Mechanical Circulatory Support

By James K Kirklin MD, Francisco Arabía, MD, MBA, Suzanne V Arnold, MD, MHA and 56 more

An update of Volume 1 of the ISHLT Monograph series, this book is not a textbook; rather each chapter tries to focus on specific topics within the field that faced in the filed of mechanical circulatory support, often on a daily basis. Many of these topics have only evolved in recent years, which makes the content more timely and of interest to the reader.

• Language: English
• Publisher: BookBaby
• Release date: Jun 16, 2021
• ISBN: 9781098385880

Book preview

Grady

CHAPTER 1

PATIENT SELECTION, TIMING, AND INDICATIONS FOR

DURABLE LVAD IMPLANTATION

Contributors: Garrick C Stewart, Finn Gustafsson

Introduction

When selecting patients for durable left ventricular assist device (LVAD) therapy, the considerable benefits of treating advanced heart failure (HF) must be weighed against significant risks and considerable costs. There is an inherent dilemma in selecting the right patients for the right device at the right time. If patients are implanted too early, outcomes after implantation may be good, but the net benefit is small since outcomes would also have been good with ongoing medical and electrical therapy. Conversely, if the device is implanted too late, outcomes may be poor, with or without device therapy (Figure 1).

Selection of LVAD candidates: overall principles

Figure 1. Overall Principles in the Selection of LVAD Candidates. Estimates of survival with (dashed green line) or without (dashed red line) LVAD therapy can provide insight into the appropriateness of device candidacy. (Reproduced with permission Lars Lund, MD, personal communication)

Consequently, selection and timing for LVAD implant first requires accurate estimation of the patient’s prognosis in the absence of advanced therapies like LVAD or transplantation. Multiple studies describe prognostic factors in groups of patients (or in an average patient in that population) and may help identify relevant risk factors. However, the incremental impact of a given factor (e.g. renal dysfunction) on an individual patient within their broader risk profile is often not well understood. Studies addressing an individual patient’s prognosis with HF are less prevalent and often come with their own limitations. The selection of the optimal candidate for LVAD therapy must begin with the identification of patients on high risk heart failure trajectories who may be a candidate for an evaluation for mechanical circulatory support (MCS).

Clinical Presentation

Hospitalization for acute decompensated HF (ADHF) is an extremely important marker of risk, as it identifies a highly vulnerable period in the patient’s journey with HF. Data from the United States reveal that mortality is 3-4 % in-hospital during an admission with HF, 10-12 % at 30 days following discharge, and 30-40 % after one year¹. Rehospitalization rates are high and generally reported to be 45-65% within the first year¹,². Among patients with ADHF, the manner of clinical presentation may further identify patients at high risk. At one end of the spectrum, patients admitted with cardiogenic shock after myocardial infarction (MI) have a 30-day mortality greater than 40%³. In patients without shock, which represents the vast majority, admission profiles associated with risk markers include acute MI, myocardial ischemia as evidenced by abnormal ECG or elevated troponin T or I,⁴ low systolic blood pressure, renal insufficiency, increased heart rate, hyponatremia, reduced ejection fraction, increased B-type natriuretic peptide (BNP) or N-terminal (NT)-proBNP, older age, and presence of comorbidities⁵,⁶. Patients who require intravenous vasodilators or inotropic support have a particularly poor prognosis⁷. The need for inotropic support was associated with an in-hospital mortality rate of 26% in the ALARM-HF (Acute Heart Failure Global Survey of Standard Treatment) registry⁸. Hypertensive patients (systolic blood pressure>160 mmHg) generally have the best prognosis with a low 60-day mortality.

In stable outpatients, the most important prognostic factors are patient age and symptoms, including the New York Heart Association (NYHA) functional class. In addition, a large number of laboratory, hemodynamic, and echocardiographic measures have been examined and shown to correlate with long term survival and hospitalization risk⁴,⁹,¹⁰ (Table 1).

Table 1. Predictors of Survival in Heart Failure

ANP, atrial natriuretic peptide; BUN, blood urea nitrogen; BNP, B-type natriuretic peptide; eGFR, estimated glomerular filtration rate; NYHA, New York Heart Association Class; PCWP, pulmonary capillary wedge pressure; RAAS, renin angiotensin aldosterone system; VO2, oxygen consumption.

Interestingly, simple clinical bedside examination can still provide prognostic information even in contemporary outpatient populations such as those included in the PARADIGM-HF (Prospective Comparison of ARNI with ACEI to Determine Impact on Global Mortality and Morbidity in Heart Failure) trial. In this study, jugular venous distention, edema, rales, and third heart sound were found to provide prognostic information independent of symptoms and natriuretic peptide levels¹¹. Left ventricular systolic function as measured by ejection fraction (LVEF) is an important predictor of outcome. It has been well documented that patients with HF and preserved ejection fraction have a better prognosis compared with patients with reduced LVEF (HFrEF)¹². However, in patients with HFrEF, the degree of impairment in LVEF, a load dependent marker of cardiac performance, may be a less robust marker of outcome. Also, it is important to realize that the inter- and intra-observer variability in LVEF using echocardiography in clinical practice is approximately 10% (absolute values). Hence, serial measurements must be interpreted with caution and a measured decline in LVEF from 25 to 15 % may not, by itself, represent a major prognostic change.

INTERMACS Classification

The Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) classification describes characteristics of patients with advanced HF, both hospitalized and ambulatory. Patients with advanced HF are categorized among 7 different profiles with specific focus on potential timing for advanced therapies, especially MCS (Table 2)¹³.

Table 2. INTERMACS Patient Profiles

INTERMACS profiles 1-3 patients are inotrope dependent and profiles 1-2 are often grouped together as they may be difficult to differentiate, especially if temporary MCS is used. The INTERMACS classification correlates with outcome of patients with advanced HF, even if they are ultimately not managed by MCS. In a study of ambulatory, non-inotrope dependent patients with advanced HF, 12-month overall survival was 60, 74, and 84% in patients in INTERMACS 4, 5, and 6/7, respectively (Figure 2)¹⁴.

Figure 2. Survival according to INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support) profile among patients enrolled in the Medical Arm of Mechanically Assisted Circulatory Support (MedaMACS) pilot study. Estimated survival was modeled through 1 year using the Kaplan–Meier method with censoring at either ventricular assist device (VAD) or transplant. Survival was significantly different between patient profile groups (P=0.039). (Reproduced with permission, Stewart et al.¹⁴)

While the INTERMACS classification has emerged as a useful tool for characterizing patients considered for MCS, some caveats should be mentioned. First, there is little documentation for the use of the classification outside of cohorts of patients undergoing consideration for implantation with MCS. Further, there is controversy over how to characterize patients stable on temporary MCS, and some decisive characteristics, for instance, threshold for inotrope dependence, may be interpreted differently in different HF programs.

Biomarkers for Heart Failure Prognosis

Cardiac Biomarkers: Natriuretic Peptides and Troponin

No ideal biomarker has been discovered yet, but of all the prognostic biomarkers currently identified, BNP and NT-proBNP come closest. This measurement entails an easily obtainable, rapidly processed, inexpensive blood test with excellent sensitivity and good specificity for HF. Another valuable biomarker in acute and chronic HF is plasma troponin both in patients with acute coronary syndrome and in patients with non-ischemic acute decompensated HF⁴. It is a class I recommendation to use natriuretic peptides both for diagnosis and prognostication in HF¹⁵. Elevated levels of natriuretic peptides are associated with worse outcomes both in patients with mild and advanced HF. However, while not necessarily linear ¹⁶, the risk function is continuous, and there is no accepted cut off limit for poor prognosis. As such, BNP values, like other biomarkers, must be interpreted in a clinical context. Female gender, low body weight, and atrial fibrillation are associated with higher levels of natriuretic peptides irrespective of cardiac function. Also, renal dysfunction increases BNP/NT-proBNP, and changes should be evaluated on this basis. Of recent relevance, it should be noted that sacubitril/valsartan specifically increases BNP, but not NT-proBNP ¹⁷.

Markers of End-Organ Function and Inflammation

Among patients hospitalized for decompensated HF, more than 60% have moderate or severely reduced renal function, as defined by estimated glomerular filtration rate (GFR) < 60 ml/min.³⁷ Serum creatinine and blood urea nitrogen (BUN) have been identified in numerous studies as powerful prognostic factors in patients with acute and chronic HF, and reduction in eGFR to less than approximately 40 mL/min/1.73 m² is associated with a doubling of mortality ¹⁸. Further, the development of worsening renal function as evidenced by serum creatinine, eGFR, or BUN during acute hospitalization for HF is associated with increased mortality rate. However, a decrease in eGFR during hospitalization for HF cannot be viewed in isolation. Changes induced by angiotensin converting enzyme inhibitors or mineralocorticoid receptor antagonists are not associated with poor outcomes¹⁹ and a decrease in eGFR during an admission for HF is only a marker of poor prognosis if accompanied by lack of decongestion²⁰. Hence, while powerful markers, the use of eGFR, creatinine, and BUN to evaluate prognosis in HF should incorporate medical therapy as well as the congestive state of the patient.

Abnormal liver function tests, such as bilirubin and international normalized ratio, reflect right heart function and correlate especially well with elevated central venous pressure. They are important markers of outcome, especially in patients undergoing LVAD implantation, as they may predict right ventricular failure following implantation ²¹.

Subclinical inflammation is often detectable in HF. Several biomarkers have been identified as indicators of poor prognosis and may reflect the degree of systemic or local inflammation (i.e., C-reactive protein, tumor necrosis factor, and interleukin-6, soluble ST-2) ²².

Physical Capacity and Mortality Risk in Heart Failure

Six-Minute Walk Test

Exercise limitation is a cardinal symptom of chronic HF. Functional capacity may be assessed by NYHA criteria, but the assessment of NYHA class remains highly subjective. The 6-minute walk test (i.e., the distance walked over a period of 6 minutes) is less subjective than NYHA functional class, but is influenced by the motivation of the patient and may be affected by non-cardiopulmonary factors such as arthritis. However, the 6-minute walk test has several advantages: it is easy to perform, reproducible, and inexpensive.

The 6-minute walk test provides prognostic information, for instance, as shown by the Study of Left Ventricular Dysfunction (SOLVD) investigators, demonstrating that mortality risk was almost 4 times higher in patients with a 6-minute walk distance less than 350 m compared with patients who walked more than 450 m ²³. Subsequent investigators have shown prognostic value of the 6-minute walk test in most, but not all, cohorts ²⁴,²⁵. Six-minute walking distance is an important end-point in trials in advanced HF and in pulmonary hypertension.

Maximal Oxygen Uptake and Prognosis in HF

Determination of peak oxygen uptake (peak Vo2) during a maximal treadmill or bicycle exercise test is the most objective method to assess maximal functional capacity in patients with chronic HF. Peak Vo2 has been shown to be an excellent indicator of prognosis in HF. Vo2 during maximal exercise can be calculated from the Fick principle: Peak Vo2 = peak cardiac output x arteriovenous oxygen difference at peak exercise. Because the maximal arteriovenous difference is similar in most sedentary individuals, peak Vo2 provides a good estimate of cardiac output reserve. Indeed, maximal cardiac output is the main limiting factor in healthy individuals as well as HF patients²⁶. This likely underlies the effectiveness of peak Vo2 in risk stratification in HF. Non-cardiac factors may affect peak Vo2, such as the metabolic activity of skeletal muscle mass and endothelial function and demographics such as age and gender. However, these factors also relate to HF severity, and this may add to the prognostic utility of peak Vo2.

To contribute valid prognostic information, measurements of peak Vo2 must be accurate, which requires that tests are performed using standardized protocols and that equipment is correctly calibrated. It is crucial to ensure that the maximal Vo2 is achieved. Patients with comorbidity, such as arthritis or lung disease, may terminate the test before achieving their peak Vo2. The most commonly used variable to document that peak Vo2 has been reached is an elevated respiratory exchange ratio (RER, ratio between expired CO2 and consumed O2). However, one should not rely on RER alone and the Vo2 and VCO2 curves should be inspected to ensure that the RER is representative. The RER defining true peak Vo2 is debated, but>1.05 is generally accepted as sufficient for use of the test result in evaluation of heart transplant candidates²⁷.

The International Society of Heart and Lung Transplantation (ISHLT) guidelines for selecting patients for heart transplantation advocates a cut-off value of 12 mL/Kg/min in patients treated with beta-blockers and 14 mL/Kg/min for those not. Peak Vo2 also plays a major role in the criteria for selecting ambulatory patients for LVAD implantation, as this was an inclusion criterion for entry into landmark LVAD trials. Analysis of peak Vo2 normalized by a predicted maximum based on age, obesity, and gender has been suggested to be useful especially in young patients, whereas others have shown no clear benefit²⁸,²⁹.

An attractive option when peak Vo2 is integrated in the management of HF is use of serial measurements, which enables identification of patients in a low-risk category over time³⁰. Since the initial reports of the value of peak Vo2 in guiding transplant candidate selection, there have been many advances in the treatment of HF. In particular, the use of beta-adrenergic blocking drugs and implantable cardioverter-defibrillators (ICDs) has had a significant impact on long-term survival without significantly improving peak Vo2. Whether Vo2 has retained its predictive power after introduction of beta blockers and ICDs has been repeatedly investigated³¹,³², but peak Vo2 remains a key parameter in the evaluation of candidates for LVADs.

Most studies on use of peak Vo2 were performed in patients with HFrEF, but the response to exercise is not very different in HFpEF³³. Indeed, peak Vo2 is also an independent predictor of mortality in HF with preserved ejection fraction (HFpEF), but more data are required to standardize the use of peak Vo2 for prognostication in HF patients without reduced ejection fraction³⁴,³⁵.

Invasive Hemodynamics

Resting hemodynamic measures (cardiac index, left and right ventricular filling pressures) are surprisingly modest indicators of mortality risk in HF. Several studies have showed that elevated filling pressures (pulmonary capillary wedge pressure and/or right atrial pressure) have predictive power, whereas cardiac index does not³⁶,³⁷. Invasively measured hemodynamic parameters obtained during exercise may offer superior prognostic information, especially in HFpEF, but acquisition is more labor intensive³⁸,³⁹. Invasive hemodynamics are, however, as discussed below, necessary for selection of patients for LVAD implantation.

Frailty

Frailty has recently been introduced in the field of advanced HF⁴⁰. Several definitions of frailty have been used in cardiovascular medicine, but Fried’s phenotype or modifications hereof have been used most consistently in HF studies⁴¹. Fried’s criteria include unintended weight loss (>5 % or 5 kg within last year), exhaustion (self-reported), inactivity (self-reported e.g. less than 2.5 hours per week), slowness (5 m walk test) and grip strength (measured by dynamometer). Frailty is present in 30-60 % of patients hospitalized for HF and patients referred electively for heart transplantation or LVAD evaluation⁴²–⁴⁴. Frailty is associated with mortality in HF⁴⁵, and in the MCS literature, most studies have shown a significant and independent association between frailty and reduced survival after LVAD implantation⁴⁶,⁴⁷. Importantly, however, frailty may be modified by LVAD implantation, implying that documentation of frailty should not always rule out LVAD implantation⁴⁸.

Risk Scores in Heart Failure

Several risk models describing outcome in both in- and outpatients have been published over the last decades. Some are based on clinical parameters available at the bedside, others incorporate specific biomarkers beyond plasma sodium and creatinine. The groups behind some of these models provide on-line calculators of the individual patient’s mortality risk. While some models like the Seattle HF Model (SHFM) or the HF Survival Score (HFSS) are often cited in scientific literature, it is less clear how they are used in clinical practice, especially in advanced HF. The models may yield divergent predictions for the same patient. The models may, however, be useful when considering advanced HF therapies, as they may qualify discussions with patients, their caregivers, as well as other stakeholders in the decision process. For this reason, many transplant- and MCS programs use the scores when patients are presented at the multidisciplinary transplant/LVAD team meeting where treatment decisions are made.

Scores most often used for inpatients are the Acute Decompensated Heart Failure National Registry (ADHERE), Enhanced Feedback for Effective Cardiac Treatment (FFECT) and Get with the Guideline (GWTG) models. These models incorporate different clinical and laboratory values, including troponin and BNP, in an updated version of the GTWG model⁴⁹. A recent analysis compared the performance of the GTWG, EFFECT, and ADHERE models in more than 13,000 patients admitted with HF and documented similar performance with C-statistics values between 0.68-0.70⁵⁰. Potentially, the importance of risk models for outpatients is greater than that of inpatients where the patients can be assessed daily anyway. The two main models used for outpatients are the HFSS and the SHFM. The HFSS was developed in the 1990s, but has been validated more recently. It resonates with the work-up of patients with advanced HF as it incorporates measurement of peak Vo2³¹,⁵¹. The model has traditionally been used to detect patients with low risk where transplant listing could safely be deferred. Similar documentation with respect to use of HFSS for timing of MCS has not been published.

The SHFM was partially derived from the patients in the Prospective Randomized Amlodipine Survival Evaluation (PRAISE) cohort, which included 1125 patients with an LVEF less than 30% and NYHA functional class IIIB and IV symptoms⁵². This model has also been updated with more contemporary interventions and was validated in five independent samples. The model is useful for showing the expected mortality benefit of adding a medication or device to a patient's HF management. In patients presenting to an advanced HF clinic for transplant evaluation,⁵³ the SHFM tends to underestimate mortality in patients with the greatest observed mortality.

Since the publication of the two first important survival models in HF outpatients, HFSS and SHFM, a multitude of models has emerged, including several important models which included international patient data, such as the Gruppo Italiano per lo Studio della Sopravvivenza nell’Insufficienza Cardiaca (GISSI-HF) and the Meta-analysis Global Group in Chronic Heart Failure (MAGICC HF) risk score⁵⁴,⁵⁵. The latter is based on data from 30 trials and registries and has gained widespread use. It can be calculated easily with an online calculator (www.heartfailurerisk.org)⁵⁵. A review from 2014 identified more than 60 published models and concluded that the design of the models was very different, but finally included a few common variables predictive of mortality: age, renal function, blood pressure, blood sodium level, LVEF, sex, natriuretic peptide, NYHA functional class, diabetes, body mass index, and exercise capacity⁵⁶. In a recent analysis the performance of MAGICC HF, GISSI-HF, SHFM, and the Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity (CHARM) was evaluated using European data, and while the models tended to overestimate mortality, the C-statistic was above 0.7 for all models⁵⁷. Whether this is sufficient for use in individual patients is not clear. Obviously, these models cannot stand alone but rather used in concert with the variables discussed above and may be applied for decision support in the discussion of treatment strategies in advanced HF.

Strategies for Durable LVAD Support

Once a patient at high risk for mortality from systolic HF is identified, despite best attempts at guideline directed medical and electrical therapy, attention can be turned to identifying a potential role for LVAD therapy. The evolving role of durable MCS in advanced HF care has been driven by improved technology, enhanced patient selection driven by an evolving appreciation of factors that contribute to perioperative risk, and proactive patient management around implant⁵⁸. LVADs were originally developed in the 1960s to provide short-term support in patients with post-cardiotomy shock unable to wean from cardiopulmonary bypass⁵⁸. As pulsatile pump component technology improved, LVADs began to be used increasingly as a bridge to transplantation in the 1990s to support the failing circulation until a donor heart could become available. The Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure (REMATCH) trial in 2001 demonstrated favorable outcomes of the electrically driven pulsatile HeartMate XVE in patients with NYHA Class IV systolic HF who were ineligible for transplant compared to medical therapy, which mainly consisted of intravenous inotropes⁵⁹. This led to the landmark approval in 2003 of this LVAD for lifelong support, also known as destination therapy, which permanently uncoupled access to a durable MCS device from transplant candidacy. A bridge to recovery strategy has also been defined for patients with transient cardiac injury (e.g., myocarditis, peripartum cardiomyopathy), although recovery to LVAD explant remains relatively rare⁶⁰–⁶².

With the advent of second and third generation continuous flow LVADs, extended durability and reduced adverse event profile have propelled this surgical HF therapy to implant volumes on par with those of cardiac transplantation. Although strategic designations—bridge to transplant, recovery, or destination therapy—have been central to clinical trial design and regulatory approval, it is now recognized that such strategic designations are fluid⁶³. For example, patients intended as destination therapy may have improvement in renal function or pulmonary vascular resistance after support, transforming them into candidates for transplant. In contrast, patients bridging to transplant may develop intercurrent illness or an adverse event such as stroke, making transplant bailout impossible. Traditionally, only transplant listed patients could receive up front biventricular support or total artificial hearts (TAH), although even that restriction is undergoing review. Focus in the field has shifted toward identifying patients on high-risk HF trajectories and delivering appropriate and timely implementation of MCS⁶⁴.

Indications for Durable LVAD Support

Candidacy for durable LVAD therapy has been defined by consensus guidelines as well as payers to include patients with ACC/AHA Stage D HF symptoms that are refractory to guideline directed medical and pacing therapies⁶⁵–⁶⁷. (Table 3)

Table 3. Definitions of Stage D (Refractory) Heart Failure by Major Cardiology Associations

HFA, Heart Failure Association; ESC, European Society of Cardiology; HFSA, Heart Failure Society of America; ACCF, American College of Cardiology Foundation; AHA, American Heart Association; GDMT, guideline directed medical therapy; LVEF, left ventricular ejection fraction; RV, right ventricular; BNP, B-type natriuretic peptide; NT-proBNP, N-terminal pro-B-type natriuretic peptide; LV, left ventricular; 6MWD, six-minute walk distance; pVO2, peak oxygen consumption; HF, heart failure; ACE, angiotensin-converting enzyme; ICD, implantable cardioverter defibrillator.

Patients who are most often considered for current generation MCS typically have manifestations of cardiogenic shock, defined by hypotension (systolic blood pressure <90mmHg), cardiac index <1.8 L/min/M², and pulmonary capillary wedge pressure>15mmHg along with evidence of end-organ hypoperfusion. Chronic HF patients with Stage D disease are also considered if they have objectively reduced functional capacity as measured by peakVo2 <12-14 ml/Kg².

Referral of patients flagged as high risk for mortality with HF to advanced HF/transplant programs remains essential to deliver timely support using shared decision making prior to a stage of illness where MCS would be futile⁶⁸. Early referral also allows for identification and relief of exacerbating factors, tailored hemodynamic therapy, and upstream multidisciplinary assessment of candidacy for advanced surgical therapies by experienced advanced HF teams. Conversations about MCS timing are often framed by the INTERMACS profiles, a validated shorthand to describe advanced HF severity and disease trajectory, which in turn can inform outcomes both with MCS and ongoing medical therapy⁶⁹–⁷².

Timing of Implantation

Identifying the optimal time to implant MCS remains a central clinical challenge in caring for patients with advanced HF. Early generation devices were implanted only in high risk patients with frank cardiogenic shock despite escalating inotropes or temporary MCS who had few alternative therapies. In such patients, the magnitude of benefit with an LVAD was profound, although longer-term outcomes remained poor due to congestion and metabolic derangements in patients with evolving shock undergoing cardiac surgery. Recognition of these relatively poor long-term outcomes in crash and burn patients compared to patients stabilized with inotropes led to a shift away from INTERMACS Profile 1 toward Profiles 2-3, which remain the predominant profiles for contemporary implant. Patients on inotropes in Profiles 2-3 represent 70% of device implants in the INTERMACS registry of approved devices in the U.S. in between 2012-17 and 64% in the IMACS registry in 2013-14⁷³,⁷⁴. Profile 1 patients in shock are now widely receiving temporary support with ECMO or percutaneous VADs to stabilize end-organ function and define eligibility for LVAD or transplant. (Figure 3)

Figure 3. Optimal timing for mechanical circulatory support (MCS). NYHA indicates New York Heart Association. (Reproduced with permission Peura et al.¹⁰⁶)

Patients on oral therapy with resting symptoms in Profile 4 are also widely accepted to benefit from MCS. Yet, implant rates in the ambulatory INTERMACS patient profiles remain low, with 13% in Profile 4, and just 3% in Profiles 5-7⁷³. With the advent of continuous flow LVADs, interest has expanded toward implanting patients with ambulatory advanced HF, in whom perioperative risk should be lower and post-implant recovery faster⁷⁵. Expansion into the ambulatory profiles remains constrained by our understanding of HF prognosis on contemporary therapies. Both patient and physician perception of HF illness severity, risk of death, and need for advanced therapies in ambulatory HF are frequently underestimated, leading to delays in acceptance of therapy¹⁴,⁷⁶. Additionally, residual adverse events on device therapy, including stroke, may raise thresholds to implant until a greater symptom burden in later stage disease justifies the perceived risk of post-implant complications⁷⁷.

For patients in INTERMACS profiles 5-7 on optimal medical therapy, it will be difficult to distinguish outcomes with medical and LVAD therapy based on survival alone⁷⁸. This has spurred interest in developing combined outcomes of survival with good quality of life, improved functional capacity, or days alive outside the hospital as ways of integrating the patient-reported outcomes that will be essential for defining the appropriateness of using this costly therapy for ambulatory patients⁷⁹. The line between a premature implant and ideal implant timing will continue to move to earlier stage HF with improving adverse event rates and a better understanding of prognosis in advanced HF. These factors highlight the importance of an annual HF review, as outlined in a recent AHA Scientific Statement on decision making in advanced HF, to estimate prognosis and begin the conversation about MCS with patients while implant can be elective⁸⁰.

Evaluation for Durable LVAD Candidacy

The leading causes of death after durable MCS continue to be stroke, infection, bleeding, and right-sided HF, eventually leading to multi-organ system failure. Careful attention to cardiac and non-cardiac comorbidities is required to identify areas for pre-implant medical optimization, to plan concurrent surgery for residual structural heart disease, and to flag patients as likely too sick to derive benefit from MCS due to high anticipated peri-operative mortality. (Table 4) A thorough evaluation can also identify patients with reversible cardiac defects likely to recover without support, with functional limitations due to non-cardiac disease (e.g., oxygen dependent chronic obstructive pulmonary disease), and can define anatomical considerations for implant approach and device selection⁸¹. Although the evaluation for durable MCS parallels that of cardiac transplantation and similarly involves a multidisciplinary heart team, decisions about MCS candidacy are not constrained by the limited donor supply that dominates heart allocation. (Figure 4)

Table 4. ISHLT Selection Guidelines for Durable Left Ventricular Assist Devices

Inclusion criteria

• AHA Stage D heart failure

• Vo2 max < 14 ml/Kg/min or <50% predicted attainment of respiratory anaerobic threshold

• NYHA functional class IIIB/IV for at least 45 of the last 60 days, despite use of maximally tolerated doses of drugs. Inability to tolerate neurohormonal antagonist medications (e.g., beta-adrenergic blockers) may lead to earlier consideration.

Exclusion criteria

• Reversible cardiac dysfunction

• Active uncontrolled coagulopathy

• Inability to tolerate anticoagulation mandated for the LVAD

• Active uncontrolled infection

• Renal disease that would significantly shorten life expectancy, including irreversible dysfunction not explained by HF; chronic hemodialysis is relatively incompatible with durable LVAD

• Hepatic disease that would shorten life expectancy, including irreversible dysfunction not explained by HF

• Lung disease that would negatively impact postimplantation survival, including recent pulmonary infection

• Diabetes uncontrolled or with evidence of significant end-organ dysfunction

• Severe peripheral vascular disease accompanied by rest pain or extremity ulceration

• Moderate to severe aortic insufficiency without plans for correction during implantation

• Mechanical aortic valve that will not be converted to bioprosthesis at time of implantation or patch closed

• Severe right ventricular dysfunction requiring permanent right VAD support

• Severe cognitive impairment or organic brain syndrome

• Unresolved drug or alcohol dependency

• History of behavioral patterns or psychiatric illness likely to interfere with therapy compliance

• Unwillingness to accept blood or blood products

• Advanced age with frailty

• Insufficiency financial means or insurance coverage

• Any other medical condition likely to limit short-term survival or quality of life following VAD implantation

Figure 4. Unique cardiovascular and non-cardiovascular considerations in the evaluation of candidates for left ventricular assist device therapy. (Reproduced with permission Wilson et al.⁸¹)

In addition to reviewing the considerations and potential contraindications outlined below, each LVAD evaluation should explore potential eligibility for transplantation to frame strategic intent of support and define bailout options for device complications, as well as explore non-MCS alternative therapies for advanced HF. Intravenous inotropes provide temporary hemodynamic support for many systolic HF patients, often at the expense of an increased risk of arrhythmia and ischemia. Although inotropes have been proven in studies to be inferior to LVAD in advanced HF, they remain a viable palliative therapy for systolic HF symptoms in patients unable or unwilling to receive durable MCS⁸². The option to shift care goals to symptom palliation should be prospectively reviewed with all patients considering durable MCS. For the overall HF epidemic, palliative care remains the most common approach for most patients dying with HF, many of whom have advanced age, comorbidities, or ventricular geometry unsuitable for LVAD placement.

Cardiovascular Considerations

Right Ventricular Dysfunction

Right-sided HF remains a major determinant of post-LVAD outcomes. The presence of pre-implant RV failure has been estimated to increase mortality after LVAD 3-4 fold⁸³,⁸⁴. Post-LVAD RV failure can lead to residual deficits in quality of life and functioning and can impair LVAD performance. RV failure requiring prolonged inotropic support or placement of a temporary RVAD still occurs in about 20% of patients following LVAD implantation despite increasing attention to the right heart. Although multiple risk scores for RV failure have been proposed, none can reliably predict freedom from post-implant right HF given the potential for intraoperative insult to the right heart along with dynamic alterations in septal configuration after LVAD implant leading to functional tricuspid regurgitation and declining RV output⁸⁵. In patients flagged as unacceptably high risk for RV failure prior to MCS, there remains an important role for biventricular assist devices or TAH (discussed in Chapter 14)⁸⁶.

Multiple domains of information are integrated to the final pre-implant assessment of RV function. Hemodynamic considerations include the ability of the RV to generate high pulmonary artery pressures (RV stroke work index <300mmHgxml/m2 or pulmonary artery pulsatility index <1.85) and inability to reduce central venous pressures (unable to reduce right atrial pressure <10mmHg)⁸⁷,⁸⁸. Imaging of the right heart via echocardiography can also identify RV geometry (e.g., increased RV to LV size ratio), reduced contractility (RV fractional area change or tricuspid annular plane systolic excursion), or residual severe tricuspid regurgitation that may need addressing at LVAD implant⁸⁹. Lastly, serologic markers of right heart congestion are also used to identify risk of post-implant RV failure, such as azotemia (BUN>40 mg/dL) or congestive hepatopathy (elevated bilirubin or prothrombin time/international normalized ratio). However, in contrast to cardiac transplant, significant pulmonary hypertension due to left HF [WHO (World Health Organization) Group 2] without associated right HF is not a contraindication for LVAD support, and many such patients with pulmonary hypertension can have improvement or resolution of pulmonary vascular resistance with LVAD unloading.

Structural Heart Disease

Pre-implant assessment of residual valvular heart disease is also required, particularly aortic insufficiency (AI). AI can progress after LVAD and lead to a futile cycling of blood from the aorta to LV cavity, leading to ineffective forward cardiac output despite normal LVAD flow⁹⁰. Any degree of AI, mild or greater, should be addressed with repair or replacement at the time of LVAD implant. Mechanical valve in the aortic or mitral position may also need to be removed and replaced at the time of LVAD given risk of valve thrombosis and stroke. The degree of mitral regurgitation does not appear to alter post-LVAD outcomes or require repair. Although tricuspid regurgitation, as discussed above, is a risk factor for post-implant RV failure, the role of tricuspid annuloplasty or replacement at the time of LVAD remains an area of active investigation⁹¹.

Intracardiac shunts must be identified and closed at LVAD implant to reduce risk of right to left shunting. If significant proximal right coronary artery stenosis is identified, a single vessel bypass may be performed for RV preservation at the time of LVAD implantation. All patients with prior cardiac surgery should have a computed tomography of the chest to define aortic calcification for LVAD outflow cannula placement, evaluate the course of bypass grafts, and review LV apical calcification to guide alternative inflow cannula placement sites.

Ventricular Arrhythmias

Patients with incessant or uncontrollable ventricular tachycardia or fibrillation are not suitable for isolated LV support, as these arrhythmias can impair right heart function. Most ventricular arrhythmias are scar mediated and will persist after LVAD, though some hemodynamically-induced arrhythmias may abate. In patients who have failed pre-implant anti-arrhythmic drug therapy or ablation, concomitant intraoperative cryoabalation or post-LVAD ablation can be completed on full support, although concerns remain about anticoagulation and thrombosis⁹². Most chronic HF patients will have a primary prevention defibrillator or resynchronization pacemaker in place at the time of LVAD implant, and early collaboration with electrophysiology team members can optimize the approach to management of implantable electrical devices⁹³. For example, ventricular arrhythmias on LVAD support are often well tolerated, so thoughtful programming of existing implantable defibrillators (e.g., high rate thresholds, extended monitoring, or multiple rates of anti-tachycardia pacing) is required to avoid shocks while the patient is awake⁹⁴,⁹⁵.

Non-Cardiovascular Considerations

Renal Function

Renal insufficiency is one of the most common comorbidities in patients with chronic systolic HF. Although renal dysfunction resulting from reduced cardiac output can be reversible after LVAD, elevated pre-implant creatinine is strongly associated with poor LVAD outcomes⁹⁶,⁹⁷. A reduced 24-hour creatinine clearance, proteinuria, and cortical thinning on ultrasound have been linked to persistent renal dysfunction after LVAD implantation. A short trial of inotropes or temporary circulatory support may be required to assess reversibility. Patients already receiving renal replacement therapy or those at high risk for becoming dialysis dependent should not be considered as candidates for durable LVAD support given limited outpatient dialysis options with high infection and mortality risk⁹⁸. In patients eligible for dual heart-kidney transplant, in hospital inotropic or temporary support could be pursued rather than durable MCS.

Hepatic Function

Underlying cirrhosis and portal hypertension are associated with poor outcomes after any cardiac surgery, including LVAD implantation. Any clinical evidence of significant liver disease should be resolved with liver biopsy before implantation. Hepatic congestion is also a marker of right heart dysfunction, placing such patients at higher risk for needing biventricular support. Chronic hepatopathy can also lead to thrombocytopenia and elevated prothrombin time, contributing to perioperative bleeding risk⁹⁹. The Model for End-Stage Liver Disease (MELD) score has also been used to risk stratify LVAD recipients¹⁰⁰. In patients with congestive hepatopathy without frank cirrhosis, reduction in central venous pressure and improved cardiac output with VAD can improve indices of liver function.

Pulmonary Function

A major predictor of post-implant mortality is the need for mechanical ventilation in those with cardiogenic shock¹⁰¹. All acute respiratory processes from pneumonia to embolism should be resolved prior to durable LVAD implant. Severe obstructive or restrictive lung disease contributes to prolonged time in the intensive care unit and extended post-operative mechanical ventilation. Accurate assessment of intrinsic lung disease can be challenging in patients with pulmonary edema. In general, candidates who require chronic supplemental oxygen to maintain saturation should be considered high risk and patients with a forced expiratory volume <1 L should not be considered for VAD.

Malnutrition, Obesity and Diabetes

Malnutrition manifest as reduced serum protein stores and cardiac cachexia is common in chronic HF and likely related to reduced appetite, early satiety, and cytokine activation. Cardiac cachexia low body mass index (BMI <18 Kg/M²) increases the risk of perioperative death¹⁰². Patients with markers of malnutrition without end-stage cachexia may benefit from preimplantation enteric feeding or stabilization with inotropes to bolster nutrition prior to cardiac surgery. Obesity itself is not associated with mortality risk after LVAD until very high BMI levels, although it may carry with it risk for delayed wound healing and infection¹⁰³. Many programs consider BMI>40 Kg/M² a relative contraindication to durable MCS. Collaboration with specialists in nutrition, metabolic support, and bariatric surgery is encouraged, although significant weight loss after LVAD placement is seldom realized. Many patients with obesity and HF develop insulin resistance and type 2 diabetes mellitus. The presence of poorly controlled diabetes can increase the risk of renal dysfunction, neuropathy, and driveline infection. Demonstration of the ability to manage diabetes as reflected in consistently reduced hemoglobin A1c levels can predict engagement in self-care after MCS.

Neoplastic Disease

Patients should undergo age appropriate screening for cancer prior to LVAD implantation where possible according to standard guidelines. Patients with a malignancy