12 Echocardiographic Evaluation of Ventricular Support Devices
Background
• Ventricular assist devices (VADs) were initially used to stabilize patients with acute cardiogenic shock (particularly post–cardiac surgery).
• Subsequently, they have been used as a “bridge to transplant,” “bridge to recovery,” or “bridge to decision” (stabilizing patients during work-up to decide transplantation candidacy) and as “destination therapy” (lifelong use).
• Early-generation left ventricular assist devices (LVADs) mimic the pulsatile action of the heart and have inlet and outlet valves (Figure 12-1A; Table 12-1).
• Automatic mode on pulsatile pumps varies heart rate to maintain constant volume; therefore, pumping is asynchronous with the electrocardiogram (ECG).
• More recent LVADs are continuous flow, involving axial (rotating propeller) pumps, and have no valves (making them more durable) (see Figure 12-1B; see Table 12-1).
• In continuous-flow LVADs, rotor speed (in rotations per minute, rpm) and pressure difference between inflow chamber and outflow chamber (usually left ventricle and aorta, respectively) determine flow.
• Even in continuous-flow LVADs, pulsatility is present as a result of residual ventricular function and the fact that LVAD flow is higher when the aorta-LV pressure differential is low (systole) and lower when the differential is higher (diastole).
• Third-generation VADs are magnetically or hydrodynamically levitated, continuous-flow centrifugal (rotating plate) pumps with no bearings or valves, theoretically making them even more durable than second-generation VADs (see Table 12-1).
• The dysfunctional RV can respond well to LVAD implantation because of reduced afterload from lower pulmonary pressures.
• Alternatively, the RV function may worsen because of high preload from increased venous return secondary to increased left-sided output.
• Some patients receiving LVADs will require RV mechanical support, either temporarily (separate right ventricular assist device [RVAD]) or permanently (biventricular assist device [BiVAD]).
• Other patients with RV dysfunction can be supported medically with pulmonary vasodilators and inotropes.
• RVADs generally involve an inflow cannula placed in the right atrium and an outflow cannula placed in the proximal pulmonary artery.
• The total artificial heart is an alternative to VADs as a bridge to transplantation and consists of right and left pumping chambers attached to the native atria.
• The intra-aortic balloon pump (IABP) reduces systolic afterload on the left ventricle and can improve diastolic coronary flow via counterpulsation in the descending aorta.
• Percutaneous ventricular assist devices (PVADs) involve cannulae inserted via peripheral arteries and veins.
• PVADs can provide 2 to 6 L/min of cardiac output support and can be placed quickly in acute cardiogenic shock from (usually left but also right) ventricular decompensation and/or acute severe mitral regurgitation.
• PVADs are used primarily as a bridge to recovery and “bridge to bridge” (acute stabilization with subsequent implantation of a standard LVAD).
• PVADs are also used in high-risk percutaneous coronary interventions in patients with severely reduced LV ejection fraction (LVEF), and in percutaneous aortic valve procedures.
• PVADs provide better hemodynamic support than IABPs, but studies have not demonstrated improvements in mortality.
• The Impella Recover (Abiomed, Danvers, MA) can support the right or left ventricle and consists of a catheter in an inflow port and an outflow port that is advanced retrograde through the aorta or pulmonary artery into the ventricle (Figure 12-2A).
• The inflow port placed below the aortic or pulmonic valve draws blood from the ventricle and ejects it from the outflow port positioned above the valve.
• The TandemHeart (Cardiac Assist, Pittsburgh, PA) involves an inflow cannula placed through the femoral vein in the RA, across an interatrial septotomy, and into the left atrium. The outflow cannula is placed through the femoral artery into the abdominal aorta (see Figure 12-2B).
• Extracorporeal membrane oxygenation (ECMO) provides temporary cardiopulmonary support for patients with reversible cardiogenic shock and/or severe pulmonary failure.
• ECMO systems include an inflow cannula placed into the right atrium via the femoral vein or right internal jugular vein, and an outflow cannula in the femoral artery or ascending aorta.
| Examples | Type |
|---|---|
| First Generation | |
| HeartMate XVE (Thoratec Corp.) | Pulsatile |
| Novacor pumps (WorldHeart Corp.) | Pulsatile |
| Second Generation | |
| HeartMate II (Thoratec Corp.) | Axial Flow |
| Jarvik 2000 (Jarvik Heart Corp.) | Axial Flow |
| Micromed DeBakey (Micromed Cardiovascular Inc.) | Axial Flow |
| Third Generation | |
| HeartWare HVAD (HeartWare Inc.) | Centrifugal |
| Duraheart (Terumo Heart Inc.) | Centrifugal |
| Levacor (WorldHeart Corp.) | Centrifugal |
| Total Artificial Heart | |
| Abiocor (Abiomed Corp.) | |
| Cardiowest (Syncardia Inc.) |
Overview of Echocardiographic Approach (Table 12-2)
Echocardiography-Guided LVAD Optimization
• Optimizing LVAD pumping rates, pumping volumes, or rotor speeds uses echocardiographic measurements to balance proper preload and adequate decompression with excessive unloading leading to “suck-down” events (obstruction of the inflow cannula by trabeculation, chamber walls, papillary muscles, or chordal structures in the setting of an underfilled chamber).
Percutaneous VADs, IABP, and ECMO
• Echocardiography is useful in preimplantation assessment, placement, and follow-up of two of the currently available PVADs, IABPs, and ECMO.
Preimplantation Echocardiographic Assessment
Anatomic Imaging
Acquisition
• Standard echocardiographic imaging techniques are used to evaluate cardiac anatomy, with particular attention to ventricles, the left atrium, the great vessels, and the interatrial and interventricular septae.
• In patients with limited echocardiographic windows, the use of microbubble contrast for LV opacification may be necessary to provide an accurate LVEF assessment.
Analysis
• Standard measurement and interpretation is done, with careful measurements of ventricular size (LV end-diastolic diameter [LVEDD]) and function, and detection of thrombi, aortopathy, and shunts.
• Real-time three-dimensional (3D) echocardiography may provide more accurate and reproducible measurements of LV volumes and LVEF.
• Severely enlarged RVs (defined in single-center experiences as > 85-mm end-diastolic diameter, > 200-mL end-diastolic volume, > 177-mL end-systolic volume, or short-axis to long-axis ratio ≥0.6), severe qualitative systolic RV dysfunction, tricuspid annular planar systolic motion (<7.5 mm in one study), or moderate to severe tricuspid regurgitation (TR), may require RVAD support (Table 12-3).
| Echocardiographic Finding | Measurement |
|---|---|
| RV enlargement |
RVEDD, right ventricular end-diastolic diameter; RVEDV, right ventricular end-diastolic volume; RVESV, right ventricular end-systolic volume; TAPSE, tricuspid annular plane systolic excursion.
Pitfalls (Box 12-1)
• Placement of a VAD inflow cannula into the apex of a small ventricle risks cannula obstruction from walls, septae, papillary muscles, chordal apparati, ridges, or trabeculae.
• Intracardiac thrombi detected by echocardiography dictate where surgeons can place inflow cannulae (e.g., LA if there is an LV apical thrombus).
• Protruding and mobile plaques, aneurysms, or dissection in the ascending aorta complicate outflow cannula placement.
• ASDs, PFOs, and ventricular septal defects (VSDs) can cause right-to-left shunting and systemic hypoxia and/or paradoxical embolus after LVAD placement.
• The sensitivity of agitated saline contrast (“bubble”) studies to detect PFOs is often reduced because the bubbles cannot pass from right to left due to high left-sided pressures.
Box 12-1 Pitfalls in VAD Imaging
• Reduced ability to detect significant AI preimplantation due to low systemic pressures and high LV filling pressures, which reduce aortoventricular diastolic gradient.
• Assessment of septal shift as a marker of LV filling is complicated by many other factors that influence septal motion.
• Increased aortic valve opening and pulsatile velocities through the VAD cannulae can be a marker for LV overload, cannulae obstruction, or LV recovery.
Physiologic Data
Pitfalls (see Box 12-1)
• Mitral stenosis compromises LVAD inflow through a left ventricular apical cannula and may dictate that the LA be cannulated instead.
• LVAD implantation will worsen baseline aortic insufficiency (AI). Aortic valve replacement or oversewing (which risks thrombus formation) may be necessary.
• In decompensated heart failure (HF), low systemic pressures and high LV pressures may lead to underestimation of AI due to a low aortic-ventricular diastolic gradient.
• Mechanical aortic valves have an increased tendency to form thrombus, despite anticoagulation, when VADs are placed because they may not open or may open minimally. Strong consideration should be given to replacing mechanical aortic valves with bioprosthetic valves at the time of LVAD implantation.
Alternative Approaches
• TEE or contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI) may be necessary to exclude aortopathy prior to outflow cannula placement.
• Dobutamine echocardiography may be necessary to differentiate low-gradient aortic stenosis (AS) from pseudo-stenosis in the presence of severely depressed LVEF.
• Invasive hemodynamic measurements may be necessary to confirm or evaluate the severity of valvular lesions.
Key Points
• A careful echocardiographic examination must be performed to exclude LV thrombi, possibly involving the use of microbubble contrast agents.
Implantation/Intraoperative Echocardiographic Assessment
Anatomic Imaging (Box 12-2)
• Intraoperative TEE guides placement of inflow cannulae away from walls/trabeculae. 3D TEE may be helpful in delineating the relationship of the cannula to the LV wall, papillary muscle, or trabeculations (Figure 12-3).
• Outflow cannulae can be positioned to avoid aneurysms and protruding plaques and to provide some “washing” of the aortic root to prevent thrombus formation.
• Intraoperative TEE directs de-airing of the heart after VAD insertion. TEE can detect bubbles following cannulae anastomosis and prior to unclamping of the aorta, and before and after termination of cardiopulmonary bypass.
• Anastomotic sites, ascending aorta, RV, and anterior LV (anterior chambers) are important areas to examine for air bubbles.
Diagnosis of the Cause of LVAD Dysfunction
Anatomic Imaging
Acquisition
• Degree and frequency of aortic valve opening is assessed by 2D and M-mode in the parasternal long-axis view.
• Cannulae are examined for vegetations, thrombi, and obstruction by papillary muscles or trabeculations in all views (parasternal long axis, parasternal short axis at the apex, and all apical views for the inflow cannula, and parasternal long axis and right upper parasternal for the outflow cannula).
• TEE is usually necessary to investigate endocarditis on valves or cannulae, and for the detection of atrial thrombi (see Box 12-2).
Analysis
• Alternatively, septal bowing into the RV may be caused by LVAD inflow or outflow cannula malposition or kinking.
• Aortic valve opening can be a marker of LV filling. By mechanically unloading the ventricle, LVADs prevent blood from being ejected through the aortic valve. Thus, the aortic valve may not open, or may open minimally and intermittently. Increased aortic valve opening may signal overfilling.
• Chamber underfilling can lead to obstruction by walls, septae, papillary muscles, chordal apparati, ridges, or trabeculae (“suck-down” phenomenon) (Figure 12-4).
• Though the total artificial heart is opaque to ultrasound, the pericardium and portions of the native atria, inferior and superior venae cavae, and pulmonary veins can sometimes be visualized and demonstrate collapse in tamponade.
• Endocarditis of internal VAD components or native or prosthetic valves can lead to VAD dysfunction and septic emboli, in addition to septic shock.
Pitfalls (see Box 12-1)
• Assessment of septal orientation and motion to assess LV filling can be complicated by acute pulmonary embolism, pulmonary arteriolar and venous hypertension, conduction delays/bundle branch blocks, pericardial tamponade and constriction, and post-sternotomy state.
• Although a marker for increased LV filling and cannula obstruction, increased aortic valve opening may, alternatively, be the result of ventricular recovery. Improved systolic function leads to increased ejection through the aortic valve. Aortic valve opening must therefore be interpreted in light of changes in ventricular function and size.
Physiologic Data
Acquisition
• Standard color and spectral Doppler assessment of all valves should be performed. (Even in the absence of aortic valve opening by 2D motion, continuous and pulsed wave [PW] Doppler interrogations through the ventricular outflow tract provide an assessment of residual cardiac function and/or quantify ventricular recovery).
• Color and spectral Doppler (both PW and continuous wave) interrogation of the inflow cannula is performed. PW velocity profile should be recorded with the sample volume within the cannula and at the cannula orifice.
• The best angle of insonation is usually from an off-axis apical view, but the orientation of the cannulae is often variable, and the best Doppler alignment may be achieved from parasternal or even subcostal views.
• Color and spectral Doppler interrogation of the outflow cannula usually requires imaging from a right or left upper sternal window.
Analysis
• Thrombotic partial occlusion produces color Doppler aliasing at the cannula orifice and high Doppler velocities (>2.3 m/s inflow and >2.1 m/s outflow for pulsatile flow pumps, and >2 m/s for continuous flow pumps) (Figure 12-7).
• In continuous flow pumps, interruption of nonpulsatile (continuous diastolic) flow can also signal obstruction.
• Aortic regurgitation can be worsened by LVAD implantation (increased pressure and flow in the proximal aorta), creating a circuit—LV → VAD → aorta → LV—that excludes the systemic circulation.
• VAD regurgitation can be seen by color and spectral Doppler analysis of the cannulae. Inflow valve regurgitation manifests as decreased flows and velocities in the outflow cannulae.
• In VAD regurgitation, there may be a high flow through the VAD compared with the unsupported ventricle.
• In pulsatile VADs, regurgitant volumes (RVVAD) can be calculated by measuring the velocity-time integral (VTI) from PW Doppler examination within the inflow and outflow cannulae, and using the known cannula diameters (diain, diaout), which are usually 12 to 25 mm:
• For nonpulsatile VADs, PW VTI and diameters of the ascending aorta (diaasc) and pulmonary artery (diapa) can be used to indirectly estimate VAD flows and regurgitant fractions (RF):
Pitfalls (see Box 12-1)
• Increases in the pulsatile velocities through the VAD cannulae suggest obstruction and overfilling but may also be seen in ventricular recovery. Velocities should therefore be interpreted in light of ventricular size and function.
• Though the aortic valve does not open or opens minimally with an LVAD, AS may or may not be present.
• Significant AI and AS have been reported months after LVAD implantation. The cause of AS is unclear (leaflets do not appear thickened or calcified but are nonetheless fused). AS may present a problem in the use of LVAD as a bridge to recovery.
• In continuous-flow LVADs, there is often some pulsatile flow through the aortic valve, but thrombus formation is a risk in pulsatile LVADs and on oversewn or stenotic aortic valves.
• After LVAD placement, color Doppler can still assess the degree of aortic and mitral valve regurgitation, but spectral Doppler gradients, other Doppler measures of valvular stenosis (such as mitral valve pressure half-time), and diastolic parameters (such as E and A velocities, tissue Doppler relaxation velocities at the mitral annulus) can be influenced by the LVAD. Normal values have not been established and likely depend on LVAD flow settings.
Alternative Approaches
• Invasive measures of intracardiac hemodynamics may be necessary to establish accurate filling pressures, cardiac output, and pulmonary vascular resistance to guide therapy in the setting of LVAD dysfunction.
Key Points
• Echocardiographic findings that suggest LVAD obstruction include enlarged LVEDD, septal bowing into the RV, spontaneous echocardiographic contrast, increased aortic valve opening, and high velocities in the cannulae.
Echocardiography-Guided LVAD Optimization and Weaning
Anatomic Imaging
Acquisition
• LV size (LVEDD) is measured as the largest cavity dimension in the parasternal long-axis view at the mitral valve leaflet tips by 2D or M-mode.
• The degree and frequency of aortic valve opening is assessed by 2D and M-mode (in the parasternal long-axis view) (Figure 12-8).
Analysis
• Neutral alignment of the interventricular septum indicates proper preload, whereas septal bowing into the RV suggests overfilling, and bowing into the LV suggests underfilling or high RV pressures (Figure 12-9).
• Quantitative and reproducible changes in LV filling in response to LVAD adjustments can be tracked by LVEDD.
• Rotor speed can be set to allow some valve opening in continuous-flow LVADs to prevent aortic root thrombosis.
• RV function and noninvasive pulmonary pressures can be assessed to examine the impact of LVAD settings on right-sided hemodynamics and function.
• Minimal change in LVEDD and increase in LVEF predicts ventricular recovery (LVEF > 45% and LVEDD < 5.5 cm in one study).
• Increase in LVEF in response to dobutamine stress testing suggests the presence of ventricular reserve.
• Increased degree and/or frequency of aortic valve opening on full LVAD support suggests ventricular recovery.
Physiologic Data
Acquisition
• Spectral Doppler measurements of the VTI in the left ventricular outflow tract from the apical 5-chamber view or apical long-axis view can be used to measure native cardiac output at different LVAD settings.
Pitfalls (see Box 12-1)
• Although increasing rotor speed may decompress the ventricle and lead to a reduction in mitral regurgitation, underfilling with “suck down” can lead to disruption of mitral valve chordal function and tethering of mitral valve leaflets, leading to an increase in mitral regurgitation.
Percutaneous Ventricular Assist Devices
Anatomic Imaging
Acquisition
• Standard echocardiographic imaging techniques are used preimplantation to perform a comprehensive 2D and Doppler examination.
• TEE is sometimes necessary to exclude LV thrombi prior to placement of the Impella Recover device, and left atrial and left atrial appendage thrombi prior to placement of a TandemHeart (see Box 12-2).
• Percutaneous VADs are usually placed under transthoracic echocardiographic (TTE), TEE, or intracardiac echocardiographic guidance, which can ensure proper placement of catheters, and can guide trans-septal puncture in TandemHeart placement.
Analysis
• Aortic abnormalities, including extreme tortuosity, aneurysms, and protruding plaques, can complicate Impella Recover placement and function.
• A redundant anterior mitral valve leaflet can be drawn into and obstruct the Impella Recover inflow port.
• The inflow port of the Impella Recover device should be positioned 3 to 4 cm below the aortic valve, away from the septum and the anterior mitral valve leaflet.
• Echocardiographic guidance during trans-septal puncture of the TandemHeart VAD can prevent inadvertent puncture of the aortic root, coronary sinus, right atrium, or left atrium.
• The TandemHeart inflow cannula should be positioned in the LA away from ridges, septae, atrial walls, and mitral valve leaflets.
• Echocardiography can confirm proper placement and diastolic inflation of the IABP, as well as diagnosing vegetations or thrombi on the IABP tip.
• Echocardiography can ensure proper placement of ECMO cannulae, particularly guiding the RA inflow cannula away from walls, ridges, valves, Chiari networks, and aneurysmal interatrial septae (Figure 12-10).
Pitfalls (see Box 12-1)
• The Impella Recover catheter can migrate forward, moving the outflow cannula below the aortic valve where it ejects blood into the LV, rather than the aorta (Figure 12-11).
• Alternatively, it may migrate backward, moving the inflow cannula above the aortic valve where it draws blood in from the aorta, rather than the LV.
• The TandemHeart inflow cannula can migrate into the left atrial appendage or pulmonary veins or across the mitral valve (potentially causing regurgitation).
Physiologic Data
Pitfalls (see Box 12-1)
• Moderate or great AS (valve area ≤ 1.5 cm2) and more than moderate aortic regurgitation complicates placement and function of the Impella Recover device.
• Severe aortic regurgitation can cause diastolic LV pressure elevation and thereby augment the reduction in coronary perfusion pressure caused by the fact that the TandemHeart shunts blood from the left atrium to the outflow catheter in the distal aorta (bypassing the coronary ostia in the aortic root).
Key Points
• Echocardiography can detect aortic stenosis or regurgitation, aortopathy, and VSD that complicate percutaneous VAD placement.
1 Catena E, Milazzo F. Echocardiography and cardiac assist devices. Minerva Cardiolangiol. 2005;55:247-265.
2 Catena E, Milazzo F, Pittella G, et al. Echocardiographic approach in a new left ventricular assist device: Impella Recover 100. J Am Soc Echocardiogr. 2004;17:470-473.
This paper provides an overview of the early experience with the Impella device.
3 Catena E, Milazzo F, Montorsi E, et al. Left ventricular support by axial flow pump: The echocardiographic approach to device malfunction. J Am Soc Echocardiogr. 2005;18:1422e7-1422e13.
4 Chumnanvej S, Wood MJ, MacGillivray TE, Vidal Melo MF. Perioperative echocardiographic examination for ventricular assist device implantation. Anesth Analg. 2007;105:583-601.
5 Horton SC, Khodaverdian R, Chatelain P, et al. Left ventricular assist device malfunction: An approach to diagnosis by echocardiography. J Am Coll Cardiol. 2005;45:1435-1440.
6 John R, Mantz K, Eckman P, Rose A, May-Newman K. Aortic valve pathophysiology during left ventricular assist device support. J Heart Lung Transplant. 2010;29:1321-1329.
7 Kirkpatrick JN, Wiegers SE, Lang RM. Left ventricular assist devices and other devices for end-stage heart failure: Utility of echocardiography. Curr Cardiol Reports. 2010;12:257-264.
8 Scalia GM, McCarthy PM, Savage RM, Smedira NG, Thomas JD. Clinical utility of echocardiography in the management of implantable ventricular assist device. J Am Soc Echocardiogr. 2000;13:754-763.
