Early outcomes of microaxial flow pump (Impella)–supported coronary artery bypass grafting for acute myocardial infarction with cardiogenic shock: national reference estimates from the J-PVAD registry

Background Contemporary national data on outcomes of coronary artery bypass grafting (CABG) performed under microaxial flow pump (Impella) support for acute myocardial infarction (AMI) complicated by cardiogenic shock are scarce. We aimed to provide contemporary national reference estimates and, as a post hoc exploratory analysis, to describe outcomes according to the setting of Impella insertion. Methods From the Japanese Registry for Percutaneous Ventricular Assist Devices (J-PVAD; February 2020 to December 2024), we analyzed 245 patients from 116 centers with AMI-related cardiogenic shock who underwent CABG under Impella support, after excluding ventricular septal perforation repair and unsuccessful insertion. The primary outcome was all-cause mortality within 30 days after Impella implantation; in-hospital mortality and adverse events were secondary. Proportions are reported with Wilson 95% confidence intervals (CIs). Operating-room insertion ( n = 19) was compared with non-operating-room insertion ( n = 226) by Firth-penalized logistic regression with propensity score overlap weighting (age, systolic blood pressure, and extracorporeal membrane oxygenation initiated before Impella). Results Thirty-day mortality was 20.8% (51/245; 95% CI: 16.2–26.3); in 24 patients discharged alive before day 30, survival to day 30 was assumed from the recorded status. In-hospital mortality was 33.1% (81/245; 95% CI: 27.5–39.2). Adverse events included bleeding (17.6%; 95% CI: 13.3–22.8), renal failure (9.8%), sepsis (9.0%), hemolysis (8.6%), and stroke (6.5%). Thirty-day mortality was 36.8% (7/19) after operating-room insertion and 19.5% (44/226) after non-operating-room insertion (weighted odds ratio 1.63; 95% CI: 0.62–4.16). This estimate had a wide confidence interval, consistent with either a substantially lower or a substantially higher mortality after operating-room insertion, and more than half of the baseline characteristics remained imbalanced after weighting. Conclusion These nationwide data provide contemporary reference estimates for Impella-supported CABG in AMI-related cardiogenic shock. The registry records the setting (operating room or elsewhere), but not the intent or clinical timing, of Impella insertion, and the comparative safety of operating-room vs. non-operating-room insertion cannot be determined from these data.