Procedural characteristics and outcomes of patients undergoing Impella-assisted high-risk percutaneous coronary...

Impella is a catheter-based, continuous blood flow left ventricle assist device used in selected patients undergoing high-risk percutaneous coronary interventions (HR PCI). We aimed to evaluate outcomes in patients undergoing Impella-assisted HR-PCI and identify independent predictors of 12-month mortality.

Consecutive HR-PCI patients enrolled in the national, multicentre, retrospective IMPELLA-PL registry (n = 253) in 20 Polish interventional cardiological centres from October 2014 until December 2021 were included in the analysis. The main endpoints were (i) procedural success defined as revascularization of all preplanned lesions, (ii) device-related complications, (iii) 12-month mortality and major adverse cardiovascular events (MACE).

The majority of patients presented with multivessel disease including left main (63.6 %). The median Syntax Score II was 43.0 (32.4-55.0). The procedural success was achieved in 83.0 % of patients. Device-related complications included access site bleeding (14.6 %), limb ischemia (2.4 %) and hemolysis (1.6 %). The in-hospital MACE included 1 cardiosurgical intervention (0.4 %), 12 exacerbations of heart failure (4.7 %), 11 myocardial infarctions (4.3 %), 32 cases of acute kidney injury (12.6 %), 35 inflammatory complications (13.8 %) and 32 major bleeding complications (13.4 %). In-hospital mortality rate was 8.3 %, 12-month mortality rate was 18.2 % and MACE rate post-discharge was 22.5 %. The 12-month-mortality was increased by pre-existing, atrial fibrillation (OR 3.50, 95 % CI 1.38-8.95) and chronic kidney disease (OR 2.77, 95 % CI 1.06-7.26) and decreased by Impella removal in the cath-lab (OR 0.11, 95 % CI 0.02-0.76) and RAAS inhibitor use (OR 0.26, 95 % CI 0.08-0.89).

Despite high anatomical complexity of coronary artery disease of patients included in the IMPELLA-PL registry, the procedural success rate was relatively high and the mortality relatively low.