585517 review-article2015
ACC0010.1177/2048872615585517European Heart Journal: Acute Cardiovascular CareLazzeri et al.
EUROPEAN SOCIETY OF CARDIOLOGY ®
Review
Extracorporeal life support for outof-hospital cardiac arrest: Part of a treatment bundle
European Heart Journal: Acute Cardiovascular Care 1–10 © The European Society of Cardiology 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/2048872615585517 acc.sagepub.com
Chiara Lazzeri1, Serafina Valente1, Adriano Peris2 and Gian Franco Gensini1,3
Abstract In recent years, an increasing number of papers have been published on the use of extracorporeal cardiopulmonary resuscitation (ECPR) in adult patients, but, although promising results have been reported in patients with in-hospital refractory cardiac arrest supported by extracorporeal life support (ECLS), data on patients with out-of-hospital (OHCA) cardiac arrest are scarce and conflicting. The present study aims at summarizing the available evidence on the use of ECPR in adult patients with OHCA, clinically focusing on the factors most often associated with outcome in these patients. Even in the absence of randomized trials, there is growing evidence from ECLS centers documenting sound clinical benefits of ECPR in selected OHCA. According to the available evidence, three factors seem to contribute strongly to the favorable outcome of ECLS supported OHCA patients: (a) selection of patients (mainly definition of age range and a witnessed cardiac arrest); (b) the availability of an ECLS team, well skilled and experienced (to reduce time of implantation and incidence of complications); (c) a multifaceted approach to the OHCA patient (the so-called ECLS-bundle) to treat the reversible cause of CA (i.e. percutaneous coronary intervention), ensure neuroprotection (hypothermia), and maintain organ perfusion (till recovery). Taking into account the promising results of ECPR in selected OHCA patients, there is a clinical need for shared protocols to reduce differences related to the center experience and mostly to increase availability of ECLS as part of a multifaceted approach for these patients. Keywords Refractory out-of-hospital cardiac arrest, extracorporeal life support, adult, outcome Date received: 10 January 2015; accepted: 8 April 2015
Introduction The term ECMO (extracorporeal membrane oxygenation), which has been used interchangeably with ECLS (extracorporeal life support), indicates the use of a modified cardiopulmonary bypass circuit for temporary life support for patients with potentially reversible cardiac and/or respiratory failure.1–4 Technological advances have improved ECLS circuits and made them widely available, so clinicians are expanding the potential life-saving indications in acute cardiovascular care and especially in patients with refractory cardiac arrest (CA).1–7 In the 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care, the term “extracorporeal life support (ECLS)” was used synonymously with ECPR.5
In recent years, an increasing number of papers have been published on the use of ECPR in adult patients, but, although promising results have been reported in patients 1Intensive
Care Unit of Heart and Vessels Department, Azienda Ospedaliero-Universitaria Careggi, Italy 2Anesthesia and Intensive Unit of Emergency Department, Azienda Ospedaliero-Universitaria Careggi, Italy 3Department of Experimental and Clinical Medicine, University of Florence, AOU Careggi, Fondazione Don Carlo Gnocchi IRCCS, Italy Corresponding author: Chiara Lazzeri, Intensive Care Unit of Heart and Vessels Department, Azienda Ospedaliero-Universitaria Careggi, Florence, Italy. Email: [email protected]
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
2
European Heart Journal: Acute Cardiovascular Care
with in-hospital refractory cardiac arrest (IHCA) supported by ECLS,7–11 data on patients with out-of-hospital (OHCA) cardiac arrest are scarce and conflicting.1,12–20
Methods The present study aims at summarizing the available evidence on the use of ECPR in adult patients with OHCA, clinically focusing on the factors most often associated with outcome in these patients.
Results Investigations on the impact of ECPR in OHCA patients have provided contrasting results, which are related to differences in selection/inclusion criteria and to protocols for the management of the OHCA patients under ECLS support that change from to one ECLS center to another. Maekawa et al.21 used propensity score analysis for the first time in OHCA patients to balance the clinical variables and to reduce selection bias when comparing the neurologic outcome in ECPR and conventional cardiopulmonary resuscitation (CCPR). The study population included 162 adult OHCA patients with witnessed cardiac arrest of cardiac origin. The propensity score matching process selected 24 OHCA patients for each subgroup and documented a significantly better outcome (that is a favorable neurological outcome at 3 months after CA) in the ECPR group than in the CCPR group (ECPR 29.2% vs CCPR 8.3%, log-rank p = 0.018). This investigation represents, to date, the strongest evidence, documenting a benefit of ECPR over CCPR in OHCA patients. When examining the available evidence on the clinical impact of ECPR in OHCA patients three factors seem to account for the discrepancies among results, especially in terms of outcomes (survival and/or favorable neurologic outcome): (a) selection/inclusion criteria; (b) ECLS team expertise; and (c) treatment bundle during ECLS support. Selection criteria. Out-of-hospital cardiac arrest is quite a complex disease as different etiologies can cause it (from acute coronary syndrome to pulmonary embolism) and external factors (i.e. the presence/absence of a witness, the time of arrival of the emergency medical team) can influence outcome. Research in this field is therefore quite difficult and randomized studies are not feasible because they would be unethical.22 Table 1 depicts the selection criteria adopted in the investigations performed in adult OHCA supported by ECLS, together with the study design and the location of the ECLS center. These studies were retrieved by a PubMed search using the terms “ECLS/ECMO” and “out-of-hospital cardiac arrest” and “adult”. The search was restricted to English language articles. Case reports and investigations enrolling patients with postcardiotomy shock, refractory cardiogenic
shock, and cardiac arrest following hypothermia, and drug intoxication were not considered. As shown in Figure 1, our search retrieved 36 patients, among whom five reviews and 16 case reports were not considered. As shown in Table 1, most investigations were retrospective and most were published in the last 2 years (2013/2014). One-third of the studies was performed in Japan (5/15, 30%), while six investigations were in European ECLS centers (6/15, 40%). Selection criteria showed great variability, although a witnessed cardiac arrest was considered an inclusion criteria in most studies. With regard to age, the upper limit was 75 years in most protocols, although Fagnoul et al.17 and Stub et al.23 did not include patients older than 65 years and Wang et al.24 enrolled patients aged 80 years. Age range was not clearly defined in the study by Leick et al.,25 although the authors stated that the study was performed in “adult patients.” Maekawa et al.21 and Kim et al.26 enrolled patients aged < 16 years and ⩾ 18 years, respectively, with no apparent upper age limit. In the study by Kim et al.,26 age was a significant predictor for good neurologic outcome in the matched ECLS group. Ventricular fibrillation was a selection criteria in all prospective studies, except those by Fagnoul et al.17 and Wang et al.24 in whom the initial rhythm was not specified. The absence of ROSC after 10-20 minutes of CPR identified in most studies the condition of “refractory cardiac arrest” and the time to alert the ECLS team. ECMO team expertise. The term “ECMO team expertise” indicates the technical skills to implant ECMO in the shortest possible time and to reduce the incidence of complications related to implantation itself. Some papers25,27,28 documented that door-to-ECLS implantation time was significantly longer in non-survivors than in survivors. Fagnoul et al.17 showed that good outcome was obtained in 15-20% of patients if time from cardiac arrest to ECLS was less than 60 minutes. Kim et al.26 observed that the rate of favorable neurologic outcome tended to decrease when CPR duration increased, even in the matched ECLS group, so those authors suggested that OHCA patients who do not respond to conventional CPR after 21 minutes must be considered for ECLS and that earlier implantation may be better for a good neurologic outcome. In fact, matched ECLS showed a higher rate of favorable outcome with 21-80 minutes of CPR duration than matched CCPR (p=0.026).26 In other words, in potential candidates for ECLS (those with a suspected reversible cause of CA and without known exclusion criteria such as malignancies) who do not respond to conventional CPR after 21 minutes, ECLS support should be considered and the ECLS team should be alerted for prompt evaluation and eventual implantation (even while the patient is being transported to hospital). Prompt alert may allow better assessment of all factors concerning the event/cardiac arrest (witnessed/not witnessed, times of no and low flow) and of the clinical characteristics of the patients (i.e. comorbidities, the possibility of a reversible
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
Germany
Japan
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
55 ECLS patients, matched ECLS 52 patients
Retrospective
Kim et al., 201426
Seoul, Korea
Japan
85 CA patients, 59 Germany OHCA 18 OHCA Monza, Italy
Mixed population: IHCA and OHCA Kagawa et al Retrospective 77 CA patients, 39 Hiroshima OHCA Japan 201027
260 patients
Prospective observational study
Haneya et al., Retrospective 201216 Retrospective Avalli et al., 201212
2010-2011
1983-2008
2008-2010
2006-2011
2007-2012
2007-2009
2006-2013
2008-2011
53 ECPR, matched Sapporo, Japan 2010-2011 group 24 patients 26 patients Pennsylvania, 2007-2014 USA
Sakamoto et al., 201428
Leick et al., 201325 Maekawa Retrospective et al., 201321 Johnson et al., Prospective 201431 registry
Morimura et al., 201120
Paris, France
Period of enrolment
Not reported
12–75 years
18–74 years
⩾ 18 years
Not reported
VF on ECG during CPR
Collapse to EMS arrive < 15 minutes; presumed cardiac origin or pulmonary embolism; ROSC not achievable within 20 minutes of conventional CPR CPR for more than 10 minutes without ROSC CPR for more than 30 minutes
Ventricular fibrillation or ventricular tachycardia as initial rhythm or obvious cardiac or other reversible etiology with non-shockable rhythm VF/VT on the initial ECG Within 45 minutes from the reception of the emergency call or the onset of CA to the hospital arrival Not reported No-flow time (time interval from presumed arrest to CPR started by the EMS provider) expected to be short
18–70 years
> 20 – < 75 years
Cardiac origin
Absence of ROSC after 10 minutes of CPR CPR for more than 20 minutes Collapse to EMS arrive < 15 minutes
Not reported
Absence of ROSC after 30 minutes of CPR
Time from CA
> 16 years
Not reported Not reported
51 (4–88)
Not reported
Rhythm on initial ECG
Selection criteria
> 30 kg < 70 years
Prospective 51 patients observational study Review of 139 patients all previous studies in Japan Retrospective 28 patients
Location of ECLS center
Le Guen et al., 201113
Number of patients Age
Study design
Investigation
Table 1. Selection criteria.
(Continued)
No other inclusion criteria Not reported
Witnessed, of suspected cardiac origin
Not witnessed in 28%
Witnessed cardiac arrest
Witnessed
Witnessed
Witnessed in 90.7%
Witnessed CA
Witnessed/not witnessed
Lazzeri et al. 3
24 patients, 14 OHCA
230 patients, 31 OHCA 26 patients, 11 with OHCA
Fagnoul et al., Prospective 201317
Prospective
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
2012–2013
2008–2011
2004–2011
Period of enrolment
Melbourne, Australia
Over 32 months
Tapei, Taiwan 2007–2012
Bruxelles, Belgium
Pavia, Italy
Hiroshima, Japan
Location of ECLS center
18–65 years
16–80 years
100 minutes (relative exclusion criteria)
Time from CA
CA: cardiac arrest; CPR: cardiopulmonary resuscitation; EMS: emergency medical system; ROSC: return of spontaneous circulation; VF: ventricular fibrillation; VT: ventricular tachycardia.
Prospective pilot study
Seven OHCA
Prospective
Mojoli et al., 201319
Wang et al., 20134 Stub et al., 201423
86 ACS CA patients, 42 OHCA
Number of patients
Retrospective
Study design
Kagawa et al 201218
Investigation
Table 1. (Continued)
4 European Heart Journal: Acute Cardiovascular Care
5
Lazzeri et al.
complications was one of the significant factors for expectation of good neurologic outcome (together with early hemodynamic stabilization and therapeutic hypothermia). This finding underscores the role of the ECLS team, with experience in preventing and managing ECLS-related complications (Table 2).
Figure 1. Flow chart.
cause such as acute coronary syndrome). Moreover, the ECLS time of implantation could be the shortest. In other terms a prompt alert may give the patients the highest possibilities of a good neurologic outcome and the lowest incidence of complications. Wang et al.24 compared the duration of ischemia (that is time from collapse to ECLS start) in IHCA in their center from a previous study29 and observed a shorter time (44 ±24.7 minutes documented in the series of 199 patients vs 55±27.0 minutes reported in their previous paper (n=135)). This finding indicated that improved organization of the ECLS team and a preequipped ECLS chart can decrease the time for deployment. Moreover, in OHCA patients the percentage of patients with favorable outcome decreased as the time of ischemia increased.24 No data exist so far on how to implement ECLS in OHCA in an institution. Taking into account the complexity of these patients and the technical skills required for ECLS implantation, it is reasonable to suppose that experienced ECLS centers should host training programs for personnel from other centers. The door-to-ECLS implantation time was the only and independent predictor of 30-day mortality in multivariate Cox regression analysis (p=0.04) in the series by Leick et al.25, and Kaplan-Meier survival analysis showed a benefit favoring patients with a door-to-ECLS implantation time less than 30 minutes (log rank 6.29; p=0.01). The available evidence strongly supports the notion that the door-to-ECLS implantation time is able to affect outcome, but this time depends not only on rapid ECLS implantation (strictly related to ECLS team expertise) but also on prompt alert of the ECLS team (which is related to implementation of a well-defined protocol for refractory OHCA). The incidence of ECLS-related complications was reported only in some studies, and variability of reported data can be related to the intercenter variability of practice habits and different criteria for identification of specified diagnosis.30 However, bleeding events were the most common, ranging from 23% to 70%. A percutaneous approach is most commonly used for cannulation as it is safe, quick, and associated with lower incidence of bleeding events.7,18,21,27,28 In the paper by Kim et al.26 in the subgroup of matched ECPR (n=52 patients), a low incidence of ECLS-related
ECLS-bundle. Recently, promising results have been reported in terms of survival and favorable neurologic outcome (that is CPC 1-2) in OHCA patients treated with ECLS. In fact, while Le Guen et al.13 observed a survival rate at discharge of 12% (and of 4% at 1 month), more encouraging survival rates are being reported in more recent studies, ranging from 15% 28 to 54%.23 Simultaneously, an improvement in good neurologic outcome was observed. Indeed, all survivors at discharge showed a CPC 1-2 in the studies by Wang et al.24 and Stub et al.23 (25.5% and 54%, respectively). Kim et al.26 provided the strongest evidence that ECLS support is associated with a favorable neurologic outcome in selected OHCA patients. By means of propensity score analysis, Kim et al.26 documented that the matched ECPR group, compared with the matched CCPR group showed an improved neurological outcome at three months post arrest, despite the survival to discharge rate not showing any difference between groups. Evidence from recent investigations suggests strongly that ECLS support, especially when included in a welldefined treatment bundle, is associated with favorable outcome in selected patients with refractory cardiac arrest. In other terms, OHCA patients supported by ECLS are submitted to a “multifaceted treatment” (the so-called ECLSbundle), which includes percutaneous coronary intervention, when needed, and hypothermia (Table 3). While Johnson et al.31 did not mention the treatments/therapeutic approaches implemented in their patients during ECLS support, all patients enrolled by Fagnoul et al.17 underwent hypothermia during ECLS support (intra-arrest cooling was performed in 17 patients) and normoxemia (PaO2 was maintained between 60 and 150 mmHg) to avoid increased production of reactive oxygen species. According to their protocol, a heat exchanger on the ECMO circuit was used to maintain body temperature at 33°C for 24 hours, while in our center, temperature is maintained at 34°C for 24 hours.7 In the series by Wang et al.,2 cardiac reperfusion was performed in almost half of the population. The clinical relevance of the ECLS treatment bundle seems to be confirmed also by the results of the refractory cardiac arrest treated with mechanical CPR, hypothermia, ECMO, and early reperfusion21 trial23 in whom survival to hospital discharge with full neurological recovery occurred in 14/26 (54%) patients, and, in particular, in the 60% of IHCA patients and in the 45% of OHCA patients. The ECLS-bundle comes from the concept that ECLS is a support therapy which gives us time to treat the reversible cause of cardiac arrest (after identification and when possible) and to maintain organ perfusion (till recovery).
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
6
European Heart Journal: Acute Cardiovascular Care
Table 2. ECLS team expertise. Number of patients
Time cardiac arrest to ECLS onset (minutes)
ECLS-related complications
ECLS implantation failure
Le Guen et al., 201113
51 patients
120 (102–149)
9/51, 18%
Morimura et al., 201120
139 patients
53 (33–70)
Leick et al., 201325
28 patients
Maekawa et al., 201321
53 ECPR matched group, 24 patients 26 patients
Door-to-ECLS implantation time was significantly longer in non-survivors [42.5 minutes (IQR 28.0-56.5) vs. 25.0 minutes (IQR 21.030.0); p < 0.01] Not reported
Transfusion 39%, severe hemorrhage 15% Bleeding 3%, ischemia of lower extremity 1.5%, malposition: 1.5% Bleeding 32%, lower limb ischemia 9%
Johnson et al., 201431 Sakamoto et al., 201428 Kim et al., 201426
260 patients 55 ECLS patients, matched ECLS 52 patients Mixed population: IHCA and OHCA Kagawa et al., 201027 77 CA patients, 39 OHCA Haneya et al., 201216 85 CA patients, 59 OHCA Avalli et al., 201212 18 OHCA Kagawa et al., (2012)18 86 ACS CA patients, 42 OHCA Mojoli et al., 201319 Seven OHCA Fagnoul et al., 201317 24 patients, 14 OHCA Wang et al., 20144 230 patients, 31 OHCA 26 patients, 11 with Stub et al., 201423 OHCA
77 ± 51 (range 12–180 minutes) Not reported Not reported
Ischemic complication 31%, bleeding 23.1%, stroke 15.4% Not reported 36.4%
Two unsuccessful cannulations
OHCA 59 (45–65)
Leg ischemia 21%, bleeding 59% Overall (85 patients), 32%
One patient out of 39 Overall, 15 cases cannulation failure
Not available 77 (69–101) 49 (30–68)
93 (74–107) All 58 (45–70) OHCA 67.5 ±30.6 56 (40–85)
Overall bleeding 25% Peripheral vascular injuries 39%, bleeding 70%
ACS: acute coronary syndrome; CA: cardiac arrest; ECLS: extracorporeal life support; IHCA: in-hospital cardiac arrest; IQR: interquartile range; OHCA: out-of-hospital cardiac arrest.
Prognostic factors Most investigations12,13,16–20,23,28,31 did not specifically investigate which factors are independently related to outcome (survival and/or favorable neurologic outcome) in OHCA patients treated with ECLS. On the contrary, few studies performed a multivariate analysis to identify the factors able to independently affect prognosis (Table 4). In the paper by Leick et al.,25 the door-to-ECLS implantation time was the only significant and independent predictor of 30-day mortality while lactate values at baseline, age and reason for cardiac arrest were not significant. Kaplan– Meier survival analysis showed a benefit favoring patients with a door-to-ECLS implantation time
ACC0010.1177/2048872615585517European Heart Journal: Acute Cardiovascular CareLazzeri et al.
EUROPEAN SOCIETY OF CARDIOLOGY ®
Review
Extracorporeal life support for outof-hospital cardiac arrest: Part of a treatment bundle
European Heart Journal: Acute Cardiovascular Care 1–10 © The European Society of Cardiology 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/2048872615585517 acc.sagepub.com
Chiara Lazzeri1, Serafina Valente1, Adriano Peris2 and Gian Franco Gensini1,3
Abstract In recent years, an increasing number of papers have been published on the use of extracorporeal cardiopulmonary resuscitation (ECPR) in adult patients, but, although promising results have been reported in patients with in-hospital refractory cardiac arrest supported by extracorporeal life support (ECLS), data on patients with out-of-hospital (OHCA) cardiac arrest are scarce and conflicting. The present study aims at summarizing the available evidence on the use of ECPR in adult patients with OHCA, clinically focusing on the factors most often associated with outcome in these patients. Even in the absence of randomized trials, there is growing evidence from ECLS centers documenting sound clinical benefits of ECPR in selected OHCA. According to the available evidence, three factors seem to contribute strongly to the favorable outcome of ECLS supported OHCA patients: (a) selection of patients (mainly definition of age range and a witnessed cardiac arrest); (b) the availability of an ECLS team, well skilled and experienced (to reduce time of implantation and incidence of complications); (c) a multifaceted approach to the OHCA patient (the so-called ECLS-bundle) to treat the reversible cause of CA (i.e. percutaneous coronary intervention), ensure neuroprotection (hypothermia), and maintain organ perfusion (till recovery). Taking into account the promising results of ECPR in selected OHCA patients, there is a clinical need for shared protocols to reduce differences related to the center experience and mostly to increase availability of ECLS as part of a multifaceted approach for these patients. Keywords Refractory out-of-hospital cardiac arrest, extracorporeal life support, adult, outcome Date received: 10 January 2015; accepted: 8 April 2015
Introduction The term ECMO (extracorporeal membrane oxygenation), which has been used interchangeably with ECLS (extracorporeal life support), indicates the use of a modified cardiopulmonary bypass circuit for temporary life support for patients with potentially reversible cardiac and/or respiratory failure.1–4 Technological advances have improved ECLS circuits and made them widely available, so clinicians are expanding the potential life-saving indications in acute cardiovascular care and especially in patients with refractory cardiac arrest (CA).1–7 In the 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care, the term “extracorporeal life support (ECLS)” was used synonymously with ECPR.5
In recent years, an increasing number of papers have been published on the use of ECPR in adult patients, but, although promising results have been reported in patients 1Intensive
Care Unit of Heart and Vessels Department, Azienda Ospedaliero-Universitaria Careggi, Italy 2Anesthesia and Intensive Unit of Emergency Department, Azienda Ospedaliero-Universitaria Careggi, Italy 3Department of Experimental and Clinical Medicine, University of Florence, AOU Careggi, Fondazione Don Carlo Gnocchi IRCCS, Italy Corresponding author: Chiara Lazzeri, Intensive Care Unit of Heart and Vessels Department, Azienda Ospedaliero-Universitaria Careggi, Florence, Italy. Email: [email protected]
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
2
European Heart Journal: Acute Cardiovascular Care
with in-hospital refractory cardiac arrest (IHCA) supported by ECLS,7–11 data on patients with out-of-hospital (OHCA) cardiac arrest are scarce and conflicting.1,12–20
Methods The present study aims at summarizing the available evidence on the use of ECPR in adult patients with OHCA, clinically focusing on the factors most often associated with outcome in these patients.
Results Investigations on the impact of ECPR in OHCA patients have provided contrasting results, which are related to differences in selection/inclusion criteria and to protocols for the management of the OHCA patients under ECLS support that change from to one ECLS center to another. Maekawa et al.21 used propensity score analysis for the first time in OHCA patients to balance the clinical variables and to reduce selection bias when comparing the neurologic outcome in ECPR and conventional cardiopulmonary resuscitation (CCPR). The study population included 162 adult OHCA patients with witnessed cardiac arrest of cardiac origin. The propensity score matching process selected 24 OHCA patients for each subgroup and documented a significantly better outcome (that is a favorable neurological outcome at 3 months after CA) in the ECPR group than in the CCPR group (ECPR 29.2% vs CCPR 8.3%, log-rank p = 0.018). This investigation represents, to date, the strongest evidence, documenting a benefit of ECPR over CCPR in OHCA patients. When examining the available evidence on the clinical impact of ECPR in OHCA patients three factors seem to account for the discrepancies among results, especially in terms of outcomes (survival and/or favorable neurologic outcome): (a) selection/inclusion criteria; (b) ECLS team expertise; and (c) treatment bundle during ECLS support. Selection criteria. Out-of-hospital cardiac arrest is quite a complex disease as different etiologies can cause it (from acute coronary syndrome to pulmonary embolism) and external factors (i.e. the presence/absence of a witness, the time of arrival of the emergency medical team) can influence outcome. Research in this field is therefore quite difficult and randomized studies are not feasible because they would be unethical.22 Table 1 depicts the selection criteria adopted in the investigations performed in adult OHCA supported by ECLS, together with the study design and the location of the ECLS center. These studies were retrieved by a PubMed search using the terms “ECLS/ECMO” and “out-of-hospital cardiac arrest” and “adult”. The search was restricted to English language articles. Case reports and investigations enrolling patients with postcardiotomy shock, refractory cardiogenic
shock, and cardiac arrest following hypothermia, and drug intoxication were not considered. As shown in Figure 1, our search retrieved 36 patients, among whom five reviews and 16 case reports were not considered. As shown in Table 1, most investigations were retrospective and most were published in the last 2 years (2013/2014). One-third of the studies was performed in Japan (5/15, 30%), while six investigations were in European ECLS centers (6/15, 40%). Selection criteria showed great variability, although a witnessed cardiac arrest was considered an inclusion criteria in most studies. With regard to age, the upper limit was 75 years in most protocols, although Fagnoul et al.17 and Stub et al.23 did not include patients older than 65 years and Wang et al.24 enrolled patients aged 80 years. Age range was not clearly defined in the study by Leick et al.,25 although the authors stated that the study was performed in “adult patients.” Maekawa et al.21 and Kim et al.26 enrolled patients aged < 16 years and ⩾ 18 years, respectively, with no apparent upper age limit. In the study by Kim et al.,26 age was a significant predictor for good neurologic outcome in the matched ECLS group. Ventricular fibrillation was a selection criteria in all prospective studies, except those by Fagnoul et al.17 and Wang et al.24 in whom the initial rhythm was not specified. The absence of ROSC after 10-20 minutes of CPR identified in most studies the condition of “refractory cardiac arrest” and the time to alert the ECLS team. ECMO team expertise. The term “ECMO team expertise” indicates the technical skills to implant ECMO in the shortest possible time and to reduce the incidence of complications related to implantation itself. Some papers25,27,28 documented that door-to-ECLS implantation time was significantly longer in non-survivors than in survivors. Fagnoul et al.17 showed that good outcome was obtained in 15-20% of patients if time from cardiac arrest to ECLS was less than 60 minutes. Kim et al.26 observed that the rate of favorable neurologic outcome tended to decrease when CPR duration increased, even in the matched ECLS group, so those authors suggested that OHCA patients who do not respond to conventional CPR after 21 minutes must be considered for ECLS and that earlier implantation may be better for a good neurologic outcome. In fact, matched ECLS showed a higher rate of favorable outcome with 21-80 minutes of CPR duration than matched CCPR (p=0.026).26 In other words, in potential candidates for ECLS (those with a suspected reversible cause of CA and without known exclusion criteria such as malignancies) who do not respond to conventional CPR after 21 minutes, ECLS support should be considered and the ECLS team should be alerted for prompt evaluation and eventual implantation (even while the patient is being transported to hospital). Prompt alert may allow better assessment of all factors concerning the event/cardiac arrest (witnessed/not witnessed, times of no and low flow) and of the clinical characteristics of the patients (i.e. comorbidities, the possibility of a reversible
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
Germany
Japan
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
55 ECLS patients, matched ECLS 52 patients
Retrospective
Kim et al., 201426
Seoul, Korea
Japan
85 CA patients, 59 Germany OHCA 18 OHCA Monza, Italy
Mixed population: IHCA and OHCA Kagawa et al Retrospective 77 CA patients, 39 Hiroshima OHCA Japan 201027
260 patients
Prospective observational study
Haneya et al., Retrospective 201216 Retrospective Avalli et al., 201212
2010-2011
1983-2008
2008-2010
2006-2011
2007-2012
2007-2009
2006-2013
2008-2011
53 ECPR, matched Sapporo, Japan 2010-2011 group 24 patients 26 patients Pennsylvania, 2007-2014 USA
Sakamoto et al., 201428
Leick et al., 201325 Maekawa Retrospective et al., 201321 Johnson et al., Prospective 201431 registry
Morimura et al., 201120
Paris, France
Period of enrolment
Not reported
12–75 years
18–74 years
⩾ 18 years
Not reported
VF on ECG during CPR
Collapse to EMS arrive < 15 minutes; presumed cardiac origin or pulmonary embolism; ROSC not achievable within 20 minutes of conventional CPR CPR for more than 10 minutes without ROSC CPR for more than 30 minutes
Ventricular fibrillation or ventricular tachycardia as initial rhythm or obvious cardiac or other reversible etiology with non-shockable rhythm VF/VT on the initial ECG Within 45 minutes from the reception of the emergency call or the onset of CA to the hospital arrival Not reported No-flow time (time interval from presumed arrest to CPR started by the EMS provider) expected to be short
18–70 years
> 20 – < 75 years
Cardiac origin
Absence of ROSC after 10 minutes of CPR CPR for more than 20 minutes Collapse to EMS arrive < 15 minutes
Not reported
Absence of ROSC after 30 minutes of CPR
Time from CA
> 16 years
Not reported Not reported
51 (4–88)
Not reported
Rhythm on initial ECG
Selection criteria
> 30 kg < 70 years
Prospective 51 patients observational study Review of 139 patients all previous studies in Japan Retrospective 28 patients
Location of ECLS center
Le Guen et al., 201113
Number of patients Age
Study design
Investigation
Table 1. Selection criteria.
(Continued)
No other inclusion criteria Not reported
Witnessed, of suspected cardiac origin
Not witnessed in 28%
Witnessed cardiac arrest
Witnessed
Witnessed
Witnessed in 90.7%
Witnessed CA
Witnessed/not witnessed
Lazzeri et al. 3
24 patients, 14 OHCA
230 patients, 31 OHCA 26 patients, 11 with OHCA
Fagnoul et al., Prospective 201317
Prospective
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
2012–2013
2008–2011
2004–2011
Period of enrolment
Melbourne, Australia
Over 32 months
Tapei, Taiwan 2007–2012
Bruxelles, Belgium
Pavia, Italy
Hiroshima, Japan
Location of ECLS center
18–65 years
16–80 years
100 minutes (relative exclusion criteria)
Time from CA
CA: cardiac arrest; CPR: cardiopulmonary resuscitation; EMS: emergency medical system; ROSC: return of spontaneous circulation; VF: ventricular fibrillation; VT: ventricular tachycardia.
Prospective pilot study
Seven OHCA
Prospective
Mojoli et al., 201319
Wang et al., 20134 Stub et al., 201423
86 ACS CA patients, 42 OHCA
Number of patients
Retrospective
Study design
Kagawa et al 201218
Investigation
Table 1. (Continued)
4 European Heart Journal: Acute Cardiovascular Care
5
Lazzeri et al.
complications was one of the significant factors for expectation of good neurologic outcome (together with early hemodynamic stabilization and therapeutic hypothermia). This finding underscores the role of the ECLS team, with experience in preventing and managing ECLS-related complications (Table 2).
Figure 1. Flow chart.
cause such as acute coronary syndrome). Moreover, the ECLS time of implantation could be the shortest. In other terms a prompt alert may give the patients the highest possibilities of a good neurologic outcome and the lowest incidence of complications. Wang et al.24 compared the duration of ischemia (that is time from collapse to ECLS start) in IHCA in their center from a previous study29 and observed a shorter time (44 ±24.7 minutes documented in the series of 199 patients vs 55±27.0 minutes reported in their previous paper (n=135)). This finding indicated that improved organization of the ECLS team and a preequipped ECLS chart can decrease the time for deployment. Moreover, in OHCA patients the percentage of patients with favorable outcome decreased as the time of ischemia increased.24 No data exist so far on how to implement ECLS in OHCA in an institution. Taking into account the complexity of these patients and the technical skills required for ECLS implantation, it is reasonable to suppose that experienced ECLS centers should host training programs for personnel from other centers. The door-to-ECLS implantation time was the only and independent predictor of 30-day mortality in multivariate Cox regression analysis (p=0.04) in the series by Leick et al.25, and Kaplan-Meier survival analysis showed a benefit favoring patients with a door-to-ECLS implantation time less than 30 minutes (log rank 6.29; p=0.01). The available evidence strongly supports the notion that the door-to-ECLS implantation time is able to affect outcome, but this time depends not only on rapid ECLS implantation (strictly related to ECLS team expertise) but also on prompt alert of the ECLS team (which is related to implementation of a well-defined protocol for refractory OHCA). The incidence of ECLS-related complications was reported only in some studies, and variability of reported data can be related to the intercenter variability of practice habits and different criteria for identification of specified diagnosis.30 However, bleeding events were the most common, ranging from 23% to 70%. A percutaneous approach is most commonly used for cannulation as it is safe, quick, and associated with lower incidence of bleeding events.7,18,21,27,28 In the paper by Kim et al.26 in the subgroup of matched ECPR (n=52 patients), a low incidence of ECLS-related
ECLS-bundle. Recently, promising results have been reported in terms of survival and favorable neurologic outcome (that is CPC 1-2) in OHCA patients treated with ECLS. In fact, while Le Guen et al.13 observed a survival rate at discharge of 12% (and of 4% at 1 month), more encouraging survival rates are being reported in more recent studies, ranging from 15% 28 to 54%.23 Simultaneously, an improvement in good neurologic outcome was observed. Indeed, all survivors at discharge showed a CPC 1-2 in the studies by Wang et al.24 and Stub et al.23 (25.5% and 54%, respectively). Kim et al.26 provided the strongest evidence that ECLS support is associated with a favorable neurologic outcome in selected OHCA patients. By means of propensity score analysis, Kim et al.26 documented that the matched ECPR group, compared with the matched CCPR group showed an improved neurological outcome at three months post arrest, despite the survival to discharge rate not showing any difference between groups. Evidence from recent investigations suggests strongly that ECLS support, especially when included in a welldefined treatment bundle, is associated with favorable outcome in selected patients with refractory cardiac arrest. In other terms, OHCA patients supported by ECLS are submitted to a “multifaceted treatment” (the so-called ECLSbundle), which includes percutaneous coronary intervention, when needed, and hypothermia (Table 3). While Johnson et al.31 did not mention the treatments/therapeutic approaches implemented in their patients during ECLS support, all patients enrolled by Fagnoul et al.17 underwent hypothermia during ECLS support (intra-arrest cooling was performed in 17 patients) and normoxemia (PaO2 was maintained between 60 and 150 mmHg) to avoid increased production of reactive oxygen species. According to their protocol, a heat exchanger on the ECMO circuit was used to maintain body temperature at 33°C for 24 hours, while in our center, temperature is maintained at 34°C for 24 hours.7 In the series by Wang et al.,2 cardiac reperfusion was performed in almost half of the population. The clinical relevance of the ECLS treatment bundle seems to be confirmed also by the results of the refractory cardiac arrest treated with mechanical CPR, hypothermia, ECMO, and early reperfusion21 trial23 in whom survival to hospital discharge with full neurological recovery occurred in 14/26 (54%) patients, and, in particular, in the 60% of IHCA patients and in the 45% of OHCA patients. The ECLS-bundle comes from the concept that ECLS is a support therapy which gives us time to treat the reversible cause of cardiac arrest (after identification and when possible) and to maintain organ perfusion (till recovery).
Downloaded from acc.sagepub.com at NORTH DAKOTA STATE UNIV LIB on January 3, 2016
6
European Heart Journal: Acute Cardiovascular Care
Table 2. ECLS team expertise. Number of patients
Time cardiac arrest to ECLS onset (minutes)
ECLS-related complications
ECLS implantation failure
Le Guen et al., 201113
51 patients
120 (102–149)
9/51, 18%
Morimura et al., 201120
139 patients
53 (33–70)
Leick et al., 201325
28 patients
Maekawa et al., 201321
53 ECPR matched group, 24 patients 26 patients
Door-to-ECLS implantation time was significantly longer in non-survivors [42.5 minutes (IQR 28.0-56.5) vs. 25.0 minutes (IQR 21.030.0); p < 0.01] Not reported
Transfusion 39%, severe hemorrhage 15% Bleeding 3%, ischemia of lower extremity 1.5%, malposition: 1.5% Bleeding 32%, lower limb ischemia 9%
Johnson et al., 201431 Sakamoto et al., 201428 Kim et al., 201426
260 patients 55 ECLS patients, matched ECLS 52 patients Mixed population: IHCA and OHCA Kagawa et al., 201027 77 CA patients, 39 OHCA Haneya et al., 201216 85 CA patients, 59 OHCA Avalli et al., 201212 18 OHCA Kagawa et al., (2012)18 86 ACS CA patients, 42 OHCA Mojoli et al., 201319 Seven OHCA Fagnoul et al., 201317 24 patients, 14 OHCA Wang et al., 20144 230 patients, 31 OHCA 26 patients, 11 with Stub et al., 201423 OHCA
77 ± 51 (range 12–180 minutes) Not reported Not reported
Ischemic complication 31%, bleeding 23.1%, stroke 15.4% Not reported 36.4%
Two unsuccessful cannulations
OHCA 59 (45–65)
Leg ischemia 21%, bleeding 59% Overall (85 patients), 32%
One patient out of 39 Overall, 15 cases cannulation failure
Not available 77 (69–101) 49 (30–68)
93 (74–107) All 58 (45–70) OHCA 67.5 ±30.6 56 (40–85)
Overall bleeding 25% Peripheral vascular injuries 39%, bleeding 70%
ACS: acute coronary syndrome; CA: cardiac arrest; ECLS: extracorporeal life support; IHCA: in-hospital cardiac arrest; IQR: interquartile range; OHCA: out-of-hospital cardiac arrest.
Prognostic factors Most investigations12,13,16–20,23,28,31 did not specifically investigate which factors are independently related to outcome (survival and/or favorable neurologic outcome) in OHCA patients treated with ECLS. On the contrary, few studies performed a multivariate analysis to identify the factors able to independently affect prognosis (Table 4). In the paper by Leick et al.,25 the door-to-ECLS implantation time was the only significant and independent predictor of 30-day mortality while lactate values at baseline, age and reason for cardiac arrest were not significant. Kaplan– Meier survival analysis showed a benefit favoring patients with a door-to-ECLS implantation time
