Haemodynamic implications of VA-ECMO vs. VA-ECMO plus Impella CP for cardiogenic shock in a large animal model

Peter H Frederiksen; Louise Linde; Emilie Gregers; Nanna L J Udesen; Ole K Helgestad; Ann Banke; Jordi S Dahl; Lisette O Jensen; Jens F Lassen; Amalie L Povlsen; Jeppe P Larsen; Henrik Schmidt; Hanne B Ravn; Jacob E Møller

doi:10.1002/ehf2.14780

Haemodynamic implications of VA-ECMO vs. VA-ECMO plus Impella CP for cardiogenic shock in a large animal model

ESC Heart Fail. 2024 Apr 22. doi: 10.1002/ehf2.14780. Online ahead of print.

Authors

Peter H Frederiksen¹, Louise Linde¹, Emilie Gregers², Nanna L J Udesen¹, Ole K Helgestad¹, Ann Banke¹, Jordi S Dahl¹, Lisette O Jensen^{1

3}, Jens F Lassen¹, Amalie L Povlsen⁴, Jeppe P Larsen⁴, Henrik Schmidt⁴, Hanne B Ravn^{3

4}, Jacob E Møller^{1

2

3}

Affiliations

¹ Department of Cardiology, Odense University Hospital, Odense, Denmark.
² Department of Cardiology, Heart Center, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark.
³ Department of Clinical Research, University of Southern Denmark, Odense, Denmark.
⁴ Department of Cardiothoracic Anaesthesiology, Odense University Hospital, Odense, Denmark.

PMID: 38649295
DOI: 10.1002/ehf2.14780

Abstract

Aims: Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) with profound left ventricular (LV) failure is associated with inadequate LV emptying. To unload the LV, VA-ECMO can be combined with Impella CP (ECMELLA). We hypothesized that ECMELLA improves cardiac energetics compared with VA-ECMO in a porcine model of cardiogenic shock (CS).

Methods and results: Land-race pigs (weight 70 kg) were instrumented, including a LV conductance catheter and a carotid artery Doppler flow probe. CS was induced with embolization in the left main coronary artery. CS was defined as reduction of ≥50% in cardiac output or mixed oxygen saturation (SvO₂) or a SvO₂ < 30%. At CS VA-ECMO was initiated and embolization was continued until arterial pulse pressure was <10 mmHg. At this point, Impella CP was placed in the ECMELLA arm. Support was maintained for 4 h. CS was induced in 15 pigs (VA-ECMO n = 7, ECMELLA n = 8). At time of CS MAP was <45 mmHg in both groups, with no difference at 4 h (VA-ECMO 64 mmHg ± 11 vs. ECMELLA 55 mmHg ± 21, P = 0.08). Carotid blood flow and arterial lactate increased from CS and was similar in VA-ECMO and ECMELLA [239 mL/min ± 97 vs. 213 mL/min ± 133 (P = 0.6) and 5.2 ± 3.3 vs. 4.2 ± 2.9 mmol/ (P = 0.5)]. Pressure-volume area (PVA) was significantly higher with VA-ECMO compared with ECMELLA (9567 ± 1733 vs. 6921 ± 5036 mmHg × mL/min × 10^-3, P = 0.014). Total diureses was found to be lower in VA-ECMO compared with ECMELLA [248 mL (179-930) vs. 506 mL (418-2190); P = 0.005].

Conclusions: In a porcine model of CS, we found lower PVA, with the ECMELLA configuration compared with VA-ECMO, indicating better cardiac energetics without compromising systemic perfusion.

Keywords: Cardiogenic shock; ECMELLA; Impella; Pressure‐volume loops; VA‐ECMO.