Extracorporeal CPR in out-of-hospital cardiac arrest

Out-of-hospital cardiac arrest (OHCA) is a leading cause of mortality and disability worldwide. Despite the progress in the field of resuscitation, survival in patients treated with conventional cardiopulmonary resuscitation (basic life support and advanced life support) remains very low (<10%). In addition, many survivors have persistent and severe neurological damage.

Reynolds J et al. Circulation 2013; 128:2488-2494

The chances of survival after out-of-hospital cardiac arrest begin to decline rapidly after 10 minutes of CPR. After 35 minutes, less than 1% of patients achieve return of spontaneous circulation (ROSC) and survive with a favorable neurological outcome.

In selected patients with refractory out-of-hospital cardiac arrest and a potentially reversible cause, the latest guidelines recommend considering extracorporeal CPR (E-CPR).

What is extracorporeal CPR (E-CPR)?

The use of extracorporeal circulation for patients in refractory cardiac arrest was first suggested in 1976. Only in recent years the use of veno-arterial ECMO during ongoing CPR, a technique called E-CPR, has been more adopted in case of refractory cardiac arrest. Although E-CPR is a promising approach for patients with OHCA who do not get ROSC during conventional resuscitation (BLS + ALS), its role has not yet been fully understood.

Proceeding too quickly with extracorporeal CPR could unnecessarily expose patients who might still get ROSC to a highly invasive, costly, and significant additional risk procedure. Conversely, delaying E-CPR could reduce its potential benefit and increase the risk of irreversible brain damage and multi-organ failure. Furthermore, in patients treated with E-CPR, ROSC is often achieved after 60 minutes of cardiac arrest and it is not yet known whether the increasing use of E-CPR is increasing the number of survivors with neurological disabilities.

Today we have 2 randomized studies: ARREST trial and Prague OHCA study

In the Prague OHCA study, a randomized study published a few days ago in JAMA (Belohlavek et al. 2022), patients randomized to the invasive strategy (which included rapid transport to hospital with ongoing mechanical CPR and VA ECMO cannulation) had neurologically favorable survival (CPC 1-2) at 180 days of 32% compared to 22% of patients randomized to conventional ALS on-site (primary outcome, p = 0.09). The neurologically favorable survival at 30 days was 31% vs 18% (p = 0.02).

An unreasonable but widespread practice is to label randomized trials as positive or negative depending on whether the p-value for the primary outcome is 0.05. On the contrary, the interpretation of a randomized study should take place at 360 degrees, that is, considering also the secondary outcomes, the study design and the context. If a signal of treatment benefit (so-called trend) is to be inferred from a P value> 0.05 requires careful consideration. When the primary outcome of a study is completely neutral (e.g. narrow confidence interval around 1), the interpretation is simple: the intervention has no benefit.

In the case of the Prague OHCA study, interrupted early and likely to become underpowered, with a p value of 0.09 in the primary outcome equal to an improvement in survival of 10% and a significant improvement in the secondary outcome of survival at 30 days (p = 0.02), the study deserves a more constructive interpretation than a simple negative trial. In particular, considering the positive results of the first randomized study: in the ARREST study (Yannopoulos et al. Lancet. 2020), survival to discharge was 43% versus 7%.

Comparison between the two RCTs

The ARREST study randomized 30 patients with refractory OHCA with shockable rhythms, while the larger Prague OHCA study randomized 256 patients with refractory OHCA of presumed cardiac origin (any rhythm). In both studies (single-center RCT), patients were transported early with mechanical CPR in progress.

In the ARREST study, all refractory OHCAs (both groups) were transported early (after 3 shocks) with ongoing mechanical CPR and randomized upon arrival at hospital (ECMO or continuation of standard ALS in the emergency room). Patients in the ECMO group gained immediate access to the cath lab while those in the ALS group remained in the emergency room.

In contrast, in the Prague OHCA study, patients were randomized on the site of cardiac arrest. Patients randomized to the standard strategy were transported with ROSC after standard ALS or declared dead on the scene (if ROSC was not obtained). Not all patients with OHCA were routinely transported, a very typical model of the US prehospital emergency system. In the invasive strategy, patients were transported to hospital with ongoing CPR with the intention of proceeding with ECMO.

Both RCTs were terminated prematurely due to the significant benefits observed with the ECPR / hyperinvasive approach to interim analyzes. Therefore, these RCTs may have a reduced power to detect beneficial effects (which is why in the Prague OHCA study the difference did not reach the significance in the primary outcome – survival at 180 days – but only in the secondary outcome).

Meta-analysis of extracorporeal CPR

Considering the importance of the randomized studies published in the last year and a half, our group (Center for Intensive Care and Anestesiology, San Raffaele Hospital, Milan) conducted a systematic review and meta-analysis on the subject. Six studies were included, two randomized and four studies with propensity-score matching.

Patients with OHCA treated with E-CPR had higher survival with good neurological outcome (CPC 1-2) at longer follow-up (14% vs. 7.8%, p <0.001) and longer survival, with any neurological outcome (22% vs. 17%, p = 0.02). The number needed to treat (NNT) was found to be 16, i.e. 16 patients with refractory OHCA must be treated with E-CPR to obtain 1 survivor with an additional good neurological outcome.

In conclusion

Brain damage after an OHCA begins within minutes and is primarily determined by the time of ischemia / no-flow. The main goal should always be to reduce the no-flow time and early CPR (bystander) always remains the most effective way! Apps to alert first responders can be of great benefit, for example.

Again, to increase survival after an OHCA it is imperative to initially focus on CPR and bystander use of AEDs through community awareness and initiatives. In the absence of these basic manoeuvres, any further advanced intervention such as E-CPR will have little or no effect on survival.

It is important to note that in the Prague OHCA study, CPR was bystander initiated in 98% of the randomized OHCA cases! The Czech Republic is the 2nd country in Europe with the highest CPR rate (83%). This is crucial for E-CPR (and survival).

There is no magic bullet in cardiac arrest

Another important point: There is no magic bullet in the treatment of cardiac arrest! Extracorporeal CPR is part of a bundle of treatments that begin in the pre-hospital, continue during transport, and complete in the hospital. Close cooperation and protocols are key.

In the pre-hospital, it is critically important to ensure that bystanders start high-quality CPR immediately and that pre-hospital and on-site time is reduced. When a potential candidate for E-CPR is identified, advanced life support should be initiated in a timely manner. The patient must be prepared for transport to a cardiac arrest center capable of E-CPR, a mechanical CPR device must be placed, valid vascular access must be obtained, and an airway must be positioned definitive (better an endotracheal tube in this particular context).

Upon arrival at the hospital, E-CPR must be initiated immediately (no wasted time / arguments, eligibility must be evaluated prior to arrival!) and followed by post-cardiac arrest care including temperature control, advanced ventilatory and circulatory support, and definitive treatment (eg, coronary angioplasty). In the post-cardiac arrest phase, a comprehensive approach to mechanical circulatory support should be available, including IABP and left ventricular assist devices.


To achieve high survival rates in E-CPR programs, careful patient selection is critical. The next steps in this field are to understand which patients could benefit most, initiation in hospital or out of hospital, logistical aspects and generalizability to other contexts.

To date, the evidence is all in support of E-CPR although it is still of too low quality (single center studies, small sample, interrupted early). Therefore, large randomized and multicentre studies are still needed to make certain and definitive recommendations and confirm the results of the meta-analysis and of the 2 studies available so far. Those who already do extracorporeal CPR with excellent results and volumes can continue to do so because it is likely that they are saving lives, those who do not must consider whether it is feasible to implement it in their system and there are the conditions for it to be effective (e.g. bystander CPR? EMS effective in times and interventions? hospital skills and resources?).

These 2 RCTs are impressive work and the authors and team (pre-hospital, ED, ICU, Cath lab) need to be congratulated for proving that extracorporeal CPR is feasible and can save lives in high performing systems.

Main references

Leave a Reply

Your email address will not be published.