Factors influencing the functional status of aortic valve in ovine models supported by continuous-flow left ventricular assist device

Abstract

Objectives: An acute animal experiment was performed to observe factors influencing the functional status of the aortic valve functional status after continuous-flow left ventricular assist device (CF-LVAD) implantation in an ovine model, and a physiologic predictive model was established.

Methods: A CF-LVAD model was established in Small Tail Han sheep. The initial heart rate (HR) was set to 60 beats/min, and grouping was performed at an interval of 20 beats/min. In all groups, the pump speed was started from 2000 rpm and was gradually increased by 50-100 rpm. A multi-channel physiological recorder recorded the HR, aortic pressure, central venous pressure, and left ventricular systolic pressure (LVSP). A double-channel ultrasonic flowmeter was used to obtain real-time artificial vascular blood flow (ABF). A color Doppler ultrasound device was applied to assess the aortic valve functional status. Multivariate dichotomous logistic regression was used to screen significant variables for predicting the functional status of the aortic valve.

Results: Observational studies showed that ABF and the risk of aortic valve closure (AVC) were positively correlated with pump speed at the same HR. Meanwhile, the mean arterial pressure (MAP) was unaltered or slightly increased with increased pump speed. When the pump speed was constant, an increase in HR was associated with a decrease in the size of the aortic valve opening. This phenomenon was accompanied by an initial transient increase in the ABF and MAP, which subsequently decreased. Statistical analysis showed that the AVC was associated with increased pump speed (OR = 1.02, 95% CI = 1.01-1.04, p = 0.001), decreased LVSP (OR = 0.95, 95% CI = 0.91-0.98, p = 0.003), and decreased pulse pressure (OR = 0.82, 95% CI = 0.68-0.96, p = 0.026). ABF or MAP was negatively associated with the risk of AVC (OR < 1). The prediction model of AVC after CF-LVAD implantation exhibited good differentiation (AUC = 0.973, 95% CI = 0.978-0.995) and calibration performance (Hosmer-Lemeshow χ² = 9.834, p = 0.277 > 0.05).

Conclusions: The pump speed, LVSP, ABF, MAP, and pulse pressure are significant predictors of the risk of AVC. Predictive models built from these predictors yielded good performance in differentiating aortic valve opening and closure after CF-LVAD implantation.

Keywords: LVAD; aortic valve; influencing factors; nomogram; predictive model.