Left ventricular hemodynamics with veno-arterial extracorporeal membrane oxygenation

Background: There is considerable debate about the hemodynamic effects of veno-arterial extracorporeal membrane oxygenation (VA-ECMO).

Aims: To evaluate the changes in left ventricular (LV) function, volumes, and work in patients treated with VA-ECMO using invasive LV catheterization and three-dimensional echocardiographic volumes.

Methods: Patients on VA-ECMO underwent invasive hemodynamic evaluation due to concerns regarding candidacy for decannulation. Hemodynamic parameters were reported as means±standard deviations or medians (interquartile ranges) after evaluating for normality. Paired comparisons were done to evaluate hemodynamics at the baseline (highest) and lowest tolerated levels of VA-ECMO support.

Results: Twenty patients aged 52.3 ± 15.8 years were included. All patients received VA-ECMO for refractory cardiogenic shock (5/20 SCAI stage D, 15/20 SCAI stage E). At 3.0 (2.0, 4.0) days after VA-ECMO cannulation, the baseline LV ejection fraction was 20% (15%, 27%). The baseline and lowest VA-ECMO flows were 4.0 ± 0.6 and 1.5 ± 0.6 L/min, respectively. Compared to the lowest flow, full VA-ECMO support reduced LV end-diastolic volume [109 ± 81 versus 134 ± 93 mL, p = 0.001], LV end-diastolic pressure (14 ± 9 vs. 19 ± 9 mmHg, p < 0.001), LV stroke work (1858 ± 1413 vs. 2550 ± 1486 mL*mmHg, p = 0.002), and LV pressure-volume area (PVA) (4507 ± 1910 vs. 5193 ± 2388, p = 0.03) respectively. Mean arterial pressure was stable at the highest and lowest flows (80 ± 16 vs. 75 ± 14, respectively; p = 0.08) but arterial elastance was higher at the highest VA-ECMO flow (4.9 ± 2.2 vs lowest flow 2.7 ± 1.6; p < 0.001).

Conclusions: High flow VA-ECMO support significantly reduced LV end-diastolic pressure, end-diastolic volume, stroke work, and PVA compared to minimal support. The Ea was higher and MAP was stable or minimally elevated on high flow.