Extracorporeal versus conventional cardiopulmonary resuscitation for refractory out-of-hospital cardiac arrest: a secondary analysis of the Prague OHCA trial

Out-of-hospital cardiac arrest (OHCA) is one of the leading causes of death in Western countries [1]. Patients without prehospital return of spontaneous circulation (ROSC) bear a grave prognosis with survival rates as low as 4% [2‐4]. An increasing number of cardiac arrest centers worldwide have established a collaboration with emergency medical services using early transport from the field and extracorporeal life support (ECLS) implantation during ongoing cardiopulmonary resuscitation (CPR) when ROSC is not achieved conventionally, a method called extracorporeal cardiopulmonary resuscitation (ECPR). However, the current 2020 American Heart Association as well as the recent European Resuscitation Council guidelines provide a weak recommendation for ECPR which may be considered as a rescue method in selected patients when conventional CPR is failing, with low certainty of evidence [5, 6]. In addition, both guidelines highlighted the need for further research to define patients who would benefit from this intervention most [5, 6].

To date, two prospective randomized trials on ECPR in refractory OHCA were published, both testing “load and go” strategy with in-hospital ECLS cannulation. The first is the ARREST trial [7] which randomized 30 patients with refractory ventricular fibrillation only and was prematurely stopped due to superiority of ECPR and showed that ECPR is a feasible rescue option after prolonged unsuccessful ACLS, where standard approach has negligible chance for success [7]. The second prospective trial is the recently published Prague OHCA study [8] which enrolled 256 patients during on-scene ongoing ACLS to invasive arm (including intra-arrest transport for in-hospital ECPR and immediate invasive assessment) or standard ACLS. The invasive treatment did not significantly improve survival with good neurologically outcome at 180 days compared to standard ACLS in the intention to treat analysis but showed a beneficial effect of the invasive approach in 30-day neurological outcome and a subgroup of patients with prolonged CPR over 45 min [8]. Importantly, the anticipated statistical scenario of expected benefit provided by invasive approach was not reached due to higher-than-expected survival in the standard group [8]. Further, crossovers were allowed and 4 out of 10 (40%) patients crossed from standard to invasive ECPR treatment survived 180 days [8]. In addition, part of the prehospitally randomized patients in both arms experienced ROSC before reaching the hospital and thus were not candidates for ECPR [8]. All these factors might have influenced the effect of the ECPR treatment in the intention to treat analysis. Therefore, we performed this secondary analysis of the Prague OHCA study to evaluate whether successful ECPR might have been associated with improved outcomes.

In this secondary analysis of the randomized refractory OHCA trial, ECPR increased both 180-day survival and favorable neurological outcome in patients without prehospital ROSC compared to patients treated with prolonged conventional ACLS only. In a multivariate Cox regression analysis, the use of ECPR was significantly associated with 180-day survival. This result is further supporting ECPR as an increasingly used method for r-OHCA and is consistent with previously published observational studies as well as the one randomized trial [7, 11‐15].

Although proper selection of patients who will benefit from ECPR is essential, to date there is no consensus about the criteria for starting intra-arrest transport and implementing ECPR [5, 6]. In addition, significant differences in ECPR protocols between cardiac arrest centers exist [5, 7, 8] and currently published data regarding predictors of survival in r-OHCA were based on evidence from observational studies only [17, 18]. The results of multivariate analysis in this study indicate that prehospital ROSC, shockable initial rhythm, shorter time of resuscitation as well as younger age are all positively associated with 180-day survival in r-OHCA confirming findings from observational studies and systematic reviews [15‐19]. However, further research is needed to achieve consensus regarding optimal ECPR strategy as excluding certain subgroup of patients without sufficient data may inappropriately limit patient care [20]. The same is true for ECPR timing as too early transport may decrease chances of achieving prehospital ROSC [4, 19], a major determinant of survival, but later transport and longer low flow time are associated with decreased survival despite ECLS implantation [15, 19]. The most relevant finding regarding the intra-arrest transport timing is derived from an observational study suggesting that ECPR should be considered between 8 and 24 min of professional on-scene resuscitation, with 16 min balancing the risks and benefits of early and later transport [19]. Prague OHCA was the first r-OHCA trial randomizing patients prehospitally during ongoing CPR in the field [8]. Patients were randomized on average after 25 min of ongoing OHCA including 15 min of ACLS, reflecting a truly refractory cardiac arrest [8]. Despite that, almost one-third of enrolled patients in the invasive arm still achieved sustained ROSC prehospitally, en route or immediately after admission, most of them having initial shockable rhythm. This highlights the need of continuous high-quality ACLS during transport to cardiac arrest center and also the urgency of further research in this area as the key question (whether conventional CPR non-responders and candidates for ECPR can be identified early during CPR) remains unanswered.

The overall 180-day survival rate for patients treated with ECPR was 23.9% in this study, which is comparable to prior observational studies reporting survival rates from 12 to 33% [11‐15]. It is lower compared to the results of the ARREST trial where 6 out of 14 patients survived (43%) as this study included patients with an initial shockable rhythm only [7]. Nonetheless, survival rate of patients with an initial shockable rhythm in the invasive arm of the Prague OHCA study was actually 48.6% [8] corresponding to the ARREST trial. The Prague OHCA trial also provided randomized data confirming a vast difference in r-OHCA outcomes between patients with initial shockable and non-shockable rhythms [8]. Results of this secondary analysis further confirm poor outcomes of non-shockable rhythms despite ECPR treatment. These findings are supporting current clinical practice in many systems which limit ECPR service to patients with an initial shockable rhythm [7, 15]. Moreover, our results confirmed that patients without prehospital ROSC have very low chances to survive even with prolonged (median time 66 min) conventional ACLS without ECPR which is in line with previous findings [3, 4, 7, 13]. Based on the current evidence from observational study [15] as well as the randomized trials [7, 8], it is obvious that the subgroup of patients with an initial shockable rhythm and prolonged CPR over 45 min benefit most from the ECPR approach [8]. However, it is important to underline that ECPR must be considered early and provided in a well-established system with close cooperation between EMS and ECPR cardiac arrest center to achieve good outcomes [7, 8] as survival rates lower than 4% were reported in patients transported without field ROSC from observational studies [2, 3].

Further, almost all randomized patients in both prospective ECPR studies [7, 8] had witnessed arrest with high rate of bystander CPR which is another important prerequisite for good outcomes. Moreover, only 6% of all OHCA patients were enrolled in the Prague OHCA trial which is in line with previous reports [21, 22]. This confirms that ECPR is not a substitute for conventional ACLS but rather complementary method for properly selected refractory OHCA patients provided in the well-organized system [7, 8]. Continuous efforts to achieve maximum rates of bystander CPR are extremely important as these are associated with favorable long-term outcomes and may also increase the pool of patients considered for ECPR [7, 8, 23].

In addition, the current analysis confirms high rates of bleeding complications associated with invasive approach and ECPR [2]. Bleeding is an important limitation of ECLS therapy in all indications, especially among ECPR patients who underwent prolonged resuscitation attempts. Despite the substantial rates of bleeding, these complications were the leading cause of death in a small proportion of ECPR patients in our study (4.2%).

Neurological outcome results at 180 days in this study revealed that majority of survivors had good neurological outcome (mainly CPC 1), and only few patients treated with ECPR survived 180 days with a poor neurological outcome, similarly to patients treated conventionally. However, brain death was the third most common cause of death in the study, and irreversible brain damage is a major barrier to achieve better outcomes in refractory OHCA [24]. Further, data regarding long-term outcomes and quality of life in ECPR survivors are scarce [25] and more information is needed.

In this secondary analysis of the randomized r-OHCA trial, ECPR was associated with improved 180-day survival in patients without prehospital ROSC. Initial shockable rhythm, younger age and shorter time of resuscitation were all associated with better 180-day survival in r-OHCA. Majority of r-OHCA survivors treated by ECPR had good neurological outcome at 180 days.