Early versus delayed impella unloading during venoarterial ECMO for cardiogenic shock: a propensity-matched analysis
Journal of Artificial Organs · 2026-12 · 29 (4)
DOI: 10.1007/s10047-026-01606-2 ↗Kurz gefasst
Vergleichsstudie in der Zeitschrift Journal of Artificial Organs, online erschienen am 13.11.2026. Verfasst von Aabha Divya und 1 weiteren Autor:innen. Beteiligte Einrichtung: Division of Cardiothoracic Surgery, Tulane University School of Medicine, 1430 Tulane Ave, New Orleans, LA, 70112, USA. adivya@tulane.edu. Hauptthemen laut MEDLINE: Shock, Cardiogenic, Heart-Assist Devices, Extracorporeal Membrane Oxygenation. Die Arbeit nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Venoarterial extracorporeal membrane oxygenation (VA-ECMO) restores perfusion in cardiogenic shock but increases left-ventricular afterload. Impella unloading (ECPELLA) is increasingly used, but optimal timing remains uncertain. We retrospectively analyzed adults with cardiogenic shock treated with VA-ECMO and Impella from 2016 to 2025. Early ECPELLA was Impella placement from the calendar day before through the day after VA-ECMO initiation; delayed ECPELLA was placement 2-7 days after VA-ECMO initiation. One-to-one propensity-score matching balanced 35 covariates. The primary outcome was 30-day all-cause mortality. Sensitivity analyses included acute kidney injury (AKI) restriction, VA-ECMO-indexed follow-up, and a 7-day landmark analysis. Matching yielded 282 per group. Mortality was similar at 30 days (48.9% vs. 49.6%; risk ratio [RR] 0.99, 95% CI 0.83-1.17), 90 days (RR 1.01, 95% CI 0.87-1.17), and 1 year (RR 1.03, 95% CI 0.89-1.18). Early ECPELLA was associated with more renal-replacement therapy (RRT) at 30 days (37.4% vs. 21.7%; RR 1.73, 95% CI 1.29-2.31), but not when follow-up began at VA-ECMO initiation in AKI-restricted cohorts (RR 1.00, 95% CI 0.81-1.22; adjusted hazard ratio 1.02, 95% CI 0.82-1.26). Major bleeding, sepsis, and limb ischemia were similar. Durable LVAD implantation was less frequent with early ECPELLA at 1 year (RR 0.57, 95% CI 0.34-0.95), but not in sensitivity analyses. Early ECPELLA was not associated with improved survival compared with delayed ECPELLA. The RRT association was not reproduced with aligned follow-up. Residual confounding and immortal-time bias limit causal inference and warrant randomized trials.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Division of Cardiothoracic Surgery, Tulane University School of Medicine, 1430 Tulane Ave, New Orleans, LA, 70112, USA. adivya@tulane.edu.
- Department of Anesthesiology and Perioperative Medicine, Tulane University School of Medicine, Murphy Building, LA, 70112, New Orleans, USA.
Zitieren
APA
Divya, A., & Dalal, N. (2026). Early versus delayed impella unloading during venoarterial ECMO for cardiogenic shock: a propensity-matched analysis. Journal of Artificial Organs, 29(4). https://doi.org/10.1007/s10047-026-01606-2
Vancouver
Divya A, Dalal N. Early versus delayed impella unloading during venoarterial ECMO for cardiogenic shock: a propensity-matched analysis. Journal of Artificial Organs. 2026;29(4). doi:10.1007/s10047-026-01606-2
BibTeX
@article{Divya2026,
author = {Divya, Aabha and Dalal, Niti},
title = {Early versus delayed impella unloading during venoarterial ECMO for cardiogenic shock: a propensity-matched analysis},
journal = {Journal of Artificial Organs},
year = {2026},
volume = {29},
number = {4},
doi = {10.1007/s10047-026-01606-2}
}
EUVIMED