Comparison of mid-term mortality after surgical, supported or unsupported percutaneous revascularization in patients with severely reduced ejection fraction: A direct and network meta-analysis of adjusted observational studies and randomized-controlled

The optimal revascularization strategy in patients with heart failure with reduced ejection fraction (HFrEF) remains to be elucidated. The aim of this paper is to compare the mid-term mortality rate among patients with severely reduced ejection fraction (EF) and complex coronary artery disease who underwent coronary artery bypass grafting (CABG), percutaneous coronary intervention (PCI) with Impella support, or without.Randomized control trials and propensity-adjusted observational studies including patients with ischemic cardiomyopathy (ICM) and severe EF reduction undergoing revascularization were selected. Different revascularization strategies (CABG, supported PCI, and PCI without Impella) were compared in pairwise and network meta-analysis. The primary endpoint was mid-term mortality (within the first year after revascularization).Fifteen studies, mostly observational (17,841 patients; 6779 patients treated with CABG, 8478 treated with PCI without Impella, and 2584 treated with Impella-supported PCI) were included in this analysis. The median age was 67.8 years (IQR 65-70.1), 21.2% (IQR 16.4-26%) of patients were female sex, and a high prevalence of cardiovascular risk factors was noted across the entire population. At pairwise analysis, CABG and PCI without Impella showed similar one-year all-cause mortality (10.6% [IQR 7.5-12.6%] vs 12% [IQR 8.4-11.5%]) RR 0.85 CI 0.67-1.09, while supported PCI reduced one-year all-cause mortality compared to PCI without Impella (9.4% [IQR 5.7-12.5%] vs 10.6% [IQR 8.9-10.7%]) RR 0.77 CI 0.6-0.89. At network meta-analysis, supported PCI showed better results (RR 0.75, 95% CI 0.59-0.94) compared to CABG.Our analysis found that supported PCI may have a benefit over standard PCI in patients in direct comparison, and over CABG from indirect comparison, and with HFrEF undergoing revascularization. Further RCTs are needed to confirm this result. (PROSPERO CRD42023425667).