Early extracorporeal CPR for refractory out-of-hospital cardiac arrest - A pre-planned per-protocol analysis of the INCEPTION-trial

BACKGROUND: Evidence for extracorporeal cardiopulmonary resuscitation (CPR) in refractory out-of-hospital cardiac arrest (OHCA) remains inconclusive. Recently, the INCEPTION-trial, comparing extracorporeal with conventional CPR, found no statistically significant difference in neurologically favorable survival. Since protocol deviations were anticipated, a pre-specified per-protocol analysis was foreseen. METHODS: The per-protocol analysis of the INCEPTION trial excluded patients not meeting inclusion or exclusion criteria, amongst which time-to-cannulation of >60 minutes, and achieving a return of spontaneous circulation before hospital arrival. Crossovers were excluded as well. The primary outcome (30-day survival in a neurologically favorable condition; cerebral performance category [CPC] 1-2) was primarily analyzed under a frequentist statistical framework. In addition, Bayesian analysis under a minimally informative prior was performed. RESULTS: Eighty-one patients were included in the per-protocol analysis (extracorporeal CPR n = 33, conventional CPR n = 48). Thirty-day survival with CPC1-2 was 15% in the extracorporeal CPR group versus 9% in the conventional CPR group (adjusted OR 1.9; 95% CI 0.4-9.3; p-value 0.393). Bayesian analysis showed an 84% posterior probability of any ECPR benefit and a 61% posterior probability of a 5% absolute risk reduction for the primary outcome. CONCLUSION: A pre-planned, pre-specified per-protocol analysis of the INCEPTION-trial, found a higher survival with favorable neurological in patients undergoing ECPR versus CCPR for refractory shockable OHCA. This difference did not reach statistical significance, but results should be interpreted with care, in the light of the small remaining sample size.