This study identified three hemodynamic phenotypes early post-VA ECMO in cardiogenic shock that were related to the underlying CS etiology and RV function.
Key Findings
Results
Animal models of cardiogenic shock treated with VA ECMO demonstrated three distinct hemodynamic phenotypes based on cluster analysis of LVEDP and arterial pulse pressure changes.
Study included 58 pigs with cardiogenic shock assessed at 2 levels of VA ECMO flows
Cluster 1 centroid showed LVEDP change of +4.99 mmHg and arterial PP change of -1.85 mmHg
Cluster 2 centroid showed LVEDP change of +0.02 mmHg and arterial PP change of -5.57 mmHg
Cluster 3 centroid showed LVEDP change of +1.31 mmHg and arterial PP change of +15.5 mmHg
These findings challenge simulation model predictions that VA ECMO uniformly results in LV distension in cardiogenic shock
Results
Clinical data from patients with cardiogenic shock on VA ECMO revealed three hemodynamic phenotypes characterized by differences in arterial pulse pressure, pulmonary artery pulse pressure, and pulmonary artery diastolic pressure.
Study included 128 patients with cardiogenic shock with data collected within 2 hours of VA ECMO initiation
The 'pulsatile' phenotype comprised 26% of patients and had the highest arterial PP
The 'low pulsatility' phenotype comprised 52% of patients and had low arterial PP and low PA pulse pressure
The 'LV distension' phenotype comprised 22% of patients and had elevated pulmonary artery diastolic pressure
PADP, PAPP, and arterial PP were the key variables used to distinguish clinical phenotypes
Results
Acute myocardial infarction as the etiology of cardiogenic shock was more common in the 'LV distension' phenotype.
The 'LV distension' phenotype was defined by elevated pulmonary artery diastolic pressure post-VA ECMO
This phenotype represented 22% of the 128-patient clinical cohort
The association suggests that acute MI-related cardiogenic shock may predispose patients to LV distension following VA ECMO initiation
This finding supports the clinical concern about LV unloading needs in acute MI patients placed on VA ECMO
Results
The 'low pulsatility' phenotype was associated with larger right ventricular diameter and tricuspid regurgitation.
The 'low pulsatility' phenotype was the most common, comprising 52% of clinical patients
This phenotype was characterized by low arterial PP and low PA pulse pressure
Larger RV diameter and tricuspid regurgitation distinguished this phenotype from the others
These findings suggest that RV dysfunction or failure may be a dominant feature in this phenotype of cardiogenic shock patients on VA ECMO
Discussion
Simulation models predicting uniform LV distension with VA ECMO in cardiogenic shock were not supported by either animal or clinical data.
Only the 'LV distension' phenotype (22% of clinical patients) and Cluster 1 in animals showed evidence of increased LV filling pressures
The majority of patients fell into phenotypes without LV distension as the dominant hemodynamic response
The animal study at 2 levels of VA ECMO flow corroborated the heterogeneity of hemodynamic responses seen clinically
The three phenotypes were related to underlying CS etiology and RV function rather than being a uniform response
Methods
Hemodynamic phenotyping of cardiogenic shock patients on VA ECMO can be performed using readily available clinical parameters within 2 hours of ECMO initiation.
Parameters used included pulmonary artery diastolic pressure (PADP), PA pulse pressure (PAPP), and arterial pulse pressure (PP)
Data collection occurred within 2 hours of VA ECMO initiation in the clinical cohort
Animal models used LV end-diastolic pressure and arterial pulse pressure as phenotyping variables
The identification of phenotypes early post-VA ECMO suggests potential for early clinical decision-making regarding LV unloading strategies
What This Means
This research suggests that when patients in cardiogenic shock (a condition where the heart cannot pump enough blood) are placed on a heart-lung bypass machine called VA ECMO, they do not all respond the same way. Previously, computer models had predicted that VA ECMO would consistently cause the left ventricle (the heart's main pumping chamber) to become overfilled and distended, which could be harmful. By studying both pigs with induced heart failure and 128 human patients, the researchers found three distinct patterns of how the heart and circulation respond to VA ECMO support.
In the clinical study, about 52% of patients fell into a 'low pulsatility' group characterized by weak pulse pressure and signs of right heart dysfunction, including an enlarged right ventricle and a leaky tricuspid valve. About 26% were in a 'pulsatile' group with stronger pulse pressure, and about 22% showed the 'LV distension' pattern — elevated pressures suggesting the left ventricle was being overfilled. Patients whose cardiogenic shock was caused by a heart attack were more likely to fall into this last group. These patterns were identifiable within just 2 hours of starting ECMO using standard monitoring measurements.
This research matters because it suggests that clinicians cannot assume all VA ECMO patients need the same interventions to protect the heart. Decisions about whether to add additional devices to unload (decompress) the left ventricle may need to be individualized based on which hemodynamic phenotype a patient belongs to. The findings also highlight that right ventricular (right heart) dysfunction appears to be a major feature in the majority of these patients, which may have important implications for how cardiogenic shock patients on VA ECMO are monitored and managed.