The Guytonian model to identify right heart failure in patients with cardiogenic shock on Impella 5.5

Background

The cardiovascular system can be described by the cardiac function curve (CFC) and the venous return curve (VRC). The point where the CFC intersects with the VRC is the operating point (OP), with coordinates of right atrial pressure (RAP) and cardiac output (CO). Physiological insights from this ‘Guytonian’ model could guide clinical management of cardiogenic shock (CS).

Purpose

To test the hypotheses that in patients with heart failure-related CS, the position of the OP can differentiate patients with vs without right heart failure (RHF) on Impella 5.5 support.

Methods

The CFC can be described by a logistic curve (CO = Cmax/(1+e-alpha(RAP-RAP0)), where Cmax is the maximum CO. The analogue of mean systemic filling pressure (Pmsa) was used to generate the VRC (resistance in the venous compartment of 0.068, 0.049 and 0.025 were used for low, normal and high SVR states). The slope was computed as the first derivative of the logistic curve. The plateau region was defined as the portion of the curve where the derivative falls below 10% of its maximum slope at the inflection point. Cardiac power output index (CPOI=(MAP-RAP)xCI/451), CI=cardiac index, MAP=mean arterial pressure.

Patients with heart failure-related CS who were hypotensive post-Impella 5.5 were included. Two fluid boluses (500mls Hartmann’s each) were administered in succession followed by up-titration of norepinephrine. Hemodynamic data before and immediately after each fluid bolus were used to generate the curves (i.e. three sets of RAP, CO and MAP data). Delta CPOI was used to measure response to norepinephrine.

Results

Seventeen patients with HF-related CS and Impella 5.5 support were included. Left ventricular ejection fraction 10 (10-15), vasoactive inotrope score was 15.0 (11.7-17.8), lactate was 3.9±0.6. Fluid boluses increased RAP 5.6±2.4 to 10.8±4.1 mmHg and CO 3.77 (3.62-4.10) to 4.86 (4.10-5.11) L/min (p<0.05). Fluid boluses produced near-parallel rightward shift of the VRC due to increase in Pmsa (10.7±2.2, 13.1±3.1, 15.8±3.7 mmHg, p<0.05). Curve-fitting yielded R2 ‘goodness-of-fit’ >0.95 in all cases.

Five patients had RHF and underwent percutaneous right ventricular assist device (RVAD) support. In patients with RHF, baseline RAP was higher (7.6±1.2 vs 4.7±2.1, p=0.012) but CO were comparable (3.7±0.1 vs 3.9±0.3, p=0.223). Patients with RHF had lower Cmax (4.1±0.2 vs 5.3±0.3 L/min) and operated close to Cmax with distance to plateau of -5.26 (-6.34- -4.75) vs 1.85 (1.26-2.34) mmHg (all p<0.05). Norepinephrine increased from 0.05 (0.02-0.06) to 0.07 (0.04-0.10) mcg/kg/min, but only increased CPOI in patients without RHF (0.37 (0.32-0.39) to 0.40 (0.37-0.43) W/m2, p=0.001 vs 0.24 (0.24-0.26 to 0.24 (0.23-0.25) W/m2, p=0.125).

Conclusion

The position of the OP in the ‘Guytonian’ model following fluid boluses could identify patients with heart failure-related CS and RHF on Impella 5.5, and guide titration of vasopressor and RVAD support.Impella 5.5 and RHF (RVAD)For image description, please refer to the figure legend and surrounding text.Impella 5.5 and no RHFFor image description, please refer to the figure legend and surrounding text.