Cardiac Output in Heart Failure Patients With Mechanical Pumps

Heart failure is a devastating disease, which affects approximately 6 million individuals in the USA alone. Five-year survival among all-comers with HFrEF is ~50%, and there are associated reductions in functional capacity and health-related quality-of-life (HRqOL). While guideline-directed medical therapy such as beta-blockers, angiotensinogen converting enzyme inhibitors, mineralocorticoid receptor antagonists and angiotensin receptor-neprilysin inhibitors improve survival, HRqOL and functional capacity, many patients progress to end-stage, advanced HFrEF, which has a five-year survival of only to 20% and is associated with reductions in functional capacity and HRqOL. While orthotopic heart transplantation is the gold-standard treatment for advanced HFrEF, only approximately 3,000 donor organs are available per year in the USA, and approximately 5,000 per year worldwide. As such, demand outweighs supply since estimates are that at least 50,000 individuals worldwide are candidates for transplant or suffer from severe HFrEF and are transplant ineligible for a variety of reasons (e.g. uncontrolled diabetes mellitus or peripheral vascular disease, or diagnosis of cancer within the past 5 years). To remedy this disparity, LVADs have emerged as an attractive alternative and are used as either a bridge-to-transplantation to stabilize and support patients until transplantation is possible, or as destination-therapy for transplant-ineligible patients. HFrEF patients frequently suffer from elevated cardiac filling pressures and reduced cardiac output (Qc) under resting conditions. LVADs normalize these hemodynamic abnormalities - at least under resting conditions - through a rotating impeller that propels blood from the left ventricle (LV) into the ascending aorta (figure 1). Total Qc is therefore determined by: 1) the LVAD, which provides the bulk of flow to the body during resting conditions; and 2) the native heart, which can still provide blood to the body by expulsion of blood through the aortic valve as the heart contracts.

Heart failure patients suffer from persistent heart failure symptoms, and impairments in functional capacity, following LVAD implantation. It has previously been demonstrated that LVAD patients have severely reduced functional capacity, as measured by serial assessments of maximal oxygen uptake (VO2Max), greater than one year following device implantation.27 Specifically, VO2Max, when assessed before LVAD implantation, 3-6 months, 1 year, and greater than 1 year following device implantation, remains below 14ml/kg/min. Submaximal exercise, measured by six-minute hall walk (6MHW) scores, improves modestly following device implantation, but on average, remains severely reduced at approximately 300-350m, and similar to VO2Max values observed in these patients, falls within a range that is observed among patients with advanced HFrEF. About half of patients report persistence of HFrEF-related symptoms 6-12 months following device implantation. These persistent reductions in VO2Max, 6MHW, and HRqOL assessments all indicate that LVAD patients suffer from persistent heart failure, which manifests with attempts to exercise. For example, HFrEF patients are considered to be eligible for a heart transplantation when their VO2Max falls below 12-14ml/kg/min.

Impairments in HRqOL and functional capacity result from an inability of the LVAD to improve cardiovascular hemodynamics during exercise. To understand how LVADs influence cardiac filling pressures at rest and during exercise, we previously evaluated patients prior to, and following LVAD implantation by invasive cardiopulmonary exercise testing (CPET) with Swan-Ganz catheterization during upright cycle ergometry prior to and following LVAD implantation (COMIRB #16-1635). In this study (figure 2), three visits were completed: 1) Visit 1: baseline exercise assessment four weeks before LVAD implantation. 2) Visit 2: post-LVAD exercise assessment with patient exercising at constant LVAD pump speed. 3) Visit 3: post-LVAD exercise assessment but with stepwise increases in LVAD rotor speed during exercise, to determine whether additional flow through device improves exercise capacity. Several novel insights resulted from this work (figure 2):

1. Cardiopulmonary performance remains severely limited following device implantation. Compared to pre-LVAD CPET (visit 1), there is no improvement in VO2Max following LVAD implantation when patients exercise to volitional exhaustion at either a constant pump speed (visit 2), or with stepwise increases in pump speed (visit 3).
2. LVAD flow increases minimally during exercise. Regardless of whether LVAD patients exercise at a constant pump speed or with stepwise increases in speed, LVADs can only increase flow by approximately 1L/min above resting levels. As such, cardiac output during exercise - and hence, VO2Max - is determined by contractile reserve of the native ventricle, as opposed to the LVAD itself.
3. Exertional pulmonary arterial and cardiac filling pressures do not improve following LVAD implantation. Pre-implantation exertional filling pressures (visit 1) are severely elevated in the setting of advanced HFrEF, and interestingly, there is no substantive improvement following device implantation when exercising at a fixed LVAD speed or with stepwise increases in pump speed.

There is a paucity of data regarding the impact of changes in hemodynamics on LVAD function in the human body. The effects of alterations in afterload, preload and contractility in the normal heart are well described. For example, left ventricular (LV) pressure-volume analysis (the gold-standard metric of describing ventricular function) indicates that there is an inverse relationship between afterload and Qc, such that increases in afterload lead to reductions in Qc, and vice versa. However, preload is directly related to Qc, such that increases in cardiac preload lead to a rise in Qc through the Frank-Starling mechanism. In vitro studies of LVADs suggest that - at least in the controlled environment of a "mock-loop" (figure 3), LVADs have a reduced preload sensitivity than the normal heart. For example, LVAD preload sensitivity is approximately half the levels observed in the normal heart (LVAD v. heart, 0.105±0.092 v. 0.213±0.003 L/min/mmHg). However, similar mock-loop studies suggest that LVADs have a much higher afterload sensitivity - approximately three times - that of the normal heart (LVAD v. heart 0.09±0.034 v. 0.03±0.01 L/min/mmHg). The main limitation with these mock-loop studies, however, is that for variables such as preload and afterload, which all contribute to changes in flow - these variables are changed in isolation (e.g. increasing preload while holding afterload constant), whereas in the human body, all variables change simultaneously during activity. As such, these mock-loop studies do not adequately describe - or explain, LVAD flow behavior during exercise, where preload increases, afterload decreases, and contractility increases, but LVAD flow increases minimally or not at all (point #2 above). It has previously been emphasized that a detailed understanding of exercise physiology in this patient population is necessary to improve HRqOL and exercise tolerance in this patient population. Therefore, the primary objective of this study is to evaluate LVAD performance in response to alterations in loading conditions (preload, afterload, contractility) in HFrEF patients supported by these devices, and characterize the determinants of LVAD flow and Qc during activity/exercise.