Reply to Zhang <i>et al</i>.

Dear Editor,We thank Zhang et al. [1] for their insightful comments on our study [2].One concern was that percutaneous coronary intervention (PCI) should be the first-line treatment in the setting of acute myocardial infarction complicated by cardiogenic shock.Our study further focused on a special group of patients who developed a Society for Cardiovascular Angiography and Intervention stage E A acute myocardial infarction complicated by cardiogenic shock and underwent extracorporeal cardiopulmonary resuscitation.Currently, no recommendations on the revascularization strategy in these challenging cases have been published [3,4].We absolutely acknowledged that more rapid revascularization in a less-invasive approach could be achieved with PCI compared with coronary artery bypass graft surgery (CABG).Indeed, the clinical predominance of PCI was reflected in our study, which included only 38 patients who received CABG among 148 patients with triple vessel disease.Despite a prolonged extracorporeal membrane oxygenation-to-revascularization time (280 min for CABG versus 90 min for PCI, P 0.005), hospital outcomes were more favourable after emergent CABG than after PCI [2].Several plausible mechanisms in favour of CABG have been discussed.First, cardiac interventionists might be reluctant to perform multivessel PCI in such patients with haemodynamic instability [5].Second, the potential benefits from active left ventricle (LV) unloading before revascularization may be more prominent for post-resuscitate myocardial stunning [6].In our practice, we preferred on-pump beating-heart technique for emergent CABG, which could offer simultaneous full circulatory support and effective ventricular unloading during revascularization.A 2nd concern was that the PCI outcomes in unmatched patients with triple vessel disease were not reported in our study.Among the 110 patients with triple vessel disease who received PCI, rates of successful extracorporeal membrane oxygenation (ECMO) weaning and hospital survival without the need for heart transplantation or durable ventricular assist device were 52.4% and 29.5%, respectively.Because the entire cohort had incorporated a heterogeneous patient population, the outcomes should not be interpreted without clarifying the inherent differences between patients receiving the 2 revascularization strategies.For example, compared with patients in the CABG group, more patients in the unmatched PCI group had experienced out-of-hospital cardiac arrest, a wellestablished unfavourable prognostic factor in patients who received extracorporeal cardiopulmonary resuscitation.Therefore, we have adjusted the inter-group differences using propensity-score matching.Third, propensity score matching (PSM) may not be the ideal matching method in studies of small sample size.To maintain the statistical power of this study, we found it difficult to perfectly match each baseline characteristic in a relatively small cohort.In agreement with Zhang et al., an appropriately powered randomized trial comparing CABG and PCI is desired to determine the optimal revascularization strategy for patients who received extracorporeal cardiopulmonary resuscitation.