Cardiac arrest (CA) affects approximately 40,000 people in France. It is still a major cause of death in a young population. Management of CA is defined by international recommendations, detailed by learned societies in each country. It includes several links that are interconnected for its optimisation. Despite all these improvements, no progress, or little has been made in the survival of CA victims over the past few years in industrialised countries, and the survival rate in France is 3% to 5%.

Refractory cardiac arrest is defined as failure, after 30 minutes of specialised resuscitation. It used to be the standard to admit that there was no hope of spontaneous cardiac activity and satisfactory neurological recovery after this period, except in cases of CA with neuroprotection (intoxication, hypothermia).

External circulatory support such as "extracorporeal membrane oxygenation" (ECMO) makes it possible to replace the circulatory activity of the myocardium and the respiratory activity of the lungs.

In in-hospital cardiac arrest (CA) some teams use ECMO with an improvement in the survival rate of 20% in comparison to standard resuscitation. This use demonstrates the possibility of neurological recovery independent of the recovery of spontaneous cardiac activity which can be differed.

These results encouraged the use of ECMOs in cases of out-of-hospital refractory cardiac arrests. Patients who are victims of CA are resuscitated for 30 minutes on the spot where the CA occurs. They are then transferred to a specialised centre. The significant improvement in survival noted in in-hospital CAs was not observed in the French series of studies concerning out-of-hospital CAs. This survival is currently estimated at 4%. This difference can be partly explained by the difference in time between the beginning of cardiac massage and the implementation of circulatory support by ECMO ("low flow" period). This time period is directly correlated to survival.

To demonstrate the superiority of this strategy in terms of survival, investigators would like to conduct a randomised comparative study of two strategies: 1) installation of an ECMO between the 20th minute to the 30 minute of CA, directly at the site of the CA, by emergency physicians and/or specifically trained resuscitators 2) On-site resuscitation optimised with secondary transfer to the hospital for the implementation of support. The purpose is to increase by 5% to 20% the survival of victims of out-of-hospital refractory cardiac arrests with a good neurological prognosis.

Main objective:

The hypothesis is that pre-hospital ECMO will result in survival for 20% of the patients, considering that the percentage of survival with in-hospital ECMO is less than 5%.

Main judgement criterion:

Survival with good neurological outcome (CPC 1 or 2) on discharge from intensive care or at 6 months

Secondary judgement criteria:

Success rate of the implementation of ECMO ECMO implementation time Immediate complications: haemorrhage, infection Number of organ harvesting The quality of survivors' neurological status according to the CPC neurological classification at D 28, 2 months and 1 year Predictive indicators of the prognosis during cardiac arrest via cerebral and biological monitoring Methodology, type of study: This is a prospective randomised study of current care Sample size (SS, power, risk): A total number of 105 patients in each group will make it possible to demonstrate at the alpha risk of 5% and a power of 1-β=90%, a significant difference in favour of early pre-hospital ECMO compared to the current practice with in-hospital ECMO.

1 STATUS OF THE QUESTION 2.1 Circumstances surrounding the problem: 2.1.1 The limits of conventional resuscitation in cardiac arrests. Sudden adult death, unexpected cardiac arrest (CA) most often related to cardiovascular causes, affects more than 40,000 people in France every year.

The prognosis of out-of-hospital sudden death is particularly sombre. Only 5 to 20% of the patients survive without neurological sequel. This prognosis can be partially improved by an efficient organization of management that implements the concept of the "survival chain". An early alert by witnesses of the CA, carrying out the first interventions, cardiopulmonary resuscitation (CPR), defibrillation by the public or persons qualified in first-aid, followed by specialized resuscitation by a medical team (such as the ambulance service) increase the amount of survivors. The reduction in the time that passes before the first interventions (called the no-flow period) and the duration of resuscitation before return of spontaneous circulation (ROSC) called low-flow, are considered to be the primary predictive factors of survival in CA.

Just recently, better knowledge of post-cardiac arrest syndrome also contributed to an improvement in the prognosis and the quality of survival. The introduction of early angioplasty and the generalisation of therapeutic hypothermia now complete management for which coding is steadily improving through international and national recommendations that are updated every 5 years. However, in order to show what it is capable of, this conventional management of CA requires spontaneous cardiac activity to be re-established as quickly as possible. Therefore, a return of spontaneous circulation (ROSC) must be obtained in the field in out-of-hospital sudden deaths.

Classically, in the absence of ROSC after 30 minutes of correctly administered resuscitation, the CA is considered to be refractory and the treatment options are limited. Just recently, it was shown that the chances of survival are non-existent after 16 minutes of resuscitation. In most cases, resuscitation is discontinued and the patient is declared dead on the spot. The regulatory inclusion of a physician on ambulance service teams relieves the difficulty of this decision. It is often simple to make and is medically indisputable when it is obvious that the prognostic factors are very unfavorable (notably, an extended no-flow period) and/or the conditions (advanced age, severe chronic disease, etc...) are not compatible with prolonged resuscitation. This decision is much more difficult when the prognostic factors are favorable and prolonged resuscitation efficiently provides spontaneous circulation (signs of the patient awaking during CPR). Under these circumstances, in France and several other European countries, the decision can be made to continue extracorporeal membrane oxygenation (ECMO) and transport the refractory CA victim. It was made possible by the development of mechanical external cardiac massage devices such as Autopulse ® and Lucas® which enable prolonged cardiac compression during transport by the emergency service. However, this continuation of resuscitation can only be considered if it enables another subsequent treatment for the patient. Two options are possible. The patient can be declared dead and become a potential organ donor in the framework of an organ harvesting procedure in a patient after "cardiac death". This harvesting, which is highly organized according to regulations, can only be done in certain hospitals authorized by the French Biomedicine Agency.

Or, resuscitation can be prolonged by the use of extracorporeal circulatory support.

2.1.2. The progress of extracorporeal circulatory support and cardiac arrest

Circulatory support is a technique that has been in common use for many years now perioperatively in cardiac surgery. One of its simplest forms, extracorporeal membrane oxygenation (ECMO) is being used more and more often outside of this field, notably in paediatrics and in the care of Acute Respiratory Distress Syndrome (ARDS) or refractory shock in adults. This technique has notably been widely introduced in general intensive care in the treatment of malignant influenzas that affect young subjects (H1N1 virus). In parallel with this extension of the indications for ECMO, the technical development of equipment was a major factor. ECMO devices, which are particularly easy to use, miniaturized and energy autonomous, are available. They make it possible to use ECMO during inter-hospital transport by ambulance or helicopter. In France, several teaching hospitals have therefore developed mobile teams called mobile circulatory support unit (UMAC) that enable the implementation of ECMO in intensive care units where there was none, and the transport of patients on circulatory and respiratory support to a reference center.

2.1.3. The implementation of extracorporeal membrane oxygenation (ECMO) in Cardiac Arrest (CA)

It quickly became evident that the possibility of having artificial circulatory activity that enables efficient perfusion by oxygenated blood was important for CA victims whose heart had stopped beating. The first research, conducted primarily during refractory CAs that occurred in the hospital setting, demonstrated the unexpected possibility for survival in patients who, without this option would be dead, and for whom resuscitation would have been stopped. In 2003 in Taiwan, Chen et al. noted a survival rate of almost 30% in a series of CAs that occurred in the hospital setting. In Caen, France, the same phenomenon was noted: the survival of 8 out of 40 patients who benefited from ECMO following refractory CA. This technique proved to be highly adapted when the cause of the CA was potentially reversible. Mégarbane et al. noted the survival of 3 out of 12 victims of CA following acute intoxication with cardiotoxic drugs.

In international recommendations, circulatory support is still only recommended in paediatrics. However, these indisputable successes in adults led to an attempt to rationalise the use of therapeutic ECMO in France. The indications considered as possible include the existence of hypothermia, intoxication, signs of life during Cardio-Pulmonary Ressucitation (CPR), and CPR (low-flow) of less than 100 minutes.

The development of ECMO programmes for the treatment of refractory CA demonstrated a difference in prognosis between in-hospital and out-of-hospital CAs. In-hospital CAs quickly benefit from the implementation of ECMO. Out-of-hospital CA victims have late access to this possibility of resuscitation. In fact they require resuscitation of at least 30 minutes in the field to be considered as refractory, followed by transport under mechanical massage until arrival at a center with ECMO. Le Guen et al. noted that in a series of patients who were victims of sudden death in Paris in out-of-hospital settings, only 2 out of 51 patients survived in good neurological condition. Most of these patients had extended low-flow periods before the implementation of ECMO. A negative correlation between the duration of resuscitation before ECMO and survival explains this poor prognostic result. In addition, resuscitation prolonged by mechanical massage is burdened by its own morbidity as Agostinucci et al. emphasized. This negative influence before access to ECMO is also noted by Chen et al. in the hospital setting. The prognosis rapidly decreases when resuscitation is prolonged: more than 40% survival if resuscitation lasts less than 30 minutes; 17% when it surpasses 60 minutes. This difference in survival between in-hospital and out-of-hospital CA is also noted in another series of French studies (Gay, AFAR abstract). The prognosis for out-of-hospital CA is even worse when it is accompanied by prolonged CPR. Morbidity is also higher among these patients. Cadarelli et al. included all the research and case histories published up until 2008 in a meta-analysis and demonstrated the harmful effect of prolonged CPR. In this analysis, the speed at which ECMO is implemented appears to be a prognostic factor similar to patients' age and the total duration of circulatory support. Therefore, ECMO that is started after more than 30 minutes of CPR results in a decrease in survival. Kilbaught et al. emphasize that it is actually the time factor that makes the difference between in-hospital and out-of-hospital CAs. In their pre-hospital emergency system, very rapid transport of patients during CPR to start ECMO upon arrival in the emergency service is possible. With this strategy, they demonstrate that the difference in prognosis between in-hospital and out-of-hospital CAs is eliminated when the time for implementation of ECMO is comparable. As a result, ECMO is used earlier and earlier in hospitals in Japan with results currently being published that appear to be very positive for survival.

2.1.4 The concepts of pre-hospital extracorporeal membrane oxygenation (ECMO)

The analysis of the international literature shows that ECMO might be a management method that improves the survival of CA victims.

However, in the context of out-of-hospital sudden death in a medicalised emergency system and in the framework of French regulations, there are two limiting factors:

• the obligation for resuscitation for 30 minutes before categorically announcing that the CA is refractory and whether or not to choose another treatment option.
• the possibility to have access to ECMO within the closest time period to the 30 minutes of Cardio-Pulmonary Ressucitation (CPR), which appears to be an important threshold in determining the prognosis.

Pre-hospital ECMO, which is the basis of the research concept being proposed, includes arterio-venous cannulation and the implementation of the extracorporeal system (pump, oxygenator) in a non-healthcare setting. It is therefore different from the in-hospital transport of patients on ECMO since the preceding steps take place in a hospital. The implementation of ECMO in hospital studies can be rapid, approximately 20 minutes in the Japanese study series and according to our experience. ECMO for out-of-hospital refractory CAs was the subject of a few clinical cases, in children and in sports events . Its feasibility by the ambulance service pre-hospital teams was confirmed in our last studies.

The improvement in survival with early ECMO, close to a 30-minute period of CPR should also be demonstrated. It is only based on the extrapolation of the results of very fast transport and almost without specialised resuscitation of victims of sudden death close to a hospital with ECMO.

Confirmation of this concept is therefore of particular importance and in fact:

• it would provide the prospect of a new treatment possibility for patients whose chances for survival are extremely slim, because prolonged CPR is required to have access to a hospital ECMO. It is an essential step before conducting a multi-centre randomised study to demonstrate the beneficial effect on survival.
• it would make it possible to stress the pertinence of the French teams' approach in this field, notably in comparison to European countries (Germany, Spain, etc...) that already have a medicalised pre-hospital emergency system, or that are currently developing it, like Japan.
• finally, it might also result in a better determination of the place of therapeutic ECMO and as a result, clarify the indications for organ harvesting after "cardiac death" in victims of pre-hospital sudden death.

In brief, the objective of this project is to evaluate the advantage of pre-hospital ECMO in improving patient survival.

Inclusion Criteria:

• Eligible patients have the following combination of criteria:
• Adults over 18 years of age and under 65 years of age
• And Refractory cardiac arrest (defined by the failure of professionals to resuscitate at the 20th minute of cardiac arrest with a minimum of 3 Automatic External Defibrillator (AED) or equivalent analyze
• And Beginning of external cardiac massage within the first 5 minutes after cardiac arrest (no flow < 5 min.) with shockable rhythm or the presence of signs of life during resuscitation (any rhythm): spontaneous movement, absence of mydriasis and/or pupillary response, spontaneous breathing movements
• And Medical cause of the cardiac arrest
• And End-Tidal CO2 (ETCO2) above 10 mm Hg at the time of inclusion
• And Absence of major co-morbidity. And Extra-corporeal Membrane Oxygenation (ECMO) team available and on-site before the 40th minute

Exclusion Criteria:

• Children under 18 years of age
• Adults over 65 years of age
• Period of more than 5 minutes without cardiac massage after collapsing
• Known co-morbidity that compromises the prognosis for short or medium-term survival
• Cardiac arrest during transportation times