Flow Protocols in VA-ECMO: Impact on PCWP in Cardiogenic Shock

The use of peripheral femoro-femoral venoarterial (VA) extracorporeal membrane oxygenation (ECMO) has globally consolidated and is the most widely used in patients with OMT refractory cardiogenic shock. However, the most recent randomized clinical trials have failed to show survival benefits in unselected populations. Therefore, correctly identifying the phenotype of critical patients that might benefit from this strategy results vital. 

From a physiological point of view, VA-ECMO high flow has been associated with left ventricular (LV) distension and increased filling pressure (pulmonary capillary wedge pressure or PCWP), consequence of a retrograde flow, which increases the risk of pulmonary edema. This is why experts recommend using the least necessary flow to ensure adequate perfusion. 

In this context, the PAPO-Flow Study group sought to determine what actually happens with incremental flow, with both invasive and non-invasively measuring protocols.

This was a single center study (Hôpital Pitié-Salpêtrière, París, France) including 80 patients with cardiogenic shock, monitored with pulmonary artery catheter (PAC or Swan-Ganz) undergoing standardized flow variation protocols from 2 to 4 L/min in 0.5-L/min increments within the first 48 hours. The remaining therapeutic measures (inotropes, ventilation, sedation) were kept constant. In patients with intra-aortic balloon pump (IABP), this was removed before protocol offset. 

Read also: Tricuspid Regurgitation, TriBicaval Registry.

Baseline patient characteristics were mean age 49, 80% men, mean SOFA score 55, most were SCAI type D, inotropic score was 20, 25% were post cardiac arrest and 11% required intra-arrest cannulation. 

And afterload flow increase was expected and, in turn, PCWP rise. Baseline mean PCWP was 15 mmHg (only one third ≥18 mmHg). The incremental flow resulted in ≥20% PCWP drop in 36% of patients, 58% showed stable PCWP and only 6% showed a rise. The global trend was a mild reduction in PCWP (from 15 to 14 mmHg; p=0.003).

This came about together with reduced LV end-systolic diameter, aortic VTI and systolic volume index, with no significant change in LV ejection fraction. There was a reduction in cardiac index (from 2.4 to 2.1 L/min/m²; p<0.01), even though left ventricular stroke work remained stable. As to the right side, central venous pressure (CVP) decreased, and RV-pulmonary artery coupling improved (FAC/PAPs, p<0.01).

Read also: IVI-Guided PCI vs Coronary Artery Bypass Surgery in Left Main Coronary or Multivessel Disease.

The authors suggest increased LV afterload is counterbalanced by reduced RV preload, leading to reduced pulmonary flow; therefore, PCWP will not rise as expected. 

Reduced PCWP predictors were: better biventricular pressure, more elevated mean arterial pressure at baseline and lower CVP. At exploratory analysis, patients with decreased PCWP presented better ICU survival (24% vs. 57%; HR 0.45; CI 95%: 0.22–0.92).

Conclusions

This study challenges the classical view of adverse hemodynamic effects of retrograde VA-ECMO flow, and highlights flow response depends on a dynamic balance between biventricular pre and afterload. Individualized flow adjustments might prevent unnecessary interventions (such as systematic venting) and the use of CAP could potentially improve responder patient identification. 

Original Title: Effect of ECMO Flow Variations on Pulmonary Capillary Wedge Pressure in Patients With Cardiogenic Shock.

Reference: Saura, O, Hékimian, G, Del Marmol, G. et al. Effect of ECMO Flow Variations on Pulmonary Capillary Wedge Pressure in Patients With Cardiogenic Shock. JACC. 2025 Sep, 86 (11) 768–778. https://doi.org/10.1016/j.jacc.2025.06.048.