Attention ASE Members:
The ASE has gone green! Visit www.aseuniversity.org to earn free continuing medical education credit through an online activity related to this article. Certificates are available for immediate access upon successful completion of the activity. Nonmembers will need to join the ASE to access this great member benefit!
Introduction
This guideline addresses the role of echocardiography during the different phases of care of patients with long-term, surgically implanted continuous-flow (CF) left ventricular (LV) assist devices (LVADs). In patients with advanced heart failure (HF) refractory to medical therapy, LVADs have been used as a bridge to transplantation (BTT), as destination therapy (DT), as a bridge to transplant candidacy, or as a bridge to recovery. Over the past three decades, tremendous progress has been made in the field of mechanical circulatory support (MCS), and more than 30,000 patients worldwide have received long-term LVADs. Recent guidelines endorse the important role of echocardiography in the clinical care of LVAD patients at several stages, including preoperative patient selection, perioperative imaging, postoperative surveillance, optimization of LVAD function, troubleshooting of LVAD alarms, and evaluation of native myocardial recovery. Despite increasing clinical use of LVADs, recognition of the central role of echocardiography in their management, and presentation of an exponentially expanding outpatient LVAD population to healthcare facilities not directly associated with implantation centers, there is a lack of published guidelines for echocardiography of LVAD recipients.
This American Society of Echocardiography (ASE) document uses both published data (albeit of limited availability at this time) and expert opinion from high-volume MCS-device implantation centers to provide consensus recommendations and sample protocols for the timing and performance of echocardiography during LVAD patient selection, device implantation, and postoperative management. The authors’ goal is to provide a general framework for the interactions between echocardiography laboratories and MCS teams. Although numerous types of LVADs are in clinical use or under development, the scope of this document is primarily limited to current surgically implanted CF-LVADs that have been approved by the United States Food and Drug Administration (FDA) for extended use in adults. Pediatric and adult patients with congenital heart disease represent a smaller but important and increasing subpopulation of patients receiving extended-use MCS devices. Comments or recommendations specifically relating to pediatric and congenital heart disease patients will be noted within the text and within the pediatric LVAD discussion section. Surgically implanted LVADs for short-term use, percutaneously implanted LVADs, right ventricular (RV) assist devices (RVADs), and/or biventricular assist devices (BiVADs) may also be encountered by echocardiographers. A brief discussion of these devices and their applications is included in Appendix A . Other MCS devices, including cardiopulmonary bypass (CPB) pumps, extracorporeal membrane oxygenation (ECMO), intraaortic balloon pumps (IABPs), and total artificial hearts (TAHs), are not covered in this report.
Introduction
This guideline addresses the role of echocardiography during the different phases of care of patients with long-term, surgically implanted continuous-flow (CF) left ventricular (LV) assist devices (LVADs). In patients with advanced heart failure (HF) refractory to medical therapy, LVADs have been used as a bridge to transplantation (BTT), as destination therapy (DT), as a bridge to transplant candidacy, or as a bridge to recovery. Over the past three decades, tremendous progress has been made in the field of mechanical circulatory support (MCS), and more than 30,000 patients worldwide have received long-term LVADs. Recent guidelines endorse the important role of echocardiography in the clinical care of LVAD patients at several stages, including preoperative patient selection, perioperative imaging, postoperative surveillance, optimization of LVAD function, troubleshooting of LVAD alarms, and evaluation of native myocardial recovery. Despite increasing clinical use of LVADs, recognition of the central role of echocardiography in their management, and presentation of an exponentially expanding outpatient LVAD population to healthcare facilities not directly associated with implantation centers, there is a lack of published guidelines for echocardiography of LVAD recipients.
This American Society of Echocardiography (ASE) document uses both published data (albeit of limited availability at this time) and expert opinion from high-volume MCS-device implantation centers to provide consensus recommendations and sample protocols for the timing and performance of echocardiography during LVAD patient selection, device implantation, and postoperative management. The authors’ goal is to provide a general framework for the interactions between echocardiography laboratories and MCS teams. Although numerous types of LVADs are in clinical use or under development, the scope of this document is primarily limited to current surgically implanted CF-LVADs that have been approved by the United States Food and Drug Administration (FDA) for extended use in adults. Pediatric and adult patients with congenital heart disease represent a smaller but important and increasing subpopulation of patients receiving extended-use MCS devices. Comments or recommendations specifically relating to pediatric and congenital heart disease patients will be noted within the text and within the pediatric LVAD discussion section. Surgically implanted LVADs for short-term use, percutaneously implanted LVADs, right ventricular (RV) assist devices (RVADs), and/or biventricular assist devices (BiVADs) may also be encountered by echocardiographers. A brief discussion of these devices and their applications is included in Appendix A . Other MCS devices, including cardiopulmonary bypass (CPB) pumps, extracorporeal membrane oxygenation (ECMO), intraaortic balloon pumps (IABPs), and total artificial hearts (TAHs), are not covered in this report.
Key Points
- •
This document addresses the role of echocardiography during the different phases of care of patients with FDA-approved long-term, surgically implanted CF-LVADs.
- •
The phases of patient care addressed include preoperative patient selection, perioperative TEE imaging, postoperative surveillance, optimization of LVAD function, problem-focused exams (when the patient has signs or symptoms of LVAD or native cardiac dysfunction), and evaluation of native myocardial recovery.
- •
Suggested protocols, checklists, and worksheets for each of these phases of care are located in the Appendices.
- •
Other types of MCS may also be encountered by echocardiographers, and these devices are discussed in Appendix A .
- •
Although echocardiography is frequently used for managing LVAD therapy, published data intended to guide timing and necessary data collection remain limited. Some of the recommendations provided herein are based on expert consensus from high-volume MCS implant centers.
- •
Most LVAD recipients are adults with dilated cardiomyopathies. Other LVAD patient populations addressed within this document include those with smaller hearts (eg, resulting from restrictive cardiomyopathies) and those with pediatric and congenital heart disease.
- •
The authors’ goal is to provide a general framework for the interactions between echocardiography laboratories and MCS teams.
Left Ventricular Assist Devices
Selection of a particular LVAD for an individual patient is a complex decision-making process and is beyond the scope of this document. Readers are referred to recent reviews for a comprehensive explanation of the structure and function of long-term, surgically implanted, intracorporeal (pump inside the body) LVADs and of short-term, surgically or percutaneously implanted, extracorporeal (pump outside the body) LVADs which are described in Appendix A . Currently, two CF-LVADs are approved by the FDA for surgical implantation in adults—the HeartMate II (HM-II) Left Ventricular Assist System (Thoratec Corporation, Pleasanton, CA) ( Figure 1 ) and the HVAD Ventricular Assist System (HeartWare International, Inc., Framingham, MA) ( Figure 2 ). The HM-II received FDA approval for BTT therapy in April 2008 and for DT in January 2010. The HeartWare HVAD received FDA approval for BTT therapy in November 2012, and a DT trial of this system is ongoing. For brevity, the abbreviation “LVAD” will be used here when referring to either of these CF-LVADs.
Common to both the HM-II and the HVAD are three components in series: (1) an inflow cannula positioned in the left ventricle near the apex, (2) a mechanical impeller, and (3) an outflow graft anastomosed to the ascending aorta ( Figures 1 and 2 ). Echocardiography allows direct visualization of the proximal inflow cannula and the distal outflow graft but not of the mechanical impeller. The HM-II impeller and its housing structure are implanted below the diaphragm, whereas the HVAD impeller and its housing structure are implanted above the diaphragm, within the pericardial sac. Discussed in further detail below, impeller positioning is the primary differentiating factor in the echocardiographic evaluation of the inflow-cannula flow of these two devices. In other respects, echocardiographic evaluation of the two pumps is similar. Furthermore, both the HM-II and the HVAD are powered by a driveline connected to an extracorporeal controller. In addition to serving as a power source, the controller continually measures and calculates a number of parameters related to LVAD function. When these parameters fall outside predetermined normal ranges, the controller alerts the patient and the HF team that there is a problem. The implications of controller alarms for echocardiography are further discussed below.
Key Points
- •
Current CF-LVADs have three intracorporeal (inside the body) components: an LV inflow cannula, a mechanical impeller, and an outflow graft that is anastomosed to the ascending aorta.
- •
The mechanical impeller is attached to an extracorporeal (outside the body) controller device via a driveline that provides power and a data link. The controller monitors several LVAD-related parameters and may generate device alarms. In turn, these alarms may indicate the need for an echocardiogram to validate the alarm and provide a definitive diagnosis.
- •
Echocardiography techniques for different devices are generally similar, except for important differences noted in the text.
The Role of Echocardiography in Candidate Selection
Optimal candidate selection is one of the most important determinants of a successful operative and long-term outcome for LVAD recipients. Transthoracic echocardiography (TTE) is generally the first-line imaging modality used to screen LVAD candidates for structural and/or functional abnormalities that represent absolute or relative contraindications to device implantation. In some cases, patients require urgent or emergent surgical LVAD placement. In these acute situations, adequate TTE information may be technically limited or unavailable. Therefore, transesophageal echocardiography (TEE) performed in the acute setting (catheterization laboratory, emergency department, intensive care unit, or operating room) should address all of the factors mentioned below with regard to TTE. Given its central role in LVAD candidate selection, preimplantation TTE or TEE (when necessary) should be performed in a laboratory that has been accredited by the Intersocietal Accreditation Commission (IAC) and should be supervised and interpreted by a skilled echocardiographer who is experienced in advanced HF evaluation and the hemodynamic assessment of MCS devices. Preimplantation TTE in LVAD candidates should include all the elements of a comprehensive examination as recommended by the ASE, with a particular focus on the high-risk or “red-flag” findings detailed below and summarized in Table 1 . A comprehensive, checklist-based preimplantation TTE protocol with the notation of red-flag findings is available in Appendix B . If preimplantation TTE yields inconclusive findings, TEE may be performed, as described below. If a recently performed high-quality TTE exam includes most but not all of the required preimplantation exam elements and there has been no interval change in the patient’s clinical status, a limited, focused follow-up exam to obtain the additional necessary information may be acceptable.
| Left Ventricle and Interventricular Septum |
| Small LV size, particularly with increased LV trabeculation LV thrombus LV apical aneurysm Ventricular septal defect |
| Right Ventricle |
| RV dilatation RV systolic dysfunction |
| Atria, Interatrial Septum, and Inferior Vena Cava |
| Left atrial appendage thrombus PFO or atrial septal defect |
| Valvular Abnormalities |
| Any prosthetic valve (especially mechanical AV or MV) > mild AR ≥ moderate MS ≥ moderate TR or > mild TS > mild PS; ≥ moderate PR |
| Other |
| Any congenital heart disease Aortic pathology: aneurysm, dissection, atheroma, coarctation Mobile mass lesion Other shunts: patent ductus arteriosus, intrapulmonary |
LV Dysfunction
Severe LV systolic dysfunction resulting from a dilated cardiomyopathy characterizes the majority of LVAD recipients. Accordingly, echocardiography laboratories must be proficient in techniques for measuring LV size, ejection fraction (LVEF), and cardiac output.
LV Ejection Fraction
Demonstration of an LVEF of <25% is a Centers for Medicare & Medicaid (CMS)-qualifying condition for LVAD implantation as DT. Additionally, the LVEF is a component of both the Seattle Heart Failure Model and the Heart Failure Survival Score, two clinical-risk scoring tools that are widely used by HF specialists to calculate patients’ expected survival times and, by extension, their suitability for an LVAD. A severely decreased LVEF is by no means the only clinical parameter used for determining whether or not a patient is referred for MCS. However, its accurate measurement by echocardiography is of paramount importance. Previous ASE guidelines describe the recommended methods for echocardiographic LV chamber quantification. On the basis of those guidelines, laboratories with the ability and expertise to perform three-dimensional (3D) assessment for determining LV volumes and the LVEF should routinely do so when imaging conditions permit; otherwise, they should use the biplane method of disks (modified Simpson’s rule) from two-dimensional (2D) images. Strong consideration should be given to the use of a microbubble contrast agent when indicated to enhance endocardial definition and improve the precision of LVEF measurement.
LV Internal Dimension at End-Diastole
In addition to the LVEF, the LV internal dimension at end-diastole (LVIDd) from 2D parasternal long-axis images is a critical measurement in LVAD candidates. For patients who eventually undergo LVAD implantation, comparison of the preoperative LVIDd to the postoperative LVIDd is the primary clinical measure of the degree of LVAD-mediated LV unloading. Whereas a comparison of pre- and postoperative LV end-diastolic volumes (LVEDVs) would better quantify LV unloading, these measurements can be extremely challenging to obtain in the immediate postoperative period, when standard echocardiographic windows are limited by supine positioning, mechanical ventilation, a recent sternotomy, bandages, and other physical barriers. While the LVIDd and LVEDV are moderately to severely increased in most patients considered for an LVAD, limited data suggest that a smaller LV cavity, defined by an LVIDd of <63 mm, is associated with increased 30-day morbidity and mortality rates after LVAD implantation. Patients who tend to have smaller LV cavities include elderly women with a smaller body habitus and persons with infiltrative cardiomyopathies (eg, amyloidosis). The latter group may also have concomitant right-sided HF, another preoperative high-risk finding that is discussed below. Whereas a small LV cavity is not an absolute contraindication to LVAD implantation, the presence of this finding should be communicated to the HF team.
Intracardiac Thrombi
An intracardiac thrombus is not an absolute contraindication for LVAD implantation but may increase the risk of stroke during the LV cannulation portion of the procedure. At particularly increased risk for LV thrombus are patients with a severely decreased LVEF and/or an LV aneurysm. In these patients, strong consideration should be given to the use of a microbubble contrast agent during assessment for LV thrombus. If such a thrombus is identified, the implanting surgeon should be made aware of its size and location so that the thrombus can be carefully removed during device implantation. In borderline cases, cardiac computed tomography (CCT) may be adjunctively used to rule out an LV thrombus. In patients with atrial fibrillation, who are at increased risk for thrombus in the left atrial appendage, adjunctive TEE may be required for complete intracardiac thrombus assessment.
RV Dysfunction
Echocardiographic signs of RV dysfunction include impaired RV systolic function and/or RV dilatation, increased RA pressure (ascertained by inferior vena cava size and collapsibility), and moderate or greater tricuspid regurgitation (TR). Previous ASE guidelines describe the recommended methods for echocardiographic evaluation of RV function and chamber quantification. On the basis of those guidelines, 3D echocardiographic assessment of RV volumes to calculate the RV ejection fraction would be ideal, but the authors realize that this approach is technically challenging and not widely available. Measurement of other secondary echocardiographic surrogates of RV systolic function, including RV fractional area change (FAC), tricuspid annular-plane systolic excursion (TAPSE), and RV free-wall peak longitudinal strain, can be difficult in patients with advanced HF. Nonetheless, quantitative measures of RV function are recommended for use whenever possible, but only when able to be properly measured in a given patient. At a minimum, a qualitative assessment of RV size and systolic function and of TR severity should be performed and communicated in the interpretation.
Echocardiographic signs of RV dysfunction should not be considered in isolation. They should be integrated with a patient’s clinical signs and symptoms of possible right-sided heart failure syndrome. Clinically severe preoperative RV dysfunction may prompt the HF team to consider planned biventricular MCS, as this may lead to better outcomes than delayed conversion of an LVAD to biventricular MCS. Some patients with less than severe RV dysfunction at preoperative assessment will develop severe RV dysfunction after LVAD implantation. This complication, defined by the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) as the requirement of an RV assist device (RVAD) or >14 consecutive days of intravenous (IV) inotropic support, has an estimated prevalence of 13% to 44% and is associated with significant morbidity and mortality. Preliminary data suggest that there may be preoperative echocardiographic parameters predictive of severe postoperative RV dysfunction. In studies that included clinical parameters in their multivariable models, an RV absolute peak longitudinal strain of <9.6% and an RV: LV end-diastolic diameter ratio of >0.75 were identified as potential independently predictive echocardiographic parameters. More recent data by Kato and colleagues suggests that the accuracy for predicting post LVAD RV failure may be improved when more than one RV echocardiographic parameter (in this case RV tissue Doppler imaging and RV speckle tracking imaging [RV longitudinal strain]) are used in aggregate. Given the lack of consensus thus far regarding the predictive value of any single echocardiographic parameter, an aggregate assessment utilizing relevant left-sided parameters (eg, indexed left atrial volume, indexed LV size) and right-sided parameters (eg, RV parameters described above, TR severity, and right atrial [RA] pressure estimation) is likely the optimal approach for now.
Key Points
- •
Nearly all LVAD candidates undergo echocardiography to screen for structural and/or functional abnormalities that preclude LVAD implantation or that may alter surgical planning.
- •
At this time, the literature does not support the use of any single echocardiographic RV parameter for predicting the post-LVAD prognosis or the need for biventricular support (RVAD use).
- •
Quantitative echocardiographic parameters of RV function (which may vary among patients, depending upon imaging conditions), should be integrated with clinical signs and symptoms to determine the degree of preoperative RV dysfunction, which may impact the operative plan and/or postoperative prognosis.
Valve Disease
Previous ASE guidelines address detection and quantitation of valvular regurgitation, valvular stenosis, and prosthetic valve dysfunction.
Valve Stenosis
In patients with advanced HF and a severely reduced stroke volume, spectral Doppler-derived valve gradients in isolation may not accurately reflect the degree of valvular stenosis. In these patients, calculation of the valvular orifice area may be more accurate. Moderate or severe mitral valve (MV) stenosis can prevent adequate LVAD cannula inflow. Accordingly, significant mitral stenosis (MS) must be corrected before LVAD implantation. In contrast, aortic stenosis (AS) of any severity may be present without affecting LVAD function, because LVADs completely bypass the native LV outflow tract (LVOT). It is important to note, however, that patients who have critical AS or who undergo surgical aortic valve (AV) complete closure to correct aortic regurgitation (AR) will have no forward flow in the event of obstructive LVAD failure, even if residual LV function is present.
Valve Regurgitation
Exclusion of significant AR before LVAD implantation is critical and sometimes challenging. When present at LVAD implantation, significant AR enables a “blind” loop of flow in which blood enters the LVAD from the left ventricle, is pumped into the ascending aorta, but then flows back into the left ventricle through the regurgitant aortic valve. It may be difficult for echocardiographers to determine the degree of AR present in patients with advanced HF and a severely reduced LV stroke volume. Heart failure patients with moderate or severe AR may have unimpressive color-flow Doppler images and low AR velocities due to low systemic pressures and high LV diastolic pressures. Additionally, the aortic regurgitant volume may be relatively low, despite a high regurgitant fraction. Accordingly, the Doppler-derived LVOT stroke volume and regurgitant fraction should be calculated routinely when possible. Furthermore, there should be a high level of suspicion for significant AR in the presence of aortic root dilatation, eccentric AR (particularly if associated with a bicommissural AV), rheumatic or calcific AV degeneration, or an aortic prosthesis. TEE should be strongly considered when there is any degree of suspected abnormal prosthetic valve regurgitation. The presence of more than mild AR should be communicated to the implanting surgeon, because recent guidelines advise confirmation by perioperative TEE and surgical correction of AR before LVAD implantation. Surgical treatment options for significant native valve AR include replacement with a bioprosthesis, completely oversewing the valve (by suturing along all coaptation zones) or by performing a central coaptation (Park) stitch. Completely over sewing the AV cusps effectively eliminates AR, but (as mentioned above) leaves the patient with no fail-safe means of LV ejection in the event of LVAD failure. When the aortic cusp integrity is good, a central coaptation (Park) stitch technique can treat central AR while allowing aortic forward flow through the residual commissural zones to occur during reduced LVAD support ( Figure 3 ) or in the event of LVAD pump failure.
Mitral regurgitation (MR) that is significant preoperatively is often markedly improved after initiation of LVAD support, because of reduced LV size, reduced filling pressures, and improved coaptation of the MV leaflets. For this reason, any degree of MR is acceptable in LVAD candidates. In contrast, moderate or greater TR is a potentially ominous finding, which may indicate RV dysfunction as mentioned above. It is important to communicate the presence of significant TR to the implanting surgeon; recent guidelines recommend that surgical tricuspid valve repair be considered at the time of LVAD implantation. Pulmonary regurgitation (PR) may be more commonly encountered in patients with congenital heart disease. In any patient, moderate or greater PR could contribute to preoperative RV dysfunction and would require repair in the event of RVAD implantation. However, PR may be well tolerated in the setting of successful LVAD implantation with adequate RV function and successful LV unloading. However, significant PR could potentially contribute to RV dysfunction after LVAD implantation if pulmonary vascular resistance were increased for any reason, including acquired pulmonary disease or an inability to adequately unload the left ventricle.
Prosthetic Valves
When indicated, prosthetic valve assessment by TTE and TEE is critical for surgical decision-making. LVAD-supported patients must receive systemic anticoagulation, regardless of the presence of mechanical prosthetic valves. However, a higher target prothrombin time international normalized ratio (PT INR) may be necessary if a mechanical valve is present. After initiation of LVAD support, the inherent reduced flow through a mechanical AV prosthesis further increases the risk of postoperative valvular or aortic root thrombosis and subsequent thromboembolic events. For this reason, replacement of even a normally functioning mechanical AV prosthesis with a bioprosthesis or valve closure should be considered at the time of LVAD implantation. Adequately functioning bioprosthetic AVs do not require removal or replacement. Similarly, surgical replacement of a normally functioning mechanical MV prosthesis is typically not recommended, even if significant MR is present, as obligatory forward transmitral flow will occur during MCS. An important exception is the presence of moderate or worse mechanical MV stenosis. In these cases, consideration should be given to MV replacement with a bioprosthesis at LVAD implantation. Although not frequently encountered, tricuspid or pulmonary valve prosthesis dysfunction is an important finding, as it could adversely affect postoperative RV function.
Key Points
- •
The position, type, and functioning of any prosthetic valve can have an important impact on surgical and postoperative management, and adjunctive TEE imaging should be performed if clinically indicated.
- •
Aortic regurgitation warrants special attention, as it can easily be underestimated in HF patients, generally worsens after LVAD activation, and impairs LV unloading due to a “blind loop” of aorta→LV→LVAD flow.
- •
Moderate or greater TR is an ominous finding, especially if accompanied by other signs or symptoms of RV dysfunction.
- •
A mechanical AV should be replaced before LVAD implantation.
- •
Severe AS and even complete AV closure can be tolerated after LVAD implantation, although either of these conditions results in the lack of a fail-safe mechanism for LV output in the event of LVAD failure.
- •
Mitral regurgitation is generally well tolerated and may improve after LVAD implantation.
Congenital Heart Disease
For all patients with known congenital heart disease of any severity, previous imaging studies documenting cardiac morphology, shunts, collateral vessels, and/or the location and course of the great vessels should be reviewed. Recent data suggest that with amenable cardiac anatomy, even patients who have complex congenital heart disease can undergo implantation of an LVAD as a BTT or as DT. Some common anomalies require correction before LVAD implantation. A patent foramen ovale (PFO), present in up to 30% of the general population, increases the risk of hypoxemia and paradoxical embolization in patients receiving LVAD support. For this reason, PFOs or any other interatrial communications should be closed at the time of device implantation. In evaluating patients with advanced HF for atrial septal defects (ASDs) and PFOs, the use of IV agitated saline combined with an appropriately performed Valsalva maneuver is necessary, because elevated left and/or RA pressures may reduce interatrial pressure gradients and preclude detection of the defect by color Doppler imaging or agitated saline injection alone. Like ASDs, congenital and post–myocardial infarction ventricular septal defects (VSDs) can also result in immediate postimplantation right-to-left shunting with hypoxemia and a risk of paradoxical embolization during LVAD support. The presence of VSDs should be systematically excluded by color Doppler interrogation of the entire ventricular septum; if identified, VSDs should be closed at LVAD implantation. In most cases, an unrepaired VSD is an absolute contraindication to device implantation. However, selected patients with single ventricle physiology (and an unrepaired VSD) may be considered for an LVAD.
Key Points
- •
In patients with congenital heart disease, echocardiography is an important complementary imaging modality after other, previous imaging studies have been reviewed.
- •
The echocardiography exam should systematically exclude the presence of a PFO or other intracardiac shunt, which should be electively repaired at the time of surgery to avoid sudden arterial oxygen desaturation after LVAD activation.
Other High-Risk Findings
Acute endocarditis (or any other active infection) is an absolute contraindication to MCS-device implantation because of the risk of bacterial seeding of a newly implanted LVAD. As a result, a mobile mass lesion suggestive of a possible vegetation is a high-risk finding. Diseases of the aorta that are relative or absolute contraindications to LVAD implantation (eg, significant aneurysmal dilatation, dissection) may be discovered on TTE. For this purpose, high parasternal long-axis, suprasternal notch, and subcostal views of the aorta should be attempted. TEE may be very useful for the diagnosis of thoracic aorta pathology. However, in the absence of contraindications to contrast agents, computed tomography (CT) or magnetic resonance imaging (MRI)—barring MRI contraindications —are preferred modalities for comprehensive imaging of the aorta before LVAD implantation.
Perioperative Transesophageal Echocardiography
Preimplantation TEE
Comprehensive perioperative TEE should be performed in the operating room before LVAD implantation, with additional imaging performed at the time of LVAD activation and after a period of stabilization. Preimplantation TEE is particularly important when urgent or emergent LVAD placement is required, in which case this modality may serve as the primary screening echocardiography examination. Previous guidelines describe the recommended approach for perioperative TEE A comprehensive; checklist-based pre- and postimplantation perioperative TEE protocol with notation of red-flag findings is included in Appendix C . The physician performing the examination should be a highly trained cardiologist with significant advanced TEE and perioperative TEE experience or a cardiovascular anesthesiologist with advanced perioperative TEE training. Among the most important aspects of preimplantation TEE are reevaluation of the degree of AR, determination of the presence or absence of a cardiac-level shunt, identification of intracardiac thrombi, assessment of RV function, and evaluation of the degree of TR. These and potentially other important conditions (eg, degree of MS, PR, prosthetic dysfunction, possible vegetations, aortic disease, etc.) may have been undiagnosed or underappreciated on previous imaging exams or may have progressed in the intervening time. Their presence may necessitate conversion of a planned “off-pump” case into one that requires CPB, a change from a limited thoracotomy to a sternotomy to enable needed repairs, or possibly biventricular MCS.
For the same reasons discussed above for TTE, the degree of AR on perioperative TEE may be underappreciated on color Doppler imaging during general anesthesia, because low mean arterial pressure and/or systemic vascular resistance may be present. As a result, adequate AR assessment may necessitate systemic blood pressure (BP) augmentation by vasopressor agents. With regard to PFO detection, thorough color Doppler scanning of the fossa ovalis margins at a low Nyquist-limit setting and IV injection of agitated saline may be inconclusive. In these cases, IV injection of agitated saline combined with a “ventilator” Valsalva maneuver may also be useful. This maneuver involves injecting agitated saline into a central IV line (eg, internal jugular) during a briefly sustained application of up to 30 cmH 2 O of intrathoracic pressure and, on opacification of the right atrium, release of the intrathoracic pressure. Even with this maneuver, in some cases, significant competitive inferior vena cava ”negative contrast” flow in the fossa ovalis region can cause a false-negative PFO evaluation after saline injection into the superior vena cava. Injection of agitated saline into a femoral vein may increase PFO detection if such access is available. Despite all efforts, a PFO may not become apparent in some cases until MCS is initiated and the left atrial pressure is decreased.
Key Points
- •
The preimplant perioperative TEE is an important confirmatory imaging study, which can identify previously underappreciated or undiagnosed pathologic conditions that may influence the surgical procedure.
- •
An LVAD perioperative TEE checklist can be useful for laboratory personnel (see Appendix C ).
- •
Preimplantation TEE should include reevaluation of AR, RV function, TR, and the aorta. Cardiac-level shunts and intracardiac thrombi should be excluded.
- •
Evaluation for PFO may require special imaging maneuvers as outlined in the text. Despite best efforts, a PFO may not be able to be diagnosed prior to LVAD implantation.
Perioperative TEE During LVAD Implantation
Both the HM-II and the HVAD require coring in the region of the LV apex for inflow-cannula insertion. This part of the procedure is inevitably accompanied by some degree of entrained air on the left side of the heart. Subsequent de-airing maneuvers require continuous TEE guidance. The left atrium, left ventricle (including the LV apex and inflow cannula ( Figures 4 and 5 ), aortic root, ascending aorta, outflow graft-to-ascending aorta anastomosis ( Figure 6 ), and transverse and descending aorta should all be directly visualized and carefully inspected for signs of air. The ostium of the right coronary artery (RCA) is situated anteriorly in the aortic root and is a common destination for air ejected from the left ventricle. Acute RV dysfunction or dilatation and/or an increase in the severity of TR should suggest the possibility of air embolization to the RCA, and this complication may resolve with watchful waiting. As during the LV coring procedure, the period immediately after separation from cardiopulmonary bypass and reinstitution of mechanical ventilation can be accompanied by the sudden appearance of new air bubbles originating from the pulmonary veins, left atrium, or left ventricle. This finding, if associated with signs of RV dysfunction from a presumed coronary air embolism, may signal the need for reinstitution of CPB and/or repeat de-airing maneuvers.
Perioperative TEE During Initial LVAD Activation and Speed Optimization
Upon LVAD activation, the device name and the initial pump speed should be annotated on the imaging screen. Although the exact order of perioperative TEE views obtained after LVAD initiation may vary among centers, it is recommended that physicians follow an LVAD checklist-based protocol ( Appendix C ) to include all of the important components unique to postoperative LVAD assessment. Table 2 lists possible abnormal findings detectable by echocardiography after LVAD implantation. Early imaging of the interatrial septum with color Doppler and with IV injection of agitated saline contrast to confirm the absence of an atrial septal communication is recommended. This is particularly important if initiation of LVAD support results in a sudden decrease in arterial oxygen saturation, the hallmark of an “unmasked” PFO or other right-to-left shunt ( Figure 7 ). Next, the degree of AV opening and the degree of AR (if any) should be assessed. When there is no AV opening, this may be apparent with standard planar imaging. In many cases, the extent and duration of aortic cusp separation may be markedly reduced or only intermittent, depending upon the degree of LVAD support (pump speed). M-mode imaging of the AV in the long-axis view can be helpful for measuring and reporting the degree of AV opening ( Figure 8 ). When there is minimal residual native LVOT forward flow, AV opening may be intermittent due to pulsus alternans in regular sinus rhythm or because of arrhythmias. A slow M-mode sweep speed (eg, 25 mm/sec to acquire more cardiac cycles) may be needed to adequately display intermittent AV opening ( Figure 8C- E).
| Pericardial effusion |
| With or without cardiac tamponade including RV compression. Tamponade : respirophasic flow changes; poor RVOT SV. |
| LV failure secondary to partial LV unloading |
|
| RV failure |
|
| Inadequate LV filling or excessive LV unloading |
| Small LV dimensions (typically <3 cm and/or marked deviation of interventricular septum towards LV). Note: May be due to RV failure and/or pump speed too high for loading conditions. |
| LVAD suction with induced ventricular ectopy |
| Underfilled LV and mechanical impact of inflow cannula with LV endocardium, typically septum, resolves with speed turndown. |
| LVAD-related continuous aortic insufficiency |
| Clinically significant—at least moderate and possibly severe—characterized by an AR proximal jet-to-LVOT height ratio >46%, or AR vena contracta ≥3 mm; increased LV size and relatively decreased RVOT SV despite normal/increased inflow cannula and/or outflow graft flows. |
| LVAD-related mitral regurgitation |
|
| Intracardiac thrombus |
| Including right and left atrial, LV apical, and aortic root thrombus |
| Inflow-cannula abnormality |
|
| Outflow-graft abnormality |
|
| Hypertensive emergency |
| New reduced/minimal AV opening relative to baseline exam at normal BP, especially if associated with new/worsened LV dilatation and worsening MR. Note: hypertension may follow an increase in pump speed. |
| Pump malfunction/pump arrest: |
|
∗ Note: based on observational data. The “normal” outflow graft peak velocities are not well defined. Because the HVAD outflow graft diameter is smaller than that of the HM II device (see discussion in text). Therefore, the normal Doppler-derived HVAD outflow velocities may be somewhat higher on average than those observed for the HM II LVAD.
Aortic Regurgitation
A pump speed–dependent reduction in LV diastolic filling pressures and increased central aortic BP can lead to the appearance of more prominent AR on color Doppler imaging than was appreciated before pump implantation ( Figure 9 A). During LVAD support, AR can be intermittent (depending upon the valve opening duration), predominantly diastolic, nearly continuous (extending into the normal systolic phase of the cardiac cycle), or continuous (holosystolic and holodiastolic). Measuring the temporal occurrence of AR can be achieved with color M-mode and continuous-wave (CW) Doppler (see Figure 9F- G). The AR duration, AR vena contracta (VC) width, LVOT AR jet height, and other evidence of hemodynamically significant AR (discussed in further detail, below) should guide the need for possible surgical intervention on the AV. Although the AR VC width may be useful in a qualitative sense, during LVAD support, the width may vary throughout the cardiac cycle with continuous AR ( Figure 9 C) and at different pump speeds ( Figure 9 D,E). Methods for assessing AR severity in the context of an LVAD problem-focused exam are further discussed below. However, in keeping with previous guideline recommendations, a VC width of ≥0.3 cm or a jet width/LVOT width of >46% at a Nyquist limit of 50-60 cm/s should be considered to indicate at least moderate (and possibly severe) AR, owing to the prolonged (if not continuous) duration of AR during LVAD support. Neither the AR pressure half-time method nor pulsed Doppler evaluation of aortic diastolic flow reversal is a reliable method for AR severity assessment after LVAD implantation. This is because the AR duration extends into the systolic ejection period. In addition, both of these methods are highly affected by LV preload, LV afterload, and aortic pulse pressure, which is diminished during LVAD support. However, CW spectral Doppler imagining may be useful for evaluating the timing and duration of AR ( Figure 9 F, G), and it’s pixel intensity may be additive to the qualitative assessment.
RV Dysfunction
Poor RV performance, with or without significant TR, immediately after LVAD initiation is not uncommon due to rapid normalization of the RV preload by the pump. Early RV dysfunction may be transient, due to CPB-related factors, or refractory, due to underlying RV dysfunction. In this setting, significant TR may be present despite an “optimal” LVAD pump speed. However, an excessive LVAD pump speed may precipitate acute severe RV dysfunction with acute severe TR. When the LVAD pump speed is set too high, the left ventricle may become small (“sucked down” or “over-decompressed”), resulting in an abnormal RV-to-LV septal shift that causes distortion of the RV geometry, including the tricuspid valve annulus; this alteration precipitates or worsens TR, which, in turn, causes or exacerbates RV dysfunction. The cascade of events resulting from an excessive LVAD pump speed may ultimately result in a “suction event,” a condition in which a segment of the LV myocardium partially occludes the inflow cannula and reduces pump inflow. Suction events, along with the noted high-risk findings, can be quickly corrected by lowering the pump speed ( Figure 10 ). Suction events can be related to other causes of reduced LV preload relative to the pump speed setting. Accordingly, rapid assessment of AV opening, the relative LV and RV sizes, degree of TR, ventricular septal position, inflow-cannula position, and flow velocities is recommended after initiation of LVAD support and after changes in the LVAD pump speed. It is important that the updated pump speed is always reannotated on the screen during the course of the perioperative TEE exam. A suction event that occurs at a relatively low pump speed or ongoing severe RV dysfunction at low levels of LVAD support is an ominous sign that may indicate the need for a return to CPB or for biventricular support. Suction events can be related to other causes of a reduced LV preload (eg, hypovolemia) or a low afterload (eg, sepsis) relative to the pump speed setting.
Inflow Cannula and Outflow Graft
Inflow Cannula
An appropriately positioned inflow cannula lies near or within the LV apex and is directed towards the MV, although some angulation towards the ventricular septum may be observed ( Figure 5 ). Assessment of the relationship of the inflow cannula to the ventricular septum is generally performed by using standard planar mid-esophageal LV views. Additional simultaneous orthogonal planes and real-time 3D imaging may be used to better identify the terminal portion of the cannula within the LV cavity ( Figures 4A,B and 5 A). Although a certain degree of inflow-cannula deviation towards the interventricular septum may be unavoidable, an excessive degree of angulation may necessitate surgical revision, given an expected decrease in LV cavity size either acutely or later in the clinical course after initiation of MCS. The combination of a smaller LV cavity and an angulated cannula can result in direct contact between the inflow cannula and the septum, which, in turn, can cause ventricular arrhythmias and/or inflow-cannula flow obstruction, as previously discussed. Additionally, the inflow cannula may directly interfere with the native submitral apparatus, and this finding should be communicated to the surgeon. Color Doppler interrogation of a properly aligned HM-II inflow cannula should reveal low-velocity (typically ≤1.5 m/sec) laminar, unidirectional flow from the ventricle to the inflow cannula, with a variable degree of uniform systolic augmentation and no regurgitation ( Figure 5 B). In some cases, the normal inflow-cannula spectral Doppler flow signal may be “contaminated” by mitral inflow and/or AR ( Figure 5 C). Using both pulsed and CW spectral Doppler for interrogating the HM-II inflow-cannula flow is recommended, in order to screen for obstructive velocities ( Figure 5 C). Any HM-II inflow cannula turbulent color Doppler or significant peak systolic velocity variability suggests the presence of mechanical obstruction by the interventricular septum, LV muscular trabeculations, or submitral apparatus. The pericardial location of the HVAD impeller results in a prominent, characteristic color and a spectral Doppler artifact that generally precludes Doppler interrogation of the inflow cannula. The HVAD Doppler artifact occurs only when the inflow cannula appears within the imaging sector. Therefore, successful color and spectral Doppler interrogation of other cardiac structures is possible whenever the imaging plane excludes the HVAD inflow cannula ( Figure 11 ). Consequently, HVAD inflow must be determined indirectly by correlating the inflow cannula anatomic imaging (ie, does the cannula appear unobstructed?) with downstream anatomic and hemodynamic parameters, as discussed in more detail below.
Outflow Graft
After interrogation of the inflow cannula, attention should be directed towards the outflow graft. Whereas the proximal outflow graft is not visible with TEE, the middle portion adjacent to the right side of the heart ( Figure 12 ) and the distal outflow graft-to-aorta anastomosis can be visualized in the majority of patients. Flow from the outflow graft into the aorta can be visualized by color Doppler interrogation near the level of the right pulmonary artery (eg, great vessel, upper esophageal view [ Figures 6 and 13 ]). Simultaneous orthogonal-plane or real-time 3D imaging may allow better characterization of the anastomosis site. Every effort should be made to perform spectral Doppler interrogation coaxially to the direction of flow. As with the inflow cannula described above, the spectral Doppler appearance should consist of low-velocity, laminar, unidirectional flow with a variable amount of systolic augmentation. However, outflow-graft-velocity benchmarks are not available. The peak systolic and nadir diastolic Doppler-derived velocities vary with pump speed in the same patient, and these speeds may also vary with the graft cross-sectional area of the particular device type. However, an outflow-graft peak systolic velocity of >2 m/s at any level (including the that of the aortic anastomosis) may be abnormal and warrant further investigation or monitoring.
Finally, it is important to note that sternal closure can change the orientation of either the inflow cannula or the outflow graft relative to their open chest positions. Accordingly, it is critical to reevaluate the inflow cannula orientation and flow characteristics and the outflow graft and/or outflow graft-to-aorta anastomosis flow immediately after sternal closure. This can be accomplished by TEE or TTE.
Pump Speed
Optimal pump speed selection is a complex topic. The early postimplantation recovery phase may be associated with significant fluctuations in LV preload and afterload. Therefore, the immediate postimplantation (operating room) pump speed that is associated with ‘normal’ LVAD function by the perioperative TEE parameters discussed above may or may not be appropriate later on. In addition (as discussed in more detail, below), selection of an “optimal” LVAD speed setting varies among implantation centers. Some centers select the speed that minimizes LVEDVs and/or the LVIDd while allowing at least intermittent AV opening (assessed best by M-mode echocardiography at the AV level). Other centers maximize LV unloading, leaving the AV closed.
Key Points
- •
Intracardiac air is a consequence of LVAD implantation, and TEE evaluation is useful for ascertaining the success of de-airing maneuvers.
- •
All images acquired after LVAD activation should be annotated with the device name and current pump speed.
- •
Postimplant perioperative TEE should include rapid assessment for possible unmasked PFO shunt, AV opening, the relative LV and RV sizes, degree of TR, ventricular septal position, inflow-cannula position, and flow velocities after initiation of LVAD support and after changes in the LVAD pump speed.
- •
A “suction event,” is a condition in which a segment of LV myocardium partially occludes the inflow cannula and reduces pump inflow. This complication is usually related to over-pumping of the left ventricle (producing a small “sucked down” LV cavity). Suction events can often be quickly corrected by lowering the pump speed.
- •
HM-II inflow cannula peak systolic flow velocities are typically <1.5 m/sec. Higher velocities suggest possible inflow-cannula obstruction.
- •
HVAD inflow-cannula velocities cannot be measured due to a characteristic Doppler artifact.
- •
TEE imaging can frequently show the anatomic contour and flow velocities of the distal outflow-graft region and the outflow-graft–to–aorta anastomosis.
- •
Outflow-graft velocities of >2 m/s at any level may be abnormal and warrant further consideration for possible obstruction, although benchmark data are lacking in this regard.
Role of Echocardiography (TTE or TEE) After LVAD Implantation
The significant variability in the clinical courses of individual patients after LVAD implantation precludes a “one-size-fits-all” approach to postimplantation echocardiography. Nevertheless, the authors believe that an overall framework can be recommended. In general, the starting point for any LVAD echocardiographic examination is a comprehensive “HF” TTE exam, which is performed at the pump’s baseline speed setting and includes LVAD-specific views and Doppler flow assessments in addition to all the elements of preoperative TTE. In some cases, outlined below, the exam also includes the systematic reacquisition of selected exam components at pump speeds above and/or below the baseline speed. The exact protocol for changing pump speeds varies, depending on the indication for examination. There are three subcategories of LVAD echo protocol indications that appear to reflect real-world clinical management:
- 1.
LVAD surveillance echocardiography, with or without LVAD optimization echocardiography.
- 2.
LVAD problem-focused echocardiography, with or without an LVAD speed-change protocol.
- 3.
LVAD recovery echocardiography.
LVAD Surveillance Echocardiography
LVAD surveillance echocardiography is performed at the pump’s baseline speed setting and includes LVAD-specific views and Doppler flow assessments in addition to all the elements of a standard HF TTE exam. Addition of an LVAD optimization protocol , may involve further limited imaging at pump speeds higher and/or lower than the baseline speed to optimize LVAD and native heart function.
The authors recommend that patients with an uncomplicated postoperative course (eg, absence of HF symptoms, successful weaning from IV pharmacologic inotropic and vasopressor agents within 14 days, absence of LVAD controller alarms, and lack of serologic evidence of hemolysis or infection) undergo follow-up surveillance TTE at prespecified intervals. Periodic LVAD surveillance echo exams are recommended, to establish patient-specific “baseline” parameters for both LVAD and native heart function. An LVAD surveillance echo exam should be considered at approximately 2 weeks after device implantation or before index hospitalization discharge (whichever occurs first), followed by consideration of surveillance TTE at 1, 3, 6, and 12 months post implantation and every 6 to 12 months thereafter. Figure 14 summarizes a sample schedule for timing postimplantation surveillance TTE. Comparison of serial surveillance-exam results to each other (for an individual patient) or to population-based benchmarks (see Appendix D ) can also help the examiner understand a patient’s response to LVAD therapy over time. Moreover, surveillance data may allow early diagnosis of occult native heart abnormalities (eg, development of LVAD-related AR) or other device-related problems, including a drift from previously optimal device speed settings. When surveillance TTE is coordinated with the patient’s routine LVAD clinic visits, HF specialists can integrate the information obtained into their clinical assessments and care plans. A putative benefit of routine LVAD surveillance echocardiograms (with optimization protocols when indicated) is improved patient outcomes, including early detection and treatment of complications and reduced hospitalizations for recurrent HF.
Key Points
- •
Patients with an uncomplicated postoperative course should undergo LVAD surveillance echocardiography at certain predetermined intervals after LVAD implantation to assess the patients’ response to MCS therapy and to screen for the development of subclinical complications.
- •
When possible, LVAD surveillance echocardiography should be coordinated with routine LVAD-clinic visits.
Clinical Data-Acquisition Standards and Sonographer Reproducibility (see Table 3 )
Before initiating any LVAD echo exam, sonographers should always annotate the LVAD type and baseline LVAD speeds in rotations per minute (rpm) on the imaging screen in addition to the standard patient demographic data. If the device speed is changed, this should be reannotated during the exam. The device type and speed information should also be routinely incorporated into reporting templates.
| √ | Sonographer Checklist / Ordering Worksheet |
|---|---|
| Study Type being ordered | |
| • Surveillance, initial (+/− optimization, pre/discharge) | |
| • Surveillance, post-discharge (+/− optimization, number months post: 1, 3, 6, 12, 18, etc.) | |
| • Problem-solving at baseline speed only | |
| • Problem-solving at baseline + other speed settings | |
| • Recovery | |
| Ordering/responsible physician identified | |
| Implant date documented | |
| Symptoms noted (if applicable) | |
| Device alarms: if present, type of alarm identified | |
| Other key clinical history/information related to indication noted | |
| Anticoagulation therapy adequate if low pump speeds tested | |
| LVAD name noted on worksheet and annotated on screen | |
| LVAD speeds (baseline and changes) noted on worksheet and annotated on screen | |
| Blood pressure (cuff or Doppler) noted on worksheet and annotated on screen (obtained by designated trained individual at time of exam) | |
| Designated person to change pump speed available | |
| Supervision: appropriate staff to perform speed changes; safety endpoint recognition (eg, low flow, suction event, hypo/hypertension) | |
| Aortic Root Thrombus detection: reason not to proceed (lowering speed could open AV) | |
|
Blood Pressure
The patient’s BP, which reflects peripheral vascular resistance, is an important parameter that greatly influences ventricular unloading and the observed echocardiographic findings. Therefore, the BP should be recorded just before the exam and immediately afterward if pump speed changes were made. Patients with CF-LVADs have a reduced and narrowed pulse pressure, and a palpable pulse may be absent. Therefore, cuff-based BP assessment may be difficult or impossible to perform. In the intensive care unit, the BP may be obtained from invasive arterial monitoring devices. In other settings in which no pulse is present, the use of a BP cuff along with handheld audible Doppler evaluation of the brachial or radial artery may be required. Note that the arterial Doppler-derived BP reading lies between the systolic pressure and the mean arterial BP. For practical purposes, if the patient has a pulse (ie, the AV is opening), the Doppler-derived BP is the same as the systolic BP. If the patient does not have a pulse (ie, the AV is not opening), the Doppler BP is considered to be the mean arterial BP. A current BP measurement is necessary for accurate echo interpretation and for safety reasons during “speed change” protocols, particularly when changing to higher speed settings. Susceptible patients may develop clinically significant hypertension in response to increased LVAD flow, and a mean arterial pressure of <85 mmHg is recommended. Hypotension is generally defined as a mean arterial pressure of <60 mmHg and may be associated with traditional symptoms and/or signs reflective of hypoperfusion. With CF-LVADs, one of the challenges is that a sonographer (or some other trained and available individual) needs to be facile at obtaining an arterial Doppler-derived BP reading. To facilitate the care of CF-LVAD recipients, there may be a need for improved BP monitoring techniques.
Key Points
- •
Although BP readings can be challenging to obtain in LVAD patients, this variable is important, as it significantly influences observed echo findings and their interpretation.
- •
In the absence of a palpable pulse, BP measurement may require audible Doppler interrogation by an appropriately trained individual before the echo exam.
- •
Susceptible patients can experience marked hypertension after the LVAD pump speed is increased. Therefore, the BP measurement should be repeated after a significant pump-speed increase, particularly if the BP is elevated at the baseline pump speed.
- •
A mean arterial BP of <85 mm Hg is recommended.
- •
Hypotension is generally defined as a mean arterial pressure <60 mmHg. It may be associated with traditional symptoms and/or signs of hypoperfusion.
LV Size and Systolic Function
Methods for determining LV size and systolic function by using linear and volumetric approaches in non-LVAD patients have been described by Lang and colleagues.
LV Size
As mentioned above, the LVIDd from the 2D parasternal long-axis image is considered the most reproducible measure of LV size after LVAD implantation ( Figure 15 A). In the presence of a normally functioning CF-LVAD, severely depressed native LV function, and altered MV opening, determination of end-diastole may be difficult. In this scenario, correlating the images to the electrocardiographic signal can be helpful. Additionally, strong consideration should be given to the use of a microbubble contrast agent when endocardial definition is insufficient for accurate LVIDd measurement. Previous data from HM-II outpatients in stable condition suggest that at least a 15% reduction in the LVIDd compared to preimplant values can be expected 3 months after implantation. Care must be taken to correlate LV end-systolic versus end-diastolic diameters with the electrocardiographic signal. The LVIDd may be paradoxically smaller than the LVIDs, and this is an important finding, as it is associated with excessive LVAD unloading and/or severe RV dysfunction. Although LV volumes, as determined by Simpson’s biplane or single-plane method ( Figure 16 ), reflect the LV size more accurately than do linear measurements, the LV size by volume assessment may be technically challenging to obtain after LVAD implantation because of apical shadowing/dropout associated with the inflow cannula. This is one reason why postimplantation LV volumes assessed by echocardiography are smaller than those assessed by CCT. A reasonable LV diastolic volume assessment is possible in many ambulatory LVAD patients, and this metric can be incorporated into the surveillance exam, particularly at the baseline pump speed setting. However, LVIDd measurement, being more expediently acquired and reproducible, is practical for tracking the relative LV size over time at a baseline pump speed (eg, Figure 15 A vs. 15 B) and in the context of a speed-change exam (see below) for quick problem solving. That the serial LVIDd measurement (combined with the degree of AV opening) can be used as a surrogate marker for the degree of LV unloading in CF-LVAD patients seems intuitive and is supported by limited available literature, which is derived primarily from HM-II studies. However, robust outcomes data are limited, and applicability to HVAD patients, for whom there is less evidence, has not been demonstrated at this time.
LV Systolic Function
Accurate determination of LV volumes is challenging after device implantation. So, too, is accurate and meaningful determination of overall LV systolic function, as based on the LVEF. Limitations for LVEF measurements are both technical with regards to imaging quality (endocardial border detection) and physiologic. The LV endocardium may be difficult to visualize because of apical foreshortening, apical shadowing from the device or acoustic dropout (signal attenuation). LVAD-related physiologic challenges include enhanced interventricular dependence and discordant septal and inferolateral wall motion, which may vary considerably in the same patient at different pump speeds. If the LV endocardium, including the apex, can be adequately visualized, with or without a microbubble contrast agent, the preferred method for calculating the LVEF is the biplane method of disks ( Figure 16 ), modified Simpson rule). Although other parameters for LV systolic function may be considered, the LVEF is an important surrogate for showing possible LV worsening or recovery. Therefore, surveillance and recovery LVAD exam reports should include an LVEF assessment, even if only a qualitative assessment is possible. However, LVAD support markedly reduces LV preload, an important determinate of LVEF. Therefore, the value of LVEF for determining systolic function during LVAD support must be taken into consideration during clinical decision-making.
Other methods: In patients with suboptimal apical but adequate parasternal views, the following methods for measuring LV systolic function may be considered, although their accuracy has not been validated in LVAD patients.
- 1.
The LV fractional area change (FAC) method at the mid-papillary muscle level on 2D short-axis views: FAC (%) = [(end-diastolic area – end-systolic area)/(end-diastolic area)].
- 2.
The Quinones method for determining the LVEF, with the assumption of an akinetic apex given the presence of the apical inflow cannula.
- 3.
The LV fractional shortening (%) method: FS = [(LVIDd-LVIDs)/(LVIDd)], where FS = fractional shortening and LVIDs = the LV internal dimension at end-systole, which has been applied in LVAD patients.
The linear and volume measurements of systolic function noted above represent possible methods for tracking the course of individual patients, serving as their own controls, over time. However, routine use of methods 1 to 3, above may not be feasible or recommended for many LVAD patients because of segmental wall-motion abnormalities, exaggerated paradoxical septal motion, ventricular dyssynergy and/or ventricular septal shift, the extent of which could change at varying pump speeds in the same patient. Note that methods of calculating the LVEF based on the LV stroke volume are not recommended, because many LVAD patients have beat-to-beat variations in this parameter. Previous data suggest that the vast majority of outpatient HM-II recipients in stable condition have persistent moderately to severely depressed LV systolic function during the first 6 months after device implantation.
Key Points
- •
After CF-LVAD activation, the LVIDd may be the most reproducible measure of LV unloading that can be tracked over time and/or at different pump speeds.
- •
The LVEDV is a more accurate representation of LV size than is the LVIDd.
- •
After LVAD implantation, measurement of LV volumes and the LVEF can be technically challenging. When the LVEF needs to be obtained (particularly to assess for LV recovery), Simpson’s biplane method of disks is recommended for use when possible.
LV Diastolic Function
It is assumed that LVAD patients have markedly abnormal baseline diastolic function. Although the standard LV diastolic function parameters can be measured and included in the report, there is a paucity of data validating their clinical usefulness in the setting of LVAD support. The use of certain diastolic parameters could be helpful, particularly when correlated with symptoms in individual patients, and at the discretion of the interpreter since they may reflect changes in the degree of LV unloading when compared to a prior study’s data or at different pump speeds during the same exam. Previous data suggest that the mitral E velocity (cm/s), left atrial volume (mL), pulmonary vascular resistance (Wood units), and pulmonary artery systolic pressure (mmHg) are significantly reduced and that the mitral deceleration time (ms) is significantly prolonged in outpatients whose condition is stable 3 to 6 months after HM-II implantation. How these parameters should be integrated into postimplantation clinical management is currently undefined, as is their prognostic value for patient outcomes. For a clinical LVAD echo reporting purposes, a practical approach at this time may be to use the following (or a similar) statement: “Interpretation of the degree of LV diastolic dysfunction (presumed abnormal) is not provided because of continuous flow LVAD support.”
Key Points
- •
It may be assumed that LVAD patients have markedly abnormal baseline diastolic function.
- •
How LV diastolic parameters measured after LVAD implantation should be integrated into the echocardiography interpretation and clinical management is currently undefined, as is their prognostic value for patient outcomes.
RV Size and Systolic Function
Many of the standard measures of RV size and systolic function, including linear dimensions, RV FAC, TAPSE, and right-sided cardiac output, can feasibly be measured in LVAD patients. However, recent data suggest that the correlation of TAPSE with overall RV systolic function may be weaker after cardiothoracic surgery and, therefore, this variable may have less clinical utility than the other measures. Current data regarding the expected response of RV systolic function after LVAD implantation are conflicting: one study showed a significant improvement in RV FAC at 3 months, but another study did not show a significant difference in this parameter at either 1 month or 6 months.
Valvular Assessment
Aortic Valve
Evaluating and reporting the degree of AV opening (if any) is important because it is affected by a number of other parameters, including LVAD speed, native LV function, volume status, and peripheral vascular resistance. In addition, whether or not the AV opens may have clinical implications. Whereas recent guidelines recommend that the LVAD speed be set low enough to allow at least intermittent AV opening, such opening may not occur at any LVAD speed in patients with extremely poor native LV function. The frequency of AV opening is most accurately assessed by recording multiple (five to six) cardiac cycles at a slow M-mode sweep speed (eg, 25-50 mm/s) ( Figure 8 D,E); the valve should be characterized as either opening with every cardiac cycle, opening intermittently, or remaining closed. Many HF teams also request that the duration of AV opening (ms) be measured from the same M-mode acquisitions. This parameter may vary from beat to beat, so it is best to measure several beats and report an average value. When the AV-opening duration is relatively constant, a faster sweep speed (eg, 75-100 mm/s) may be appropriate ( Figure 8 A,B). An important potential pitfall of using M-mode to assess the presence and duration of aortic cusp separation is illustrated in Figure 17 . The AV semilunar cusp conformation, combined with cardiac translational motion or slightly off-axis imaging, can create the false appearance of aortic cusp separation, even when the cusps are not separating. Careful attention and the additional use of color M-mode may be useful in difficult cases to avoid M-mode “pseudo AV opening” or an exaggerated AV-opening duration. However, an additional interesting observation is that in some cases of “minimal” AV opening, the duration of AV cusp separation and duration of forward systolic flow are not always the same, and color M-mode can help to document this finding ( Figure 9 G,H). In patients whose AV remains closed, it is important to evaluate for aortic root thrombus, which may be transient or associated with commissural fusion. Continuously closed aortic cusps have been associated with the development of aortic root thrombosis and LVAD-associated AR, as discussed below. Fusion of the aortic cusps, either surgical or secondary to chronic aortic cusp closure, can be recognized on speed-change echocardiograms (discussed below).
New-onset (“de novo”) AR occurs in approximately 25% to 33% of patients 12 months after LVAD implantation and is a key finding, given its known adverse effects on LVAD performance, morbidity, and mortality. Several studies suggest that persistent AV closure is a risk factor for de novo AR after LVAD implantation, even without the presence of aortic root thrombus ( Figure 18 ). For the reasons noted above in the postimplant TEE section, standard methods for quantifying AR may be challenging to use after LVAD implantation. In the absence of definitive cutoff criteria to define mild, moderate, and severe AR after LVAD implantation, one should perform an aggregate assessment based on duration (predominantly diastolic vs. continuous), AR jet VC width, jet height relative to the LVOT, comparative LVAD and native circuit flow measures and LV chamber size. Additionally, significant AR noted on LVAD surveillance echocardiography may be further evaluated with device controller data and the cardiac response during LVAD problem-focused echocardiography with speed changes, as described below.
Key Points
- •
Recording multiple cardiac cycles with color M-mode at a sweep speed of 25-50 mm/s is recommended to accurately assess the frequency and duration of AV opening.
- •
Persistent AV closure can be associated with aortic root thrombus and de novo AR.
- •
If aortic root thrombus is suspected, a decrease in the LVAD pump speed should be avoided, as it could result in sudden AV opening (eg, during a planned speed-change exam).
- •
After LVAD implantation, the presence of AR is not uncommon. Assessment of severity is partly based on careful color Doppler analysis in the parasternal long-axis view.
Mitral Valve
As noted above, LV unloading generally leads to reduced MV annular dilatation, improved leaflet coaptation, and, ultimately, reduced MR severity. Persistence of significant MR after initiation of LVAD support may indicate inadequate LV unloading or inflow cannula malposition and interference with the submitral apparatus. If MR is present, it can be quantified by using standard methods. Incidental post-LVAD MR may also represent LVAD malfunction and should be discussed with the clinical team.
Tricuspid and Pulmonary Valves
Like MR, moderate or greater TR is an important finding on LVAD surveillance echocardiography, as this condition may be associated with insufficient LV unloading (functional TR), excessive LV unloading with a leftward shift of the interventricular septum (eg, a suction event), elevated systolic pulmonary pressures, and/or intrinsic RV systolic dysfunction. Distinguishing between these etiologies by utilizing echocardiographic parameters is discussed in further detail below. Regardless of the etiology, TR after LVAD implantation can generally be assessed with standard methods. Furthermore, the native pulmonary valve typically remains functionally normal after LVAD implantation and can be interrogated by using standard methods when significant stenosis or regurgitation is suspected. As noted in the foregoing discussion of perioperative TEE, the presence of significant preexisting or acquired PR may have implications with regards to RV function and/or the ability to perform RVAD implantation if needed.
Interventricular Septal Position
The end-diastolic interventricular septal position, which is dependent on the interventricular pressure gradient, should be routinely reported as neutral, leftward-shifted, or rightward-shifted . A leftward shift can be due to elevated RV end-diastolic pressures, reduced LV preload, or LV over-decompression resulting from excessive LVAD speed; differentiation of these etiologies is further discussed below. A rightward shift is generally due to elevated LV end-diastolic pressures resulting from an inadequate LVAD speed setting, pump dysfunction, severe AR, or an increased LV afterload.
Inflow-Cannula and Outflow-Graft Interrogation
Inflow Cannula
Usually, the apically inserted inflow cannula can be adequately imaged in standard or modified 2D parasternal and apical TTE views. The sonographer’s objective is to reveal the inflow cannula’s location and orientation in relation to the interventricular septum and other LV structures. The inflow cannula can often be visualized with 3D echo techniques, and this approach may be used as a complementary imaging method by examiners experienced in 3D imaging. As noted above in the section on perioperative TEE, color Doppler interrogation of a properly aligned inflow cannula should reveal laminar, unidirectional flow from the ventricle to the inflow cannula, with no evidence of turbulence or regurgitation. Pulsed and CW spectral Doppler interrogation may require “off-axis” modification of a standard parasternal, apical, or short-axis TTE view to achieve true coaxial alignment between the sampling beam and inflow-cannula flow; such interrogation should additionally reveal the flow to have a low peak velocity (<1.5 m/s). Due to native LV contractility, cannula flow generally remains pulsatile to some degree even when the AV does not open. Recording both the peak systolic and nadir diastolic velocities over at least three to four cardiac cycles is recommended ( Figures 5 and 19 ).
