Acute myocardial infarction (AMI) vs. Non-AMI cardiogenic shock treated with microaxial flow pump support: insights from the multicenter MMERGE registry

Background percutaneous microaxial flow pump support (mAFP – Impella®) is increasingly adopted in cardiogenic shock (CS) for its ability to actively unload the ventricles, reduce end-diastolic pressure, minimize cardiac work and augment cardiac output. However, pathophysiological differences exist between acute myocardial infarction-related CS (AMI-CS) and non-AMI-CS. Purpose To assess differences in baseline features, in-hospital complications, and outcomes after Impella implantation according to CS etiology in a real world contemporary cohort of patients with CS treated with mAFP. Methods We analyzed 449 CS patients treated with Impella from the internatonal multicenter IMMERGE registry, including 266 AMI-CS and 180 non-AMI-CS. The primary outcome was all-cause mortality at 180 days and at long term follow-up. Additional outcomes included a composite of all-cause death, durable LVAD implantation, or heart transplantation. Baseline characteristics, in-hospital complications and follow up outcomes were compared between the two groups, with further analyses stratified by SCAI shock stage. Results At baseline, AMI-CS patients were older (64 versus 52 years) and had a higher prevalence of cardiovascular risk factors (hypertension 62% vs 27%, hypercholesterolemia 57% vs 26%, diabetes 40% vs 22%). Nearly half of those presented with STEMI while 15% had NSTE-ACS. Non-AMI-CS cohort was predominantly due to dilated (48%) or ischemic (27%) ADHF in cardiomyopathy, myocarditis (20%) or postcardiotomy shock (5%). Compared with AMI-CS, non-AMI-CS patients had larger LV dimensions and more frequent right ventricular dysfunction, longer support duration (7 vs 3 days, p<0.0001) and greater need for Impella 5.0 and 5.5 (27% vs 7%). Intrahospital complications were similar, except from higher incidence of hemolysis and device thrombosis in non-AMI CS. At 180 days, mortality was higher in AMI-CS (65.7% vs 51.7%, p=0.001), a difference persisting at long-term (p=0.01). Non-AMI-CS patients had lower mortality but higher rates of LVAD (20% vs 6%, p<0.0001) and heart transplantation (19% vs 4%, p<0.0001). According to SCAI stratification, 180-day mortality in AMI-CS cohort rose progressively with advancing SCAI stage (50% in C, 69% in D, and 78% in E), showing a clear stage-dependent gradient of risk. Conversely, in non-AMI CS cohort, mortality was 34% in C, 56% in D, and 64% in E, with overlapping D/E curves. Conclusions AMI-CS and non-AMI CS treated with mAFP in real world practice have different baseline characteristics and different outcomes. In AMI-CS, Impella may facilitate myocardial recovery, but outcomes are poor if implantation is delayed to advanced shock stages. In non-AMI-CS, Impella provides a safe bridge to LVAD or transplantation. Early implantation (SCAI C) should be pursued in both settings before progression to refractory shock.