univariable Cox regression analyses). The proportional hazards assumption was tested by examining the log (−log survival) curves and partial (Schoenfeld) residuals. The length of hospital and ICU stay was compared with the log-transformed Student’s t-test. All comparisons were 2-sided, and P values <0.05 were regarded as statistically significant. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 22.0 (IBM Corp., Armonk, NY, USA).

Results

Baseline Characteristics

A total of 160 patients (mean age: 58.2±14.3 years, 76.3% male, 48.1% ischemic etiology) were diagnosed with cardiogenic shock. The clinical presentation and specific etiology of cardiogenic shock for the patients are presented in Supplementary Table 1. The study population received VA-ECMO at a median 2 days (interquartile range [IQR], 0.0–8.6) from hospital admission and was maintained during for median of 4.8 days (IQR, 2.2–10.5). The median VIS before VA-ECMO was 31 (IQR, 8–62) and the time-dependent change of VIS before and after VA-ECMO is presented in Figure 1. The most commonly used vasoactive or inotropic agent was norepinephrine, followed by dopamine and dobutamine (Table 1). ROC analysis showed that a VIS before VA-ECMO had an AUC of 0.60 (95% CI, 0.51–0.69; P=0.03) for in-hospital death (Figure 2); specifically, a VIS of 32 had a sensitivity of 61.6% and a specificity of 66.2% for predicting in-hospital death, and this value was used for stratification.

Figure 1.

Change of vasoactive-inotropic score before and after initiation of VA-ECMO. (A) overall population; (B) high VIS group; (C) low VIS group. VA-ECMO, venoarterial extracorporeal membrane oxygenation; VIS, vasoactive-inotropic score.

Table 1. Baseline Characteristics of Patients According to the Vasoactive-Inotropic Score

Variables	Overall (N=160)	VIS ≥32 (N=79)	VIS <32 (N=81)	P value
Age, years	58.2±14.3	58.1±15.4	58.4±13.2	0.92
Body mass index, kg/m2	23.6±3.8	23.0±3.7	24.2±3.8	0.03
Male	122 (76.3)	57 (72.2)	65 (80.2)	0.23
Hypertension	68 (42.5)	32 (40.5)	36 (44.4)	0.61
Diabetes	60 (37.5)	33 (41.8)	27 (33.3)	0.27
Prior coronary artery disease	73 (45.6)	33 (41.8)	40 (49.4)	0.33
Prior PCI	40 (25.0)	20 (25.3)	20 (24.7)	0.93
Prior cardiac surgery	29 (18.1)	13 (16.5)	16 (19.8)	0.59
Chronic kidney disease	32 (20.0)	18 (22.8)	14 (17.3)	0.38
Renal replacement therapy	7 (4.4)	4 (5.1)	3 (3.7)	0.72
Mechanical ventilation*	77 (48.1)	51 (64.6)	26 (32.1)	<0.001
Intra-aortic balloon counterpulsation	11 (6.9)	5 (6.3)	6 (7.4)	0.79
Extracorporeal CPR	80 (50.0)	27 (34.2)	53 (65.4)	<0.001
Non-ischemic etiology	83 (51.9)	45 (57.0)	38 (46.9)	0.20
Clinical presentation				0.09
Acute cardiogenic shock	98 (61.3)	44 (55.7)	54 (66.7)
Acute decompensation of chronic heart failure	57 (35.6)	34 (43.0)	23 (28.4)
Post-cardiotomy shock	5 (3.1)	1 (1.3)	4 (4.9)
At the time of ECMO
Hemoglobin, g/dL†	11.6±2.9	11.1±2.8	12.2±3.0	0.03
Prothrombin activity, %†	58.3±29.6	51.9±29.1	65.3±28.7	0.005
Creatinine, mg/dL†	1.6±1.1	1.8±1.1	1.4±0.9	0.02
eGFR, mL/min/1.73 m2	56.4±27.1	48.2±24.9	64.3±26.9	<0.001
Systolic BP, mmHg	63 (0–78)	68 (52–74)	55 (0–81)	<0.001
Diastolic BP, mmHg	39 (0–50)	40 (27–49)	35 (0–50)	<0.001
Heart rate, beats/min	106±29	100±24	108±31	0.24
Lactate, mmol/L	8.2±4.5	8.9±4.4	7.5±4.6	0.07
Lactate, mmol/L (6 h after ECMO)	6.6±4.4	7.3±4.8	6.0±4.0	0.08
ECMO variables
Duration, h	115 (53–251)	123 (53–265)	111 (50–225)	0.78
ECMO flow, L/min	4.1±0.5	4.0±0.5	4.2±0.5	0.03
Vasoactive-inotropic score
VIS	31 (8–62)	62 (45–74)	8 (0–20)	<0.001
Duration from initial need for vasoactive or inotropic agent to ECMO, days	1 (0–4)	1 (0–8)	1 (0–4)	0.20
Dobutamine	65 (40.6)	46 (58.2)	19 (23.5)	<0.001
Dose, μg/kg/min	12.5±5.7	14.4±5.3	7.8±3.7	<0.001
Dopamine	78 (48.8)	49 (62.0)	29 (35.8)	0.001
Dose, μg/kg/min	12.9±8.2	15.3±8.1	8.9±6.8	0.001
Epinephrine	54 (33.8)	42 (53.2)	12 (14.8)	<0.001
Dose, μg/kg/min	0.15±0.11	0.17±0.12	0.07±0.05	<0.001
Norepinephrine	107 (66.9)	76 (96.2)	31 (38.3)	<0.001
Dose, μg/kg/min	0.27±0.16	0.33±0.14	0.11±0.05	<0.001
Vasopressin	49 (30.6)	44 (55.7)	5 (6.2)	<0.001
Dose, IU/kg/min	0.05±0.03	0.05±0.03	0.04±0.04	0.60
Milrinone	1 (0.6)	0 (0.0)	1 (1.2)	1.00
LVEF, %‡	27.2±17.9	25.7±15.7	28.7±19.9	0.31

Data are shown as mean±SD, median (interquartile range), or n (%). *Use of mechanical ventilation was defined as the adoption of invasive mechanical ventilation at least an hour before the initiation of venoarterial extracorporeal membrane oxygenation (VA-ECMO). ^†Data of hemoglobin, prothrombin time, and creatinine were available in 154 (96.3%), 149 (93.1%), and 153 (95.6%) patients, respectively. ^‡Echocardiographic data before the initiation of ECMO were available in 125 (78.1%) patients. BP, blood pressure; CPR, cardiopulmonary resuscitation; ECMO, extracorporeal membrane oxygenation; eGFR, estimated glomerular filtration rate; LVEF, left ventricular ejection fraction; PCI, percutaneous coronary intervention; VIS, vasoactive-inotropic score.

Figure 2.

ROC curve of the pre-ECMO vasoactive-inotropic score for in-hospital death in patients (N=160) with cardiogenic shock who received VA-ECMO. ROC, receiver-operating characteristics; VA-ECMO, venoarterial extracorporeal membrane oxygenation.

When the patients were divided according to the VIS value of 32, baseline characteristics were generally comparable between the 2 groups, except for the higher proportion of patients with extracorporeal cardiopulmonary resuscitation (eCPR) in the low (<32) VIS group (n=81) compared with the high (≥32) VIS group (n=79) (65.4% vs. 34.2%; P<0.001; Table 1). In addition, the laboratory results prior to VA-ECMO showed significant differences in that the high VIS group had lower values of hemoglobin and prothrombin activity and a higher value of creatinine. Serum lactate was measured at a median of 30 min before the initiation of VA-ECMO; the mean level of pre-ECMO lactate was 8.2±4.5 in the entire cohort and did not show a significant inter-group difference (P=0.07). Serum lactate after 6±2 h from the initiation of VA-ECMO was also not significantly different between the high and low VIS groups (7.3±4.8 vs. 6.0±4.0; P=0.08).

In-Hospital Outcomes

The median length of hospital and ICU stay was 24 (IQR, 9–70) and 13 (IQR, 5–27) days, respectively. The median length of hospital stay (29 vs. 21 days, P=0.79) and ICU stay (16 vs. 11 days, P=0.31) were not different between the high and low VIS groups (Table 2). A total of 83 (52.9%) patients were successfully weaned off from VA-ECMO, and the rate of VA-ECMO weaning was significantly lower in the high VIS group (P=0.004), which was mostly driven by the spontaneous recovery of myocardial function (Table 2). In-hospital death occurred in 86 (53.8%) patients, of whom the median time to death was 7 hospital days (IQR, 1–17) after the application of VA-ECMO. The cumulative rate of in-hospital death was significantly higher in the high VIS group (65.1%) than in the low VIS group (37.5%; P=0.002) (Figure 3). The overall rate of complications related to VA-ECMO was not significantly different between the 2 groups (P=0.65) (Table 2).

Table 2. Event Rates of In-Hospital Outcomes for Patients According to the Vasoactive-Inotropic Score

	Overall (N=160)	VIS ≥32 (N=79)	VIS <32 (N=81)	P value
Clinical outcomes
Length of hospital stay, days	24 (9–70)	29 (9–76)	21 (10–59)	0.79*
Length of ICU stay, days	13 (5–27)	16 (6–34)	11 (4–24)	0.31*
Weaning from VA-ECMO	83 (52.9)	32 (40.5)	51 (60.0)	0.004†
Spontaneous recovery	61 (38.1)	21 (26.6)	40 (49.4)	0.003†
Bridge to heart transplantation	20 (12.5)	11 (13.9)	9 (11.1)	0.59†
Bridge to LVAD	2 (1.3)	0 (0.0)	2 (2.5)	0.50†
In-hospital death	86 (53.8)	53 (67.1)	33 (40.7)	0.002
Complications
Any complication	68 (42.5)	35 (44.3)	33 (40.7)	0.65†
Bleeding	45 (28.1)	22 (27.8)	23 (28.4)	0.94†
Access site	11 (6.9)	7 (8.9)	4 (17.4)
Lung	8 (5.0)	4 (5.1)	4 (17.4)
Gastrointestinal	8 (5.0)	5 (6.3)	3 (13.0)
Brain	3 (1.9)	2 (2.5)	1 (4.3)
Others	15 (9.4)	4 (5.1)	11 (47.8)
Limb ischemia	9 (5.6)	5 (6.3)	4 (4.9)	0.74†
Oxygenator failure	16 (10.0)	8 (10.1)	8 (9.9)	0.96†
Ischemic stroke	6 (3.8)	2 (2.5)	4 (4.9)	0.68†
Systemic thromboembolism	5 (3.1)	4 (5.1)	1 (1.2)	0.21†

Data are shown as n (%) or median (interquartile range). *Calculated using a Student’s t-test from log-transformed data. ^†Calculated using a chi-squared or Fisher’s exact test, as appropriate. ICU, intensive care unit; LVAD, left ventricular assisted device; VIS, vasoactive-inotropic score; VA-ECMO, venoarterial extracorporeal membrane oxygenation.

Figure 3.

Event curves of in-hospital death according to the vasoactive-inotropic score. (A) overall population; (B) patients with non-ischemic etiology; (C) patients with ischemic etiology. VIS, vasoactive-inotropic score.

When the patients were divided into 2 subgroups according to the etiology for cardiogenic shock (i.e., ischemic or non-ischemic), the proportion of patients with eCPR for initiation of VA-ECMO was higher among those with ischemic etiology (61.0% vs. 39.8%, P=0.007) (Supplementary Table 2). Among the patients with ischemic etiology (n=77), the risk of in-hospital death was not significantly different (P=0.08) between the high and low VIS groups (Figure 3). In contrast, among those with non-ischemic etiology, the high VIS group had a significantly higher risk of in-hospital death (P=0.01). Those with ischemic etiology showed a higher VA-ECMO weaning rate in the low VIS group than in the high VIS group (P=0.03) (Supplementary Table 3). Those with non-ischemic etiology showed no significant difference in VA-ECMO weaning rate (P=0.06), although the low VIS group had a non-significant trend to favor a higher weaning rate.

Multivariable analyses with a Cox regression model showed that VIS ≥32 (HR, 2.46; 95% CI, 1.51–4.01; P<0.001), eCPR (HR, 1.99; 95% CI, 1.22–3.24; P=0.006), age (HR, 1.02; 95% CI, 1.01–1.04; P=0.008), and lactate level before VA-ECMO (HR, 1.06; 95% CI, 1.00–1.11; P=0.048) were independently predictive of in-hospital death (Table 3).

Table 3. Univariable and Multivariable Analyses for In-Hospital Death

Variables	Univariable analysis		Multivariable analysis
	HR (95% CI)	P value	HR (95% CI)	P value
Extracorporeal CPR	1.85 (1.21–2.84)	0.005	1.99 (1.22–3.24)	0.006
Age	1.02 (1.004–1.04)	0.01	1.02 (1.01–1.04)	0.008
Male	0.94 (0.58–1.54)	0.82
Hypertension	1.58 (1.03–2.41)	0.04	N/S	N/S
Diabetes	1.26 (0.82–1.94)	0.29
Coronary artery disease	0.97 (0.63–1.48)	0.88
Chronic kidney disease	1.46 (0.90–2.37)	0.13
eGFR*	0.996 (0.99–1.004)	0.37
Lactate, pre-ECMO	1.08 (1.03–1.14)	0.001	1.06 (1.00–1.11)	0.048
Left ventricular EF	1.00 (0.98–1.01)	0.48
VIS ≥32	1.96 (1.26–3.03)	0.003	2.46 (1.51–4.01)	<0.001

*Data of eGFR were gathered immediately before the initiation of ECMO. CI, confidential interval; CPR, cardiopulmonary resuscitation; ECMO, extracorporeal membrane oxygenation; EF, ejection fraction; eGFR, estimated glomerular filtration rate; HR, hazard ratio; VIS, vasoactive-inotropic score.

Discussion

By investigating the association between VIS and in-hospital outcomes in patients with cardiogenic shock requiring VA-ECMO, we found that: (1) a high (≥32) VIS before the initiation of VA-ECMO was associated with the risk of in-hospital mortality; and (2) this difference in the in-hospital mortality according to the VIS was only significant in patients with cardiogenic shocks of non-ischemic etiology.

The VIS was reported to be predictive of in-hospital outcomes in pediatric patients undergoing cardiac surgery.^10,11 Furthermore, the VIS was also able to predict adverse in-hospital outcomes among adult patients with cardiogenic shock,⁹ post-cardiotomy shock with VA-ECMO support,¹² and those who underwent heart transplantation.¹³ However, limited data are available regarding the clinical value of the VIS in those with refractory cardiogenic shock requiring MCS. Na et al recently reported the predictive value of the VIS in patients with cardiogenic shock, of whom 26.5% received VA-ECMO support,⁹ and suggested the possible benefit of VA-ECMO in a selected group of patients with a VIS >130. However, the clinical application of a VIS of 130 as a cut-off point for the initiation of VA-ECMO in those with cardiogenic shock may be limited because: (1) the study did not directly compare the patients stratified by a VIS value of 130; and (2) a VIS value of 130 may be too high to follow in real clinical practice. Our study is different from the previous study in that the study population was restricted to those with cardiogenic shock requiring VA-ECMO support, and our results are consistent with the study by Samuels et al, who reported a similar value of the VIS as a criterion for ventricular assist device in patients with post-cardiotomy shock.¹⁴

The present study suggested that a high VIS of ≥32 at the time of VA-ECMO application for cardiogenic shock was associated with poor in-hospital outcomes after VA-ECMO. Successful weaning from VA-ECMO depends on multiple clinical factors including age, preexisting comorbidities, the status and reversibility of underlying cardiac disease-causing shock, and lactate level.^3,4,15,16 Delayed initiation of VA-ECMO that leads to prolonged exposure to higher doses of various vasoactive and inotropic agents may precipitate extensive tissue hypoxia and subsequent multiple organ dysfunction, which may further worsen the clinical outcomes.^17,18 Therefore, timely intervention for hemodynamic support is crucial for treating refractory cardiogenic shock before the development of multiorgan dysfunction. However, there is no clear consensus on the optimal timing for the initiation of VA-ECMO in terms of achieving the balance between improving clinical outcome with avoidance of multiorgan dysfunction by swift initiation and avoiding the potential complications of VA-ECMO by delaying the initiation. The current Extracorporeal Life Support Organization (ELSO) guidelines for VA-ECMO do not have a detailed recommendation on this issue, and limited data are available on the optimal time-point for the initiation of VA-ECMO for patients with cardiogenic shock. Nevertheless, dysregulated organ function represented by a high Sequential Organ Failure Assessment score is related to high risks of mortality in patients with cardiogenic shock requiring VA-ECMO support.¹⁹ Furthermore, several studies reported the possible benefit for early initiation of VA-ECMO.^5,6,17,18 As an example, Basir et al reported that the amount of inotropic support was inversely associated with in-hospital survival and suggested that early hemodynamic support with MCS might be a key factor in improving clinical outcomes.⁶ In this respect, our study results are valuable in that the VIS, an easily calculable score, was revealed as a potentially useful clinical marker for determining the timely initiation of VA-ECMO and independently predictive of in-hospital death. In particular, a high VIS could predict the clinical outcome after VA-ECMO independent of the pre-VA-ECMO lactate level, which is a surrogate marker for tissue oxygenation and widely used for determining VA-ECMO initiation.

In the present study, the trend for the VIS after VA-ECMO insertion was different between the 2 groups. VA-ECMO can support blood pressure and tissue perfusion. Therefore, high doses of vasoactive or inotropic agents were reduced after VA-ECMO insertion in the high VIS group. However, the amount of drug use increased slightly after VA-ECMO insertion in the low VIS group. Even though the VIS increased in the low VIS group, the average values of the VIS were very low compared with those in the high VIS group. Furthermore, the absolute differences of the VIS between at ECMO and 6 h after ECMO were also small in the low VIS group. We assumed that VA-ECMO was initiated in patients in the low VIS group before overt shock with multiorgan damage was evident, which might lead to the absence of prominent changes in the VIS. Also, the use of analgesic and/or sedative agents to relieve pain associated with VA-ECMO and relative volume depletion after the initiation of VA-ECMO might have led to the slight increase in the VIS in the low VIS group. Although the role of VA-ECMO was evident because it prominently decreased the VIS in the high VIS group, the clinical outcomes were worse in the high VIS group. Therefore, despite significant hemodynamic support with VA-ECMO with a marked decrease in the VIS, delayed initiation of VA-ECMO may be related to poor outcomes.

In addition, the duration from the initial requirement for vasoactive or inotropic agents to initiation of VA-ECMO was not significantly different between the 2 groups, as shown in Table 1. However, the duration from drug use to VA-ECMO may not be the only marker to represent the optimal time-point for initiating VA-ECMO because the clinical course and etiology of cardiogenic shock are heterogeneous. In our study, the proportion of eCPR was higher in the low VIS group. Physicians are not convinced that lower blood pressure during eCPR can be promptly recovered by increasing vasoactive or inotropic agents, and thus more likely to decide to insert VA-ECMO. However, in the high VIS group, physicians may have a relatively longer time to monitor the blood pressure with the increase of vasoactive or inotropic agent use, even though the requirements of the drugs are rapidly increasing. For this reason, even though the duration from the initial need for the drugs to the initiation of VA-ECMO was not significantly different, the blood pressure levels were significantly different between the 2 groups. Thus, we assume that various factors including clinical course, reversibility of the etiology for cardiogenic shock, preference of the physician for performing an invasive, risk-harboring procedure, and the time and location (e.g., general ward, ICU, emergency room) of shock development may affect the levels of the VIS.

Another notable finding of our study is that the potential benefit of initiation of VA-ECMO before reaching a higher VIS was prominent in patients with cardiogenic shocks of non-ischemic etiology. Patients with a cardiogenic shock resulting from ischemic etiology may be different from those with non-ischemic etiology in terms of clinical presentation, the nature and reversibility of underlying cardiomyopathy, and the clinical situation at the time of VA-ECMO initiation. As an example, eCPR was significantly higher in patients with ischemic etiology than those with non-ischemic etiology (61.0% vs. 39.8%, P=0.007), which suggests a higher proportion of patients with ischemic etiology were started on VA-ECMO in an acute clinical setting without clinical awaiting with the infusion of vasoactive or inotropic agents. The inclusion of a significantly higher proportion of eCPR in the ischemic etiology group might have resulted in blunting the differential outcomes between the low and high VIS groups in this population. Therefore, caution is needed when interpreting our study results that the possible benefit of the initiation of VA-ECMO in patients with a low VIS may be suitable for those titrating vasoactive or inotropic agents for hemodynamic support. In addition, although the in-hospital survival was higher in the low VIS group, the overall rates of complication related to VA-ECMO were not significantly different between the low and high VIS groups. Therefore, the initiation of VA-ECMO should be determined by comprehensively considering the risk of VA-ECMO-related complications in the setting of each center.

There are several limitations to our study. First, the study population with cardiogenic shock was heterogeneous in terms of the clinical setting and underlying cause of cardiomyopathy. Second, the study has inherent limitations stemming from its retrospective design. Although the baseline characteristics were generally comparable between the 2 groups, unmeasured confounders might have influenced the observed results. Therefore, a prospective study is needed to determine the optimal time for the initiation of VA-ECMO and establish the role of a VIS in the treatment of patients with cardiogenic shock. Third, this study could not determine the exact mechanism of the beneficial effect of VA-ECMO initiation in those with a low VIS. Fourth, vasoactive or inotropic agent use can invariably change over time and may simply reflect the severity of shock. However, Choi et al reported that the VIS could predict in-hospital outcomes among patients with acute cardiogenic shock due to myocardial infarction.²⁰ Therefore, despite its property that changes over time, a VIS may have a predictive role and be useful for deciding the initiation of VA-ECMO. Fifth, the study population only included those with cardiogenic shock who had already received VA-ECMO. Patients with cardiogenic shock may recover with medical therapy, and VA-ECMO support is not mandatory in some of those patients. Such a restricted study population may limit the interpretation of the potential benefit of VA-ECMO initiation in those with a low VIS. Finally, combined hemodynamic support with intra-aortic balloon counterpulsation was rarely used in our study population; moreover, percutaneous axial flow pump (Impella^®), which was reported to improve the clinical outcomes when used concomitantly with VA-ECMO,²¹ was not utilized due to unavailability in South Korea.

Conclusions

In the present study comparing the benefit of VA-ECMO initiation according to a VIS of 32, hemodynamic support with VA-ECMO at a lower VIS showed better in-hospital outcome and higher weaning rate in patients with refractory cardiogenic shock. A VIS may be a potential clinical marker for determining the timing of VA-ECMO initiation.

Disclosures

The authors declare that they have no conflicts of interest.

IRB Information

This work was approved by the IRB of Asan Medical Center, Seoul, Korea.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-21-0614

References

• 1.   Pagani FD, Lynch W, Swaniker F, Dyke DB, Bartlett R, Koelling T, et al. Extracorporeal life support to left ventricular assist device bridge to heart transplant: A strategy to optimize survival and resource utilization. Circulation 1999; 100: II206–II210.
• 2.   Burrell AJ, Pellegrino VA, Wolfe R, Wong WK, Cooper DJ, Kaye DM, et al. Long-term survival of adults with cardiogenic shock after venoarterial extracorporeal membrane oxygenation. J Crit Care 2015; 30: 949–956.
• 3.   Schmidt M, Burrell A, Roberts L, Bailey M, Sheldrake J, Rycus PT, et al. Predicting survival after ECMO for refractory cardiogenic shock: The survival after veno-arterial-ECMO (SAVE)-score. Eur Heart J 2015; 36: 2246–2256.
• 4.   Muller G, Flecher E, Lebreton G, Luyt CE, Trouillet JL, Bréchot 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.
• 5.   Choi KH, Yang JH, Hong D, Park TK, Lee JM, Song YB, et al. Optimal timing of venoarterial-extracorporeal membrane oxygenation in acute myocardial infarction patients suffering from refractory cardiogenic shock. Circ J 2020; 84: 1502–1510.
• 6.   Basir MB, Schreiber TL, Grines CL, Dixon SR, Moses JW, Maini BS, et al. Effect of early initiation of mechanical circulatory support on survival in cardiogenic shock. Am J Cardiol 2017; 119: 845–851.
• 7.   Takagi K, Kimmoun A, Sato N, Mebazaa A. Management of acute heart failure during an early phase. Int J Heart Fail 2020; 2: 91–110.
• 8.   Vahdatpour C, Collins D, Goldberg S. Cardiogenic shock. J Am Heart Assoc 2019; 8: e011991.
• 9.   Na SJ, Chung CR, Cho YH, Jeon K, Suh GY, Ahn JH, et al. Vasoactive inotropic score as a predictor of mortality in adult patients with cardiogenic shock: Medical therapy versus ECMO. Rev Esp Cardiol (Engl Ed) 2019; 72: 40–47.
• 10.   Gaies MG, Gurney JG, Yen AH, Napoli ML, Gajarski RJ, Ohye RG, et al. Vasoactive-inotropic score as a predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med 2010; 11: 234–238.
• 11.   Gaies MG, Jeffries HE, Niebler RA, Pasquali SK, Donohue JE, Yu S, et al. Vasoactive-inotropic score is associated with outcome after infant cardiac surgery: An analysis from the Pediatric Cardiac Critical Care Consortium and Virtual PICU System Registries. Pediatr Crit Care Med 2014; 15: 529–537.
• 12.   Chen F, Wang L, Shao J, Wang H, Hou X, Jia M. Survival following venoarterial extracorporeal membrane oxygenation in postcardiotomy cardiogenic shock adults. Perfusion 2020; 35: 747–755.
• 13.   Barge-Caballero E, Segovia-Cubero J, González-Vilchez F, Delgado-Jiménez J, Pérez-Villa F, Almenar-Bonet L, et al. Evaluation of the preoperative vasoactive-inotropic score as a predictor of postoperative outcomes in patients undergoing heart transplantation. Int J Cardiol 2015; 185: 192–194.
• 14.   Samuels LE, Kaufman MS, Thomas MP, Holmes EC, Brockman SK, Wechsler AS. Pharmacological criteria for ventricular assist device insertion following postcardiotomy shock: Experience with the Abiomed BVS system. J Card Surg 1999; 14: 288–293.
• 15.   Keebler ME, Haddad EV, Choi CW, McGrane S, Zalawadiya S, Schlendorf KH, et al. Venoarterial extracorporeal membrane oxygenation in cardiogenic shock. JACC Heart Fail 2018; 6: 503–516.
• 16.   Haile DT, Schears GJ. Optimal time for initiating extracorporeal membrane oxygenation. Semin Cardiothorac Vasc Anesth 2009; 13: 146–153.
• 17.   Chaturvedi RR, Macrae D, Brown KL, Schindler M, Smith EC, Davis KB, et al. Cardiac ECMO for biventricular hearts after paediatric open heart surgery. Heart 2004; 90: 545–551.
• 18.   Ge M, Pan T, Wang JX, Chen ZJ, Wang DJ. Outcomes of early versus delayed initiation of extracorporeal life support in cardiac surgery. J Cardiothorac Surg 2019; 14: 129.
• 19.   Dangers L, Bréchot N, Schmidt M, Lebreton G, Hékimian G, Nieszkowska A, et al. Extracorporeal membrane oxygenation for acute decompensated heart failure. Crit Care Med 2017; 45: 1359–1366.
• 20.   Choi KH, Yang JH, Park TK, Lee JM, Song YB, Hahn JY, et al. Differential prognostic implications of vasoactive inotropic score for patients with acute myocardial infarction complicated by cardiogenic shock according to use of mechanical circulatory support. Crit Care Med 2021; 49: 770–780.
• 21.   Pappalardo F, Schulte C, Pieri M, Schrage B, Contri R, Soeffker G, et al. Concomitant implantation of Impella(®) on top of veno-arterial extracorporeal membrane oxygenation may improve survival of patients with cardiogenic shock. Eur J Heart Fail 2017; 19: 404–412.