Computationally Managed Bradycardia Improved Cardiac Energetics While Restoring Normal Hemodynamics in Heart Failure

Abstract

In acute heart failure, systemic arterial pressure (AP), cardiac output (CO), and left atrial pressure (P _LA) have to be controlled within acceptable ranges. Under this condition, cardiac energetic efficiency should also be improved. Theoretically, if heart rate (HR) is reduced while AP, CO, and P _LA are maintained by preserving the functional slope of left ventricular (LV) Starling’s curve (S _L) with precisely increased LV end-systolic elastance (E _es), it is possible to improve cardiac energetic efficiency and reduce LV oxygen consumption per minute (MVO ₂). We investigated whether this hemodynamics can be accomplished in acute heart failure using an automated hemodynamic regulator that we developed previously. In seven anesthetized dogs with acute heart failure (CO < 70 mL min⁻¹ kg⁻¹, P _LA > 15 mmHg), the regulator simultaneously controlled S _L with dobutamine, systemic vascular resistance with nitroprusside and stressed blood volume with dextran or furosemide, thereby controlling AP, CO, and P _LA. Normal hemodynamics were restored and maintained (CO; 88 ± 3 mL min⁻¹ kg⁻¹, P _LA; 10.9 ± 0.4 mmHg), even when zatebradine significantly reduced HR (−27 ± 3%). Following HR reduction, E _es increased (+34 ± 14%), LV mechanical efficiency (stroke work/oxygen consumption) increased (+22 ± 6%), and MVO ₂ decreased (−17 ± 4%) significantly. In conclusion, in a canine acute heart failure model, computationally managed bradycardia improved cardiac energetic efficiency while restoring normal hemodynamic conditions.

Appendix

Feedback Control Algorithms of the Automatic Hemodynamic Regulator

To minimize the difference between target and subject’s S _L (ΔS _L = target S _L—subject’s S _L) and those of R (ΔR = target R—subject’s R), the proportional-integral (PI) feedback controllers adjust the infusion rates of DOB and SNP, respectively (Fig. 2). In the PI controller (Fig. 6), ΔS _L (or ΔR) and the difference integrated with an integral gain (K _i) are summed and scaled by a proportional gain (K _p) to give the infusion rate of DOB (or SNP). PI gain constants for DOB infusion [K _i = 0.01 s⁻¹, K _p = 0.06 μg kg⁻¹ min⁻¹ (mL min⁻¹ kg⁻¹)⁻¹] and for SNP infusion [K _i = 0.007 s⁻¹, K _p = −1.37 μg kg⁻¹ min⁻¹ (mmHg min ml⁻¹ kg)⁻¹] were determined on the basis of open-loop response of S _L and R to the infusion of DOB and SNP, respectively.30

Figure 6

Block diagram of the PI controller in the automated hemodynamic regulator. ΔS _L and ΔR denote the difference between target and subject’s S _L, and between target and subject’s R, respectively. K _i and K _p represent the integral and proportional gain constants, respectively. s is a Laplace operator

To minimize the difference between target and subject’s V (ΔV = target V—subject’s V), a nonlinear (N-L) feedback controller (Fig. 2) adjusts the infusion of DEX or injection of FUR based on the following “if-then” rules:

• Rule 1: If ΔV ≥ 1 mL kg⁻¹ then infuse DEX at 10 mL min⁻¹

• Rule 2: If ΔV ≤ −2 mL kg⁻¹ then inject FUR (10 mg) at 10 min intervals

The “if-then” rules were determined on the basis of the open-loop response of V to the infusion of DEX and FUR.30