https://orcid.org/0000-0002-6030-3957
This editorial refers to ‘The association between time to extracorporeal cardiopulmonary resuscitation and outcome in patients with out-of-hospital cardiac arrest', by Shoji Kawakami et al., https://doi.org/10.1093/ehjacc/zuac010.
Extracorporeal cardiopulmonary resuscitation (ECPR) remains a debated treatment for out-of-hospital cardiac arrest (OHCA). The ARREST trial which randomized 30 cardiac arrest patients to ECPR or conventional resuscitation, demonstrated that significantly more ECPR patients (43 vs. 0%) were alive with favourable neurological outcome [cerebral performance category (CPC) Scale 1 or 2] 3 months after hospital discharge.1 It should be noted that the ARREST trial was conducted in Minneapolis, MN, USA, by a group having an extensive expertise in the management of cardiac arrest, including the use extracorporeal membrane oxygenation (ECMO).2 More recently, the Prague OHCA Study Group randomized 264 patients with witnessed OHCA of presumed cardiac aetiology to either an invasive strategy including a bundle of early intra-arrest transport, ECPR, and invasive assessment or to the continuation of on-site conventional resuscitation.3 The primary endpoint of CPC 1–2 survival at Day 180 was not statistically different between groups (ECPR: 31.5%, standard care: 22%, P = 0.09), while the cumulative probability of overall survival was higher in the ECPR group (log-rank test P = 0.01). Since this study was terminated early at the recommendation of the data and safety monitoring board when prespecified criteria for futility were met, it was possibly underpowered to detect a clinically relevant difference. It should also be noted that 8% of standard group patients crossed over to ECPR and that the invasive approach was associated with an increased risk of bleeding complications.
Yet, there are still many obstacles to the replication of such encouraging results in everyday practice (Figure 1). Hospital survival was only 8% in the largest real-life ECPR registry study reporting on 525 patients of the Paris region,4 while long-term neurological outcome was not evaluated in this study. Heterogeneity in patients’ selection and ECPR indications, pre-hospital systems of care, time from arrest to ECPR, and post-arrest interventions such as coronary angiography and targeted temperature management (TTM) may explain important differences between series in the rate of favourable outcome after ECPR. As underlined in recent resuscitation guidelines which provided only weak recommendation for ECPR,5 the indications and the risk-to-benefit ratio of this technique should now be carefully evaluated.
Unanswered questions regarding indications and implementation of ECPR. ECPR, extracorporeal cardiopulmonary resuscitation; CA, cardiac arrest; CPR, cardiopulmonary resuscitation; etCO2, end-tidal carbon dioxide concentration; TTM, targeted temperature management; PCA, percutaneous coronary angiography. This figure has been designed using resources freely available at Flaticon.com.
In this issue of the European Heart Journal: Acute Cardiovascular Care, Kawakami and colleagues report the largest cohort of ECPR patients to date. Patients with OHCA of suspected cardiac cause admitted to 87 centres participating in the Japanese Association for Acute Medicine-OHCA registry were prospectively included between June 2014 and December 2017. Of the 34 754 patients with OHCA, 695 received ECPR, with only 77 (11%) reaching the primary endpoint of survival with favourable neurological outcome (CPC 1–2) 30 days after cardiac arrest. This study provides a nationwide evaluation of ECPR in a resource-rich country. It confirms that ECPR is a niche indication applied to only 2% of OHCA, as already reported by the Sudden Death Expertise Center in Paris.4 While the rate of favourable outcome was better than in the French series, it is still lower than in studies published by the Minnesota group (33–43%)1,2 or than in other retrospective series (19–32%).6,7
Kawakami and colleagues found that the call-to-ECPR initiation interval [per minute increase odds ratio (OR) 0.96, 95% confidence interval (CI) 0.94–0.99, P = 0.001] was one of the most important predictor of a favourable outcome. This observation is in agreement with most previous ECPR series that demonstrated better outcomes with shorter no-flow or low-flow times.8 Noticeably, the rate of survival with good neurological outcome dropped drastically when this time interval exceeded 70 min (Figure 2 and Graphical abstract). It should also be noted that this specific time interval may be more accurate and more widely applicable than the collapse-to-ECPR interval, as the exact time of patient’s collapse is less frequently precisely recorded. This variable may indeed be included in future algorithms and guidelines defining the indications of ECPR.
Other independent predictors of a better outcome were percutaneous coronary intervention (PCI, OR 2.3, 95%CI 1.1–4.7, P = 0.02) and TTM (OR 2.3, 95%CI 1.1–4.6, P = 0.02). Of note, 80% of patients with TTM underwent targeted hypothermia. Although the association of TTM with the more favourable neurological outcome has already been reported in previous ECPR series,9 it should be interpreted with caution. The decision of TTM was not protocolized and made by the physician in charge, suggesting that a selection bias may have persisted despite multivariable adjustment. Furthermore, results of recent large randomized controlled trials do not support the use of hypothermia vs. targeted normothermia at 36°C after cardiac arrest.10,11 In the recent randomized HYPOECMO trial12 in cardiogenic shock patients under venoarterial ECMO, moderate hypothermia (33–34°C) did not significantly increase survival compared with strict normothermia (36–37°C) [adjusted OR, 0.71 (95%CI 0.45–1.13), P = 0.15]. It should also be noted that although 50% of HYPOECMO patients had cardiac arrest before ECMO, this was not a trial of ECPR for OHCA and patients who had received CPR for longer than 45 min were excluded.
In the study by Kawakami and colleagues, nearly all survivors (99%) underwent percutaneous coronary angiography and the positive association between PCI and better survival was also previously reported in a large Korean nationwide ECPR database.9 Alternatively, two recent randomized control trials found no benefit of immediate vs. delayed coronary angiography in cardiac arrest without ST-segment elevation13,14 but none included ECPR patients. The optimal timing of coronary intervention in ECPR patients may be evaluated in future trials.
In summary, Kawakami and colleagues should be commended for reporting the results of this large nationwide cohort, which provides important information regarding the care of patients with OHCA refractory to conventional resuscitation. The indications and the risk-to-benefit ratio of this technique should continue to be carefully evaluated using similar structured national or international registries and future large randomized trials.
Conflict of interest: none declared.
References
Author notes
The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.
Comments