Chronic Management of Patients with Left Ventricular Assist Devices

Fig. 10.1
HeartMate II left ventricular assist device . CT scan film demonstrating appearance of the device on radiography, which is important to recognize in patients with long-term devices in place. The inflow cannula is seen entering the left ventricle, and the pump is positioned outside the pericardial space


Fig. 10.2
HeartWare HVAD left ventricular assist device . CT scan demonstrates the appearance of the device. The pump is implanted in the pericardial space


Fig. 10.3
HeartMate III left ventricular assist device . CT scan images shown. The device is implanted in the pericardial space. Compared to the HVAD, it is larger with a shorter inflow cannula

The HeartMate II device (Thoratec, Pleasanton, CA) was initially approved for use as a bridge-to-transplant (BTT) therapy in 2008 and then as a destination therapy (DT) for patients not eligible for transplant in 2010. HeartWare became approved as a BTT device in 2012 and is currently completing trials to obtain DT approval. Increasing use of LVADs for DT has been seen in the most recent Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS ) data , which recorded an increase in patients receiving DT devices from 28.3% in the 2008–2011 era to 45.7% in 2014. However, approximately 60% of patients still receive an LVAD as a BTT therapy with immediate listing for heart transplantation or plans to place on the list in the near future [2].

Those receiving a device as DT have been demonstrated to have improved survivals compared to those end-stage heart failure patients who have no intervention. For DT patients, after implantation of an LVAD, there is estimated survival of 76% and 57% at 1 year and 3 years, respectively. Technical improvements in continuous-flow LVADs have improved long-term survival in patients with end-stage heart failure who receive a device for a BTT or DT indication. According to the most recent estimates, combined survival at 2 years with either an axial or centrifugal LVAD is 83%. Despite improvements in devices, challenges continue to emerge in the management of these patients.

Device Management

General Management

Current generation devices utilize a single part rotor in either an axial-flow rotor or centrifugal design to counter earlier generation complications of mechanical failure, but the continuous flow of these devices changes the nature of monitoring this subset of patients. While speed is the main parameter adjusted in an LVAD patient based on their hemodynamic status, there should also be continued optimization of the CHF regimen, mean arterial pressure, and possibly newly initiated anticoagulation regimen. This should include maintenance of optimally tolerated neurohormonal blockade, volume optimization, and afterload reduction.

After implantation of an LVAD, monitoring of the device parameters and alarms can provide insight into the function and hemodynamic status of the patient. HeartMate II device parameters include speed (RPMs), power (Watts), pulsatility index (PI), and estimated flow (L/min). These parameters should be measured and recorded at each visit to assess for trends or acute changes in LVAD function. Additionally, review of any recent alarms and the timing of those alarms is an important evaluation of an LVAD patient. The specific interpretation of alarms will be reviewed later in this chapter.

Generally flow and power carry a linear relationship compared at a given speed. However, this may not hold true in certain clinical situations. Given the linear relationship between flow and power, flow is a calculated number based on a direct measurement of the power. Therefore, if the power increases due to mechanical failure, it may not truly reflect increased flow from the LVAD. For example, when a thrombus is present in the inflow cannula, it can produce an increase in power without increased flow due to the obstruction by the thrombus. This may cause some confusion related to reported device readings as both power and flow will be incorrectly reported as increased. Similarly, if there was outflow obstruction, this would cause a decrease in flow as well as erroneously low power.

The PI is a measure of the assistance the LVAD is providing to the LV and is provided in a range from 1 to 10. The PI is a specific parameter for only the HeartMate devices. This PI is calculated based on flow pulses during systole, which are sensed by the LVAD and averaged over duration of 15 s. Lower values indicate the pump is contributing more to systemic flow and will translate to less pulsatility, whereas higher values indicate less pump contribution to this flow and higher pulsatility. PI values should remain relatively constant, and a decrease should lead clinicians to consider a decrease in circulating blood volume. Significant increases in PI should prompt further evaluation for possible fluid retention, hyperdynamic states (such as sepsis), significant aortic valvular insufficiency, and, in rare instances, cardiac recovery.

Additionally, after implantation, the necessary speed of the LVAD should be evaluated at each visit based on mean arterial pressure, echocardiography, and, if necessary, further testing such as a right heart catheterization (RHC) . The speed should be aimed to maintain peripheral pressure and perfusion while minimizing right ventricular (RV) overload, aortic regurgitation (AI), or left ventricular (LV) collapse [3, 4]. Since RV function, AI, and LV size are dynamic factors, frequent assessment should occur at least every 6 months following implantation of an LVAD or sooner if symptoms arise.

Role of Hemodynamic Assessment

A reduction in exertional capacity is a characteristic feature of advanced heart failure. The use of cardiopulmonary exercise testing (CPET) in patients with severe left ventricular impairment prior to cardiac transplantation or LVAD placement will demonstrate a markedly reduced exercise time and peak oxygen consumption. Patients with marked reduction of VO2 below 14 mL/kg/min have been demonstrated to have a reduced survival when compared to those with left ventricular impairment and a VO2 greater than 14 mL/kg/min [5]. While completion of CPET is typically completed prior to cardiac transplantation, its use is also recommended in patients after implantation of an LVAD. It may be useful in providing clinicians an objective assessment of exercise capacity as well as helping guide recommendations for exercise regimens. Studies have consistently demonstrated improvement in exercise time in the first 6–8 weeks following LVAD implantation; however, there has been variable response to VO2. Age may be a large predictor of VO2 improvement with patients reaching only 50–60% of their age- and sex-predicted VO2 after LVAD implantation [6, 7].

Current guidelines recommend regular interval testing of hemodynamics by right heart catheterization, particularly in those patients awaiting transplantation. Serial evaluation can identify those patients with pulmonary hypertension (pHTN), which when irreversible has been associated with a higher risk of allograft dysfunction [810]. No specific recommendations currently exist on the interval of testing; however, data has suggested that those with pHTN on right heart catheterization 2–3 months after LVAD implantation are at the greatest risk for persistent or progressive pHTN in the following 6 months. In addition, there appears to be little predictive value of a right heart catheterization prior to LVAD implantation and development of pHTN following surgical implantation .

Right heart catheterization has also proven useful for hemodynamic-guided LVAD optimization and diagnosis of inadequate pump speed, right heart failure, or volume overload. Studies have demonstrated the ability to decrease pulmonary capillary wedge pressure and central venous pressures as well as increase cardiac output/index by evaluating increasing pump speeds while in the catheterization lab [11]. Interval evaluation may also provide a way to differentiate patients with persistent heart failure symptoms into categories of right heart failure (relatively normal PCWP with persistently elevated central venous pressures) versus those with left-sided volume overload (persistently elevated PCWP despite increasing pump speeds) (Fig. 10.4). Also some patients may respond to increasing pump speeds with decreasing PCWP and central venous pressures suggesting inadequate LVAD speeds. While hemodynamic improvements have been achieved with right heart catheterization-directed studies, no direct correlations have been drawn to symptomatic improvement or decreases in morbidity and mortality .


Fig. 10.4
Interpretation of right heart catheterization in LVAD patients. CVP central venous pressure; PCWP pulmonary capillary wedge pressure; RHF right heart failure


This section will identify the basics of echocardiography in the chronic management of patients following implantation of an LVAD , while prior chapters will address preoperative assessment and issues immediately postoperatively. Transthoracic echocardiography (TTE) is essential in the optimization of LV decompression, reduction of aortic insufficiency, and evaluation of possible device malfunction. Using standard echocardiographic views, 2D measurements, color Doppler, and spectral Doppler can collectively be used to provide LV size, valvular function, and interrogation of inflow/outflow cannulae. Guidelines by the American Society of Echocardiography recommend surveillance TTE at postoperative week 2 and at 1, 3, 6, and 12 months if the patient remains clinically stable. Evaluation should then take place every 6–12 months thereafter [12].

Measurement of the left ventricular internal diastolic dimension (LVIDd ) is an important parameter in the assessment of LV unloading following LVAD implantation. While end-diastolic volumes as obtained by Simpson’s method of disks have in some cases been shown to be a more accurate measurement of LV unloading, reproducibility can be an issue due to shadowing from the outflow cannula of the LVAD. Therefore, the use of the parasternal long-axis images to obtain the LVIDd has been found to be the most reproducible measurement. Following LVAD implantation, approximately a 15% reduction in LV size has been shown to be expected at 3 months post-LVAD [13, 14]. This data was obtained solely in HM II patients; therefore variation may exist between devices.

In regard to LV ejection fraction, serial measurement can provide clinicians data to evaluate for myocardial recovery or worsening over time. As mentioned above, measurement of LVEF by Simpson’s method of disks is the recommended method; however, it can be difficult due to shadowing in the apex from the LV outflow cannula, paradoxical septal motion, or significant regional wall motion abnormalities. Therefore, in cases when endocardial visualization is difficult, the use of alternative methods including LV fractional shortening, LV fractional area change, or the Quinones method has been suggested (for the latter bearing in mind the LV apex should be considered akinetic due to the presence of the LV inflow cannula) (Table 10.1). It is important to remember that these methods are not validated in patients with LVAD in situ.

Table 10.1
Alternative methods of evaluation of LVEF in LVAD patients





Fractional area change (%)

FAC = [(end-diastolic area − end-systolic area)/(end-diastolic area)]

Can be used in the absence of adequate apical visualization

Less reliable in the setting of significant LV wall motion or distortion (i.e., aneurysm)

Fractional shortening

FS = [(LVIDd − LVIDs)/(LVIDd)]

Can be used in the absence of adequate apical visualization

Less reliable in the setting of significant wall motion or LV distortion (i.e., aneurysm)

Quinones method

LVEF = [(LVIDd2 − LVIDs2)/(LVIDd2)]

Decreases error by using multiple areas of measurement

Less reliable in the setting of significant wall motion or LV distortion (i.e., aneurysm)

FAC fractional area change; FS fractional shortening; LVIDd left ventricular internal diastolic dimension; LVIDs left ventricular internal systolic dimensions; LVIDd 2 indicates measurement of eight different LV dimensions at different levels in the parasternal long axis, four-chamber, and long-axis views of end diastole; LVIDs 2 indicates measurement of eight different LV dimensions at different levels in the parasternal long axis, four-chamber, and long-axis views of end systole

Assessment of the aortic valve mobility is an alternative parameter in serial evaluation of LV unloading. If LVAD speed is set above which allow aortic valve opening, it may lead to aortic regurgitation (AR) which impairs LVAD function. This occurs as a continuous loop of flow from the LV outflow cannula to the aorta followed by regurgitation back into the LV. Therefore, significant AR can affect the unloading of the LV and thus LVAD effectiveness. Development of AR has important implications in morbidity and mortality, which will be discussed in more detail later in this chapter. Additionally, if the aortic valve remains closed, it can predispose patients to aortic root thrombus and/or fusion of the aortic valve cusps) .

Assessment of the aortic valve during LVAD surveillance TTE should begin with evaluation of opening of the valve. This is most accurately achieved with the use of M-mode echocardiography by recording the aortic valve in up to 5–6 cardiac cycles, as the valve can open with every cardiac cycle, open intermittently, or remain closed with every cycle. While aortic valve opening can occur with each cardiac cycle, it may only occur for a short duration. Therefore, the duration of aortic valve opening should also be addressed by averaging the duration of opening in multiple cardiac cycles, usually in milliseconds (ms).

Parameters for defining the severity of AR have not been specifically validated; however, the use of prior guidelines is generally followed. When there is a vena contracta of ≥0.3 cm or an AR jet width of >46% of the LVOT diameter, there is likely at least moderate if not severe AR. The AR may also be present during only diastole, nearly continuous when it extends into systole, or continuous when it is holodiastolic and holosystolic. Due to the ability of AR to be present into systole and the extracardiac circuit of the LVAD, neither the pressure halftime nor the presence/absence of aortic diastolic flow reversal is a reliable method to quantitate AR.

Mitral regurgitation (MR) is also regularly evaluated on surveillance TTE after LVAD implantation as it can have implications for device management. Quantification of the severity of MR can be made based on the general echocardiography guidelines. Importantly, the presence and severity of MR can be an indicator of adequate unloading provided by an LVAD. Appropriate LVAD speeds will ideally lead to reduction in LV size and in turn the mitral annulus. This reduction in the mitral annular size will improve coaptation and thus reduce mitral regurgitation. However, if mitral regurgitation is persistent despite increasing LV unloading, evaluation for LVAD malfunction should be sought as the outflow cannula may interfere with the submitral apparatus in some cases .

The tricuspid and pulmonic valves are reliably interrogated using standard methods in patients following LVAD implantation. Tricuspid regurgitation (TR) , present in moderate to severe ranges, can provide indirect data on the function of the LVAD. Assuming the absence of a concurrent right ventricular assist device, significant TR in appropriate clinical scenarios can suggest inadequate LV unloading, RV dysfunction, or excessive LV unloading leading to intraventricular shift and distortion of the tricuspid valve morphology. The presence of significant TR should therefore prompt review of serial changes in LV size, ejection fraction, measures of RV function, and intraventricular motion .

Lastly, TTE can be useful in the evaluation of the inflow and outflow cannula of the LVAD. The inflow may be visualized in the parasternal long axis or LV four-chamber views and should be evaluated for its position in reference to the septum or submitral apparatus. Flow through the cannula can also be interrogated using pulsed and continuous wave Doppler and should be obtained over 3–4 cardiac cycles. Normal Doppler waveforms will be pulsatile due to the contribution of the LV to flow even if the aortic valve is closed (Fig. 10.5). Doppler velocities should also be ≤1.5 m/s, and when higher flows are present, it may indicate obstruction or the presence of thrombus.


Fig. 10.5
Pulsed wave Doppler is seen of a HeartMate II inflow cannula. Although it is a continuous-flow device, contribution of left ventricular contraction leads to a systolic peak (lined arrow) and a diastolic nadir (arrow head). Typically these can be obtained in a standard four-chamber echo view or from the parasternal long view (upper left panel). Velocities should be ≤1.5 m/s

Evaluation of the aortic outflow graft anastomosis is more difficult but can be seen in the modified parasternal views, which focus on the ascending aorta. In cases where this is not sufficient, positioning patients in the right lateral decubitus position and obtaining right parasternal views may be helpful. Spectral velocities through the graft can be used in calculation of flow using velocity time integral (VTI) and outflow graft area method, keeping in mind that velocities can vary between graft sizes. For example, HeartMate II tends to have a larger outflow graft (16 mm) than HeartWare devices (10 mm). In general, flows >2 m/s are considered abnormal for the outflow graft, and further evaluation for obstruction should be undertaken.

Optimization of a patient’s LVAD speed can be completed in asymptomatic ambulatory patients using TTE to obtain the above parameters. These optimization studies, sometimes referred to as a “speed change” echo, can be obtained by completing a TTE study at the baseline speed and either lower or incrementally increasing speeds. At each speed, the patient’s mean pressure as well as echo parameters should be obtained, including LVIDd, septal position, frequency/duration of aortic valve opening, and quantitative or qualitative AR, MR, and TR. It should be noted that if a thrombus is visualized in the aortic root, the LVAD speed should not be changed as it can aid in mobilization of the thrombus, especially at lower speeds.

HeartMate II and HeartMate III Device Parameters

After implantation of a HeartMate II or HeartMate III device, patients will receive a system controller with two sets of batteries, a primary operating set, and a backup. Additionally, the patients will have a system controller and a power base unit (PBU). A control monitor is required at the initial implantation and subsequent encounters to review settings and alarms but is not a required component for discharge. Patients will be educated during their admission on the operation of the system controller, which has several display icons and buttons seen on the face (Fig. 10.6). There are two buttons, which include the “Test Select” and “Silence Alarm” buttons, which allow patients to interact with the system. Lighted icons on the controller include a power symbol, battery symbol with fuel gauge, battery module symbol, and red heart symbol.


Fig. 10.6
HeartMate II control system

The battery symbol and fuel gauge are the most vital for day-to-day patient usage. The gauge has four green markers, which can provide an approximate amount of battery life remaining. When all four are lighted, it signifies that 75–100% of the battery energy remains with a reduction of 25% of battery energy with the loss of each marker. When a single green lighted marker remains, it signifies <25% of battery energy remains, and once the battery symbol appears as a yellow or red indicator, it signifies <15 min or <5 min of battery energy remains. Not all patients can be provided with similar durations of battery life as this can vary depending on the set speed or the age of the battery. Higher set speeds will deplete a battery more quickly, and as the battery ages, it will hold its charge for less time.

As previously described, the LVAD speed, flow, power, and pulsatility index are displayed when the device is hooked to the control monitor and can be adjusted by the clinician through this interface (Fig. 10.7). The minimum and maximum operating speeds for the HeartMate devices are 6000 and 15,000 RPMs, but typical operating speeds usually range between 8800 and 10,000 RPMs. When making speed adjustments or speed optimization on TTE, incremental changes of 200–400 RPMs are used .
Feb 24, 2018 | Posted by in CARDIOLOGY | Comments Off on Chronic Management of Patients with Left Ventricular Assist Devices
Premium Wordpress Themes by UFO Themes