Pulmonary Critical Care Fellow Houston, Texas, United States
Disclosure(s):
Moe Ameri, MD, M.Sc: No financial relationships to disclose
Background: Weaning from venoarterial extracorporeal membrane oxygenation in cardiogenic shock is limited by hemodynamic surrogates that may miss evolving systemic venous congestion. Hepatic and portal vein Doppler (venous excess ultrasound framework) may provide real-time physiologic congestion assessment, but its role during weaning is not well described.
Methods: We report two cardiac intensive care unit cases supported with venoarterial extracorporeal membrane oxygenation for cardiogenic shock. Each underwent a structured bedside weaning trial with stepwise reduction in circuit flow. During each step, we recorded mean arterial pressure and standard invasive hemodynamics (including central venous and pulmonary artery pressures) alongside echocardiography. Simultaneously, we obtained hepatic vein Doppler (assessing direction and phasicity of venous flow) and portal vein Doppler (assessing degree of pulsatility as a marker of systemic venous congestion). Doppler findings were interpreted in real time and integrated with end-organ perfusion markers to support multidisciplinary decisions regarding decannulation readiness.
Outcome: Case 1: A 74-year-old woman developed cardiogenic shock and cardiac arrest after high-risk coronary intervention and was supported with venoarterial extracorporeal membrane oxygenation plus intra-aortic balloon pump. On day 3, flows were reduced stepwise. Mean arterial pressure remained adequate, and right-sided filling pressures did not meaningfully rise. Hepatic vein Doppler remained predominantly antegrade, and portal vein pulsatility stayed low, suggesting minimal venous congestion despite reduced support. With improving end-organ function, she was decannulated on day 4 and discharged to rehabilitation without recurrent shock. Case 2: A 67-year-old woman developed postoperative cardiogenic shock after coronary bypass, mitral valve repair, and left ventricular aneurysm repair, requiring venoarterial extracorporeal membrane oxygenation with vasoactive support and intra-aortic balloon pump. On day 6, flow reduction from 3.5 to 1.0 L/min led to a gradual mean arterial pressure decline (75 to 60 mm Hg) while invasive pressures remained relatively stable. Portal vein pulsatility increased from minimal to moderate and then moderate-to-severe, indicating worsening venous congestion at lower flows; decannulation was deferred.
Conclusion: Hepatic and portal vein Doppler provided real-time, physiologic insight into systemic venous congestion during VA-ECMO weaning. Stable waveforms supported safe decannulation, while rising portal pulsatility identified weaning intolerance despite reassuring invasive pressures. Integrating venous Doppler into structured weaning may improve decision-making and avoid premature decannulation.
