Left ventricular unloading with impella versus IABP in cardiogenic shock requiring VA-ECMO
R Ito, T Kondo, S Kazama, T Araki, T Imaizumi, T Nabeta, Y Ikeda, H Matsuzoe, K Matsumura, K Kanaoka, T Nagai, T Fujino, R Morimoto, T MuroharaAbstract
Background
Venoarterial extracorporeal membrane oxygenation (VA-ECMO) is increasingly used for refractory cardiogenic shock (CS), but it augments left ventricular (LV) afterload and may impede myocardial recovery. Mechanical LV unloading with intra-aortic balloon pump (IABP) or Impella is frequently adopted; however, the optimal unloading strategy during VA-ECMO remains controversial, and etiology-specific comparative data are limited.
Purpose
To compare clinical outcomes of Impella versus IABP as LV unloading strategies in patients with CS requiring VA-ECMO and to examine its differences according to CS etiology.
Methods
From the Japanese Registry of All Cardiac and Vascular Diseases–Diagnosis Procedure Combination database (2012–2023), we identified adult patients with CS treated with VA-ECMO plus Impella or IABP. The primary outcome was in-hospital mortality. Secondary outcomes included length of hospital stay, duration of mechanical circulatory support (MCS), and hospitalization costs. Propensity score–based inverse probability of treatment weighting (IPTW) was applied to adjust for baseline differences.
Results
Among 18,906 patients (Impella n=1,688; IABP n=17,218), in-hospital mortality was 57.3% in the Impella group and 67.2% in the IABP group. After IPTW adjustment, Impella use was independently associated with lower in-hospital mortality (odds ratio [OR] 0.78, 95% confidence interval [CI] 0.70–0.88). In etiology-specific analyses, Impella was associated with lower mortality in heart failure (OR 0.62, 95% CI 0.44–0.88) and arrhythmia (OR 0.53, 95% CI 0.37–0.76), whereas associations were attenuated and not statistically significant in acute myocardial infarction and other etiologies. Impella use was also associated with longer hospital stay (median 22 vs 12 days), longer MCS duration (7 vs 4 days), and higher hospitalization costs (56.1 vs 27.3 thousand USD; p<0.001).
Conclusion
In this large nationwide cohort, Impella unloading during VA-ECMO was associated with lower in-hospital mortality compared with IABP, particularly in selected etiologies such as heart failure and arrhythmia, while requiring greater resource utilization. These findings support the need for etiology-informed patient selection and prospective randomized trials to clarify the optimal LV unloading strategy in CS.figure2