Selection and Treatment of Candidates for Heart Transplantation

Etiology of Heart Disease

With the exception of one report in which nonischemic heart disease predicted a poorer outcome,¹⁰ patients with heart failure due to coronary artery disease (CAD) generally have a higher mortality than those with congestive heart failure without ischemic heart disease.^{9 11 12 13} When considered in selection of candidates for heart transplantation, this information supports the concept that when all other clinical and hemodynamic variables are similar, patients with CAD should be actively listed for heart transplantation before patients without CAD, provided revascularization is not feasible.

When evaluating candidacy for heart transplantation in patients with cardiomyopathy who have neither ischemic nor valvular heart disease, it is important to consider that prognosis may differ, depending on the etiology of the cardiomyopathy. Spontaneous remission of cardiac dysfunction has been reported in patients with active lymphocytic myocarditis and peripartum cardiomyopathy.^{14 15} The cardiac function of patients with dilated cardiomyopathy associated with excessive consumption of alcohol may improve with abstinence. Caution should be exercised when judging prognosis in these patient subgroups, and a period of observation and intense pharmacological therapy should be undertaken before heart transplantation is considered. The heterogeneity of idiopathic dilated cardiomyopathy must be taken into account when attempting to predict outcome of patients with this condition. One study showed that survival was dramatically better in a population-based idiopathic dilated cardiomyopathy cohort than in a referral-based idiopathic dilated cardiomyopathy cohort both at 1 year (95% versus 69%, respectively) and at 5 years (80% versus 36%, respectively) (P<.001).¹⁶ Furthermore, 1- and 5-year survival of 137 patients with idiopathic dilated cardiomyopathy who were referred for heart transplantation between 1982 and 1987 were significantly better than that of 85 patients with idiopathic dilated cardiomyopathy referred in the preceding 22 years.¹⁷ These findings suggest that both referral bias and improvement in diagnosis and treatment affect the natural history of idiopathic dilated cardiomyopathy.

Effect of Gender

Multicenter randomized trials of new therapies for heart failure have been conducted primarily in Caucasian males. Prognostic variables and therapeutic responses have not been adequately evaluated in female and non-Caucasian patients. Recently reported data from the Studies in Left Ventricular Dysfunction (SOLVD) Registry and the Survival and Ventricular Enlargement (SAVE) and Cooperative New Scandinavian Enalapril Survival Study (CONSENSUS) II trials^{18 19 20 21 22} suggest that the benefits of ACE inhibitor therapy are fewer in females than in males. Because the effects of gender and race on heart failure prognosis are ill defined at best, it is not known whether gender and race-related differences in heart failure outcome should influence the timing of heart transplantation. Clearer guidelines will be provided by clinical trials planned to evaluate the effects of gender and race on response to heart failure therapy.

Duration of Illness

A short duration of heart failure symptoms appears to be associated with a greater likelihood of spontaneous improvement. In one study heart failure symptoms improved in 6 of 11 patients (51%) with heart failure duration of less than 7 months and in none of 17 patients (0%) with duration more than 7 months.²³ In another study, 16 of 55 patients (29%) with a heart failure duration of less than 12 months improved, compared with 17 of 114 patients (15%) with heart failure duration greater than 12 months (P<.02).²⁴ However, patients with recent onset cardiomyopathy require close clinical surveillance and aggressive medical therapy because mortality rates are highest early after referral in patients with severe heart failure.^{25 26} If the condition of patients with recent onset cardiomyopathy deteriorates, such patients should be listed for heart transplantation.

Hemodynamics, Functional Capacity, and Neurohumoral Activation

Four categories of variables have been shown to independently predict the prognosis of patients with heart failure due predominantly to systolic dysfunction. These are (1) hemodynamic measures of cardiac pump performance, (2) functional capacity, (3) indexes of neurohumoral stimulation, and (4) spontaneous arrhythmias.

Left ventricular ejection fraction (LVEF) is a measure of global ventricular function that is among the most powerful predictors of survival in patients with cardiac disorders due to diverse causes. Left ventricular ejection fraction is widely used because it can be measured noninvasively by radionuclide ventriculography, echocardiography, or magnetic resonance imaging. In large populations of patients with a wide spectrum of symptoms, LVEF has a significant relation to mortality.^{7 10 27 28}

In V-HeFT-I, which enrolled patients with mild to moderate heart failure, LVEF was dichotomized at the mean value of 28%. Patients with LVEF below this value had an annual mortality of 22%, compared with a rate of 13% in patients with LVEF above the mean.¹² Preliminary results from a combined analysis of data from V-HeFT-I and II suggest that serial determinations of LVEF may enhance the prognostic value of this measure. Patients in whom LVEF decreased by more than 5% demonstrated nearly twice the number of deaths during the follow-up period than did patients in whom LVEF remained unchanged. Improvement of LVEF by more than 5% was a very favorable prognostic sign and identified a subgroup of patients with a mortality risk of less than 10% over a 1-year period.²⁹

Studies confined to patients with advanced heart failure (New York Heart Association [NYHA] functional Classes III and IV) have failed to demonstrate a difference in LVEF between survivors and nonsurvivors.³⁰ However, patients with low LVEF values (less than 20%) may be further stratified on the basis of peak exercise oxygen consumption (VO_2max)³¹ and response to optimal preload and afterload reduction.^{13 32} Right ventricular ejection fraction (RVEF) may also influence survival in patients with heart failure. Of 34 patients with LVEF below 40%, those with RVEF less than 35% had a 71% mortality over a 2-year follow-up period, whereas those with an RVEF greater than 35% had mortality rates of only 23%.³³

The relation between invasively obtained hemodynamic parameters and mortality in heart failure patients has also received considerable attention. In most studies nonsurvivors have a lower systolic blood pressure, higher left ventricular end-diastolic pressure, right atrial pressure, and pulmonary artery wedge pressure, and a lower cardiac output and stroke work index than survivors.³⁴ Although individual studies have reported the superior prognostic value of one or the other parameters, in general, pulmonary artery wedge pressure and stroke work index appear to be the most useful of these prognostic indexes.^{13 31} A recent report pointed out that pulmonary artery wedge pressure achieved after optimal reduction in preload and afterload predicts survival better than pulmonary artery wedge pressure measured when patients are first referred as potential heart transplant candidates.¹³ According to the findings of this study, patients achieving pulmonary artery wedge pressure less than 16 mm Hg while maintaining adequate cardiac output and systemic arterial blood pressure had a 1-year survival rate that was significantly greater than those in whom these hemodynamic effects could not be achieved by maximally tolerated therapy (83% versus 38%, P=.0001).

Hemodynamics measured during exercise may have greater prognostic accuracy than values obtained at rest. In one study survivors had lower pulmonary artery wedge pressure and stroke work index at rest than nonsurvivors. The response of the stroke work index with exercise significantly enhanced ability to separate survivors from nonsurvivors: in patients with a stroke work index greater than 20 g/m² at peak exercise, the mortality rate was 13%; in patients with peak exercise stroke work index below this value, the mortality rate was 67%.³⁵

Clearly the severity of hemodynamic impairment is among the most powerful predictors of mortality in patients with heart failure. Of the available variables, LVEF is the most widely applicable and clinically useful measure. However, all hemodynamic variables, with the possible exception of those related to right ventricular function, lose predictive value in patients with symptoms of refractory heart failure.

Valvular Heart Disease

Consideration of heart transplantation in patients with severe valvular heart disease requires that the outcome of valve replacement surgery be weighed against that of heart transplantation. Patients with mitral valve stenosis should very seldom be considered for heart transplantation, because even patients with severe pulmonary hypertension and NYHA functional Class IV symptoms have excellent surgical outcome with mitral valve replacement.^{55 56} Most patients with aortic stenosis, depressed ventricular performance, and severe congestive heart failure will benefit from aortic valve replacement.^{56 57} In approximately 85% of these patients, left ventricular dysfunction results from both decreased contractility and afterload mismatch. Surgical relief of the outflow obstruction decreases the pressure gradient between the left ventricle and the aorta, thereby reducing the afterload mismatch and acutely improving left ventricular performance and cardiac output.⁵⁶ Patients with a mean transvalvular gradient greater than 50 mm Hg and severely reduced LVEF may tolerate surgery that produces rapid hemodynamic and functional improvement. However, there is a small group of patients with severe left ventricular dysfunction, intraventricular pressure less than 140 mm Hg, and transvalvular gradient less than 30 mm Hg, in whom cardiac performance is not significantly improved by surgery. It is likely that these patients have mild to moderate aortic stenosis and a coincident cardiomyopathy. If patients are unresponsive to optimal vasodilator therapy, it may be necessary to consider heart transplantation.⁵⁶

Valvular surgery relieves symptoms, increases long-term survival, and prevents irreversible left ventricular dysfunction in 70% to 80% of patients with chronic aortic and mitral valve regurgitation.⁵⁸ Aortic valve replacement is recommended for patients with chronic aortic regurgitation when the symptoms of heart failure have progressed to NYHA Class III or IV or when signs of systolic left ventricular dysfunction have developed, even when only NYHA Class I or II symptoms are present.

Despite compromised preoperative left ventricular function (left ventricular end systolic volume greater than 60 mL/m² and LVEF less than 45%), aortic valve replacement is not precluded and should not be postponed in symptomatic patients.^{59 60} Most patients improve symptomatically even without marked postoperative changes in left ventricular mass or volume or improvement in left ventricular systolic function. Reversibility of left ventricular dysfunction depends greatly on its duration. If aortic valve replacement is performed more than 18 months after development of left ventricular dysfunction, myocardial damage is irreversible.^{59 60} While a delay in performing aortic valve replacement lessens the chance for recovery and allows further deterioration of left ventricular function, aortic valve replacement can and should be performed as long as LVEF exceeds 20% to 30%.^{59 60}

When surgery is contemplated for patients with chronic mitral valve regurgitation, the often slowly progressive nature of mitral regurgitation must be weighed against the immediate risks and long-term uncertainties of surgery. Surgical mortality depends on the patient’s left ventricular function and the presence of comorbid conditions, as well as on the skill and experience of the surgical team. The decision to replace or reconstruct the valve is critically important, because surgical mortality associated with mitral valve reconstruction appears to be lower than that associated with replacement, although patients selected for the two procedures differ.⁶¹ Reconstructive procedures have been useful in patients who have severe, noncalcific mitral regurgitation with pliable valves, a dilated mitral annulus, mitral regurgitation secondary to ruptured chordae to the posterior leaflet, or perforation of a mitral leaflet due to infective endocarditis in the absence of severe subvalvular chordal thickening and major loss of leaflet substance.

Long-term survival in patients with predominant mitral regurgitation who undergo mitral valve replacement is poorer than in those with mixed stenotic and regurgitant lesions, because left ventricular dysfunction may be more advanced and largely irreversible when patients with pure mitral regurgitation become symptomatic. However, surgery is still indicated in the majority of patients with mitral regurgitation and severe left ventricular failure, because conservative therapy has little to offer. Although the 5-year survival rate is approximately 30% in patients with mitral regurgitation secondary to CAD, some postoperative improvement can be expected even in patients with heart failure refractory to medical therapy, as long as the cardiac index is greater than 1.5 L/min/m² and LVEF is greater than 35%. When left ventricular dysfunction is more severe, the risk of death after mitral valve surgery becomes prohibitive, and heart transplantation should be considered instead.⁵⁸ Patients with aortic or mitral valve disease and CAD pose a special therapeutic challenge.⁶¹ The results of several studies indicate that mortality is greater in patients requiring both aortic valve replacement and coronary revascularization than in those undergoing aortic valve replacement alone. However, ignoring coexisting CAD at the time of aortic valve replacement increases perioperative risk and late mortality. Poor prognostic indexes in patients undergoing combined procedures include older age, cardiothoracic ratio greater than 50%, preoperative NYHA Class III or IV, and use of a mechanical prosthesis without adequate anticoagulation.⁶¹ These characteristics identify patients whose condition may deteriorate enough postoperatively to require consideration for heart transplantation. If CAD is present, aortic valve replacement is performed for severe aortic insufficiency. When left ventricular function is normal and aortic insufficiency is mild to moderate, a valve-conserving procedure rather than aortic valve replacement is preferred.⁶²

The best approach to patients with mild to moderate aortic insufficiency and abnormal left ventricular function remains controversial. Early and late mortality are greater when coronary revascularization is combined with mitral valve replacement than with aortic valve replacement.⁶¹ Indexes of poor prognosis in patients undergoing combined mitral valve and coronary revascularization surgery include ischemic etiology of mitral valve dysfunction, age greater than 60 years, preoperative NYHA Class IV, decreased LVEF, and concomitant ventricular arrhythmias.⁶¹ Since the most common late complication is congestive heart failure, heart transplantation may become a therapeutic consideration for these patients. In one study, 77 of 278 subjects had late heart failure, and 55 of 77 died. The high heart failure and death rates suggest that mitral valve disease continues to have a negative impact on survival even with improved surgical techniques (mitral valve repair) and perioperative critical care.⁶¹

Arrhythmias

How arrhythmias should influence the decision about and timing of listing heart failure patients for heart transplantation is an important question. While most studies have identified a relation between asymptomatic ventricular arrhythmias and total mortality, only a few have shown a correlation with sudden death, which may be the mode of death in 22% to 86% of patients with heart failure.⁶³ In the digoxin/captopril study, only nonsustained ventricular tachycardia (NSVT) frequency (more than 2.1 episodes per 24 hours), but not the mere presence of NSVT, was an independent predictor of mortality and the strongest predictor of sudden death.¹⁰ Electrophysiological testing has not significantly improved the ability to identify heart failure patients at high risk for sudden death. Except in patients with ischemic cardiomyopathy, in whom lack of inducible ventricular tachycardia predicts a low risk of arrhythmic death, electrophysiological testing is not useful in risk stratification of heart failure patients with NSVT.⁶⁴ In comparison with patients with a normal LVEF, a larger number of patients with LVEF less than 50% have inducible ventricular tachycardia, which is not necessarily a specific predictor of the subsequent risk of arrhythmogenic death.⁶⁵ The prognostic value of noninvasive markers obtained with signal-averaged electrocardiography remains uncertain. The specificity and predictive value of late potentials in forecasting an increased risk of life-threatening ventricular arrhythmias is greater in post–myocardial infarction than in nonischemic heart failure patients (60% versus 45% and 60% versus 52%, respectively).^{65 66}

Two other important issues must be considered when determining candidacy for heart transplantation in patients with heart failure complicated by ventricular arrhythmias. Therapeutic agents shown to reduce mortality of patients with heart failure, such as ACE inhibitors and β-receptor blocking agents, have not decreased the incidence of sudden death.²¹ Due to altered pharmacodynamics and pharmacokinetics, the potential of drugs used to suppress ventricular arrhythmias to worsen heart failure and arrhythmias increases as LVEF decreases.^{68 69} In one study the risk of proarrhythmic events was 0.8% in patients with LVEF greater than 30% and increased sixfold in patients with LVEF less than 30%.⁶⁹ No data support the notion that suppressing arrhythmias will prevent sudden death. Rather, the Cardiac Arrhythmias Suppression Trial (CAST) found an increased mortality in post–myocardial infarction patients treated with encainide and flecainide.⁶⁹ The addition of low-dose amiodarone (300 mg/d) to digoxin, diuretics, and vasodilators reduced mortality by 28% and hospitalizations by 31% in patients with severe heart failure independent of the presence of ventricular arrhythmias.⁷⁰

In summary, the impact of arrhythmias on survival in heart failure patients must be carefully considered when evaluating candidacy for heart transplantation. Sudden death due to tachyarrhythmias or bradyarrhythmias accounts for a large number of deaths in heart failure patients. The prognostic value of both invasive and noninvasive arrhythmia markers is greater for ischemic than nonischemic heart failure patients. The proarrhythmic effects of drugs used to suppress arrhythmias increase as left ventricular dysfunction worsens. To date, only low-dose amiodarone has been shown to decrease both sudden and progressive heart failure death rates in patients with severe left ventricular dysfunction. The implantable cardioverter defibrillator is often the best option for patients with severe left ventricular dysfunction who are at high risk of dying suddenly and in whom antiarrhythmic agents are ineffective or poorly tolerated.

Selection of Pediatric Patients for Heart Transplantation

Unfortunately, similar objective prognostic indexes have not been identified for infants and children with complex forms of congenital heart disease and dilated cardiomyopathy who are potential heart transplant recipients.

Heart transplantation may provide a survival benefit for patients with hypoplastic left heart syndrome who rarely live beyond a month without surgical treatment.⁵ It is unknown at this time whether heart transplantation is superior or inferior to the long-term results of the Norwood procedure, which consists of creating a communication between the right ventricle and the aorta and enlargement of the ascending aorta. However, therapeutic strategies sequentially incorporating both these options, depending on the anatomic details of the defect and availability of donor organs, seem reasonable. In addition, continued efforts to identify the determinants of good long-term survival after heart transplantation for infants with hypoplastic left heart syndrome are critical to allow appropriate selection of patients for heart transplantation, efficient use of donor organs, and use of other treatments for patients unlikely to benefit from a heart transplant. Clinical experience suggests that in patients with hypoplastic left heart syndrome, prognosis after heart transplantation worsens as preoperative waiting times lengthen, and thus infants should have priority.

Patients with congenital heart defects other than hypoplastic left heart syndrome may also be candidates for heart transplantation. In general, heart transplantation is considered when these patients develop myocardial failure in addition to the underlying congenital cardiac defect.⁵ Superficially, it would seem reasonable to select these patients for heart transplantation using guidelines similar to those used for patients with dilated cardiomyopathy. However, as detailed below, guidelines for heart transplantation for children with dilated cardiomyopathy and no congenital heart disease are unclear. Also, the outcome of patients with congenital heart disease and myocardial dysfunction may be different than that of patients with dilated cardiomyopathy without congenital heart disease. Finally, measurement of ventricular function in a patient with congenital heart disease and myocardial dysfunction with unusually shaped ventricles is much more difficult than in a patient with an anatomically normal heart.

The natural history of dilated cardiomyopathy in children is unclear.⁷¹ As illustrated in Fig 1⇓, 5-year survival can range from less than 40% to 80%. Undoubtedly this variability is due to the source of patients in individual studies. For example, Taliercio et al⁷² studied patients evaluated in a major tertiary referral practice. The 5-year survival rate was low. Wiles et al,⁷³ however, reported a population-based study; the 5-year survival rate was approximately 80%. Predictors of survival and death are not consistent among the studies.^{73 74 75 76} Thus, at present there appears to be no reliable way to determine the likelihood of death for an individual patient subsequent to cardiac evaluation. In view of these unknowns, the following guidelines may be helpful in identifying the child or adolescent with end-stage cardiac disease who may benefit from a heart transplant:

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Survival in children with dilated cardiomyopathy according to findings of six retrospective studies.

1. Progressive deterioration of ventricular function or functional status despite optimal medical therapy, including use of digitalis, diuretics, and ACE inhibitors

2. Growth failure secondary to severe congestive heart failure unresponsive to conventional medical treatment

3. Malignant arrhythmias or survival of cardiac arrest, unresponsive to conventional medical treatment and not likely to be successfully treated with an implantable cardioverter defibrillator

4. Need for ongoing intravenous inotropic support

5. Unacceptably poor quality of life

6. Progressive pulmonary hypertension that would predictably preclude heart transplantation at a later date.⁷⁷

Evaluation of Comorbid Conditions

Evaluation of potential heart transplant recipients should include a careful analysis of comorbid conditions that may negatively affect outcome after transplantation (Table 2⇓).^{5 6} Any coexistent systemic illness that limits survival independent of heart disease should be considered a possible contraindication to heart transplantation.^{5 6}

Effect of Medical Therapy on Prognosis of Advanced Heart Failure

In the past decade certain types of medical therapy have been shown to improve survival in subjects with heart failure. The first trial to demonstrate a mortality reduction with medical therapy was V-HeFT-I, in which the vasodilator combination of isosorbide dinitrate and hydralazine decreased 2-year mortality by 34%.³² Since V-HeFT-I numerous mortality studies in heart failure have been conducted, some showing improved outcome and some showing worsened outcome (Fig 2⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Outcome (vs placebo) of survival trials in heart failure or left ventricular dysfunction.

As seen in Fig 2⇑, ACE inhibitors are the only treatment that has consistently reduced mortality in heart failure. The reduction in mortality at 6 months in CONSENSUS (enalapril) was 40%²⁵ ; in SOLVD (enalapril), it was 29%.²¹ ACE inhibitors have also been shown to lower mortality in the setting of asymptomatic left ventricular dysfunction. In the SAVE trial, captopril reduced mortality by 19% at 42 months after an acute myocardial infarction.²⁰ In addition, ACE inhibitors have been shown to delay progression to symptomatic heart failure in asymptomatic patients and to reduce the incidence of myocardial infarction in subjects with CAD and left ventricular dysfunction.^{20 21}

These data indicate that activation of the renin-angiotensin system is part of the pathophysiology of both heart muscle disease and CAD, and the adverse effects of this activation may be attenuated by ACE inhibitors.¹¹⁷ Enalapril and captopril enhance functional status in patients with heart failure, with 40% to 80% of patients showing improvement in NYHA functional class. Although some patients may improve more than one functional class, the average improvement is one half to one functional class. The SOLVD treatment trial also found that enalapril decreased the number of hospitalizations by 15% over the study period.

Despite the proven mortality-reducing effects of ACE inhibitors, the natural history of heart failure remains progressive, with a 1-year mortality of 25% to 50% in medically treated subjects (Fig 3⇓). Fig 3⇓ provides a comparison of the effects of medical treatment with digoxin, diuretics, and ACE inhibitors to surgical treatment with heart transplantation. The only ACE inhibitor trial conducted in a patient population with advanced heart failure in which prognosis approaches that of a pre–heart transplantation population is the CONSENSUS Trial.²⁵ In this study, as shown in Fig 3⇓, 1-year mortality exceeded 40% in subjects treated with digoxin, diuretics, and enalapril. This contrasts with 1-year survival after heart transplantation according to data from United Network for Organ Sharing (UNOS), in which 50% of the heart transplant candidates were on intravenous inotropic treatment and 20% on mechanical assist devices before surgery. As seen in Fig 3⇓, there is a marked difference between medical and surgical treatment of advanced heart failure, with 1-year survival in the heart transplant group being around 90% versus 50% to 60% with medical treatment. However, because fewer than 5% of subjects with NYHA Class III and Class IV heart failure can actually receive a heart transplant because of the inadequate donor supply, the only way to improve outcome in the majority of patients with advanced heart failure is to improve medical therapy and/or implement the use of ambulatory mechanical assist devices. Fortunately, recent data indicate that this is becoming a reality.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Survival of patients with severe congestive heart failure receiving medical therapy or heart transplantation according to 1994 data from the United Network for Organ Sharing.

Two types of agents have recently been shown to improve prognosis on a background of ACE inhibition, and the results are relevant to treatment of patients close to or in the heart transplantation range. In a study conducted on an ambulatory heart failure population composed largely of NYHA Class III patients that used the combined end-point of all-cause mortality and morbidity (defined as admission to the hospital for intravenous inotropic support), the weak inotropic agent vesnarinone reduced the probability of reaching the combined end-point by 50% over a 6-month period (Fig 2⇑).¹¹⁸ In addition, patients treated with this agent improved symptomatically to the extent that could be detected by quality of life measurements. At the dose used in the trial (60 mg), vesnarinone has a mild positive inotropic effect due to type III phosphodiesterase inhibitory properties.¹¹⁸ However, unlike other phosphodiesterase inhibitors, this agent tends to lower heart rate because of the compound’s activity on ion channels, particularly potassium channels.^{118 119} Whether an ancillary property of the agent, such as cytokine inhibition, explains the reduction in mortality is not known. The answers to some of these questions may come from the ongoing placebo-controlled trial (VESnarinone Trial), in which mortality is the primary end-point.

The major side effect of vesnarinone is agranulocytosis, noted in 2.5% of patients treated with this agent. Another problem is a dose-related increase in mortality (at 120 mg/d).¹¹⁸ Therefore, the drug has a narrow therapeutic index, much like digoxin and other agents routinely used to treat heart failure. For these reasons, additional clinical trials will be required before vesnarinone is approved by the Food and Drug Administration (FDA).

Another class of agents recently associated with improved outcome in heart failure patients symptomatic on digoxin, diuretics, and ACE inhibitors is the class of β-adrenergic receptor blocking compounds. The Metoprolol in Dilated Cardiomyopathy (MDC) trial used the combined end-point of all-cause mortality and deterioration to the point of needing heart transplantation.⁶⁷ In the MDC trial the β₁-selective blocker metoprolol reduced by 90% the probability of deterioration in clinical status to the point of requiring heart transplantation (P=.0001). In the metoprolol-treated patients, the probability for the combined end-point was .058 in favor of metoprolol and did not achieve significance because all-cause mortality was not lowered by the agent. Based on the trial design, nearly all mortality in the study was due to sudden death, and metoprolol did not lower the risk of sudden death (19 deaths on placebo, 23 on metoprolol, P=NS).

A similar lack of effect on sudden death was recently reported in a study of bisoprolol (Cardiac Insufficiency BIsoprolol Study),¹²² in which all-cause mortality in subjects with idiopathic dilated cardiomyopathy was reduced by 56%. This was entirely due to reduction in death from progressive pump dysfunction. In the CIBIS trial subjects with ischemic cardiomyopathy had no reduction in mortality, and the effect of β-blocking agents in this important subset of patients is somewhat uncertain. However, the nonselective β-blocking agent bucindolol has recently been shown to improve left ventricular function in ischemic cardiomyopathy,¹²¹ and propranolol has been shown to decrease the incidence of sudden death in CAD populations.¹²²

Currently no β-blocker has been approved by the FDA for use in idiopathic dilated cardiomyopathy and advanced heart failure refractory to standard medical treatment, and its use remains experimental. Bisoprolol is not generally available in this country. The β-blocker/vasodilator carvedilol, which has been shown to substantially improve left ventricular function in heart failure patients, was stopped during phase III development because of a favorable effect on survival in several medium-sized trials. It may be available for heart failure therapy in the foreseeable future.

The issue of whether β-blockade can lower mortality in heart failure patients was not addressed by either the MDC or CIBIS trial. The MDC trial was not designed for mortality as an isolated primary end-point, and the CIBIS trial was underpowered. The answer to the mortality question will presumably come from the BEST (β-Blocker Evaluation of Survival Trial), which began in 1995 and will enroll 2800 patients randomly assigned to bucindolol versus placebo. Until this trial is complete, the issue of the use of β-blockers in advanced heart failure remains open and generally should be approached in the context of a clinical investigation with protocol approved by an institutional review board. When β-blockers are used, extreme caution is warranted, as heart failure may worsen acutely. Nevertheless, small doses of a β-blocker (eg, metoprolol, 5 mg twice a day), gradually and carefully increased over many weeks, may benefit selected patients with dilated cardiomyopathy and post–acute myocardial infarction patients, including many with heart failure.⁶⁷

If β-blockade proves to be effective in lowering mortality in advanced heart failure in both idiopathic and ischemic cardiomyopathy, and if β-blockade can be combined with vesnarinone (as can be predicted on the basis of their respective pharmacological properties), in the foreseeable future patients with NYHA Class III heart failure symptoms (UNOS Status II) may, with the addition of these two types of drugs, survive as long with medical therapy as with heart transplantation. However, current medical therapy cannot replace heart transplantation, as the natural history of cardiomyopathy remains progressive even with these new therapies. The ultimate goal of medical therapy should be to delay progression of heart failure symptoms or death to the point where a good quality and quantity of life could be achieved for a majority of patients, with heart transplantation reserved for those who do not respond to or progress with medical therapy. A randomized trial has been initiated in the United Kingdom in which patients with advanced heart failure are randomly assigned to medical therapy, heart transplantation, or a “wearable” electrically driven left ventricular assist device.

Optimization of Current Medical Therapy for Severe Heart Failure

The accepted standard medical treatment for severe heart failure includes therapy with ACE inhibitors, digoxin, and diuretics. Large clinical trials have shown that ACE inhibitors provide symptomatic benefit in most patients and reduce mortality by 18% to 30% in all degrees of symptomatic heart failure.^{21 25} Because ACE inhibitors are well tolerated in 90% of patients who receive them and provide unequivocal survival benefit, these drugs should be used in all patients with left ventricular systolic dysfunction unless specific contraindications exist. Contraindications include history of intolerance or adverse reactions to these agents, serum potassium greater than 5.5 mEq/L that cannot be reduced, and symptomatic hypotension. After initiation of ACE inhibitor therapy, doses should be titrated upward over 2 to 3 weeks with the goal of reaching the doses used in large-scale clinical trials: captopril, 50 mg three times a day, or enalapril, 10 mg twice a day. Patients who tolerate these doses may benefit from higher doses (eg, captopril, 100 mg three times a day or enalapril, 20 mg twice a day). Volume status should be assessed if hypotension or a rise in serum creatinine of 0.5 mg/dL occurs as the dose is increased. If there is evidence of volume depletion, the dose of ACE inhibitors should be reduced to the highest dose previously tolerated and the diuretic dose decreased. The ACE inhibitor dose should then be increased again. If higher doses are not tolerated even after the dose of diuretics has been decreased, then the lower dose should be continued or a trial of hydralazine-isosorbide instituted. It should be noted that relatively low blood pressure (less than 90 mm Hg), moderate renal insufficiency (serum creatinine greater than 3 mg/dL), and mild hyperkalemia that can be corrected are not contraindications to ACE inhibition. In these instances ACE inhibitors should be titrated upward slowly, as tolerated, to a maximum of half of the maintenance doses. If physicians are uncomfortable starting ACE inhibition therapy in patients with low blood pressure, they should refer such patients to centers with expertise in treating heart failure rather than abandoning attempts to use ACE inhibitors or other vasodilators.

Potassium-sparing diuretics should be stopped in all patients who are started on ACE inhibitors, regardless of the level of serum potassium. These agents can be resumed if patients remain hypokalemic on full therapeutic doses of ACE inhibitors. Potassium supplements should also be withheld unless serum potassium is less than 4 mEq/L. If potassium supplements are continued, serum potassium levels must be followed every few days until stable. Patients who develop a cough while taking ACE inhibitors should be evaluated to determine whether the cough is a result of pulmonary congestion before considering discontinuing ACE inhibitors. Most patients are willing to tolerate the nuisance of a cough in exchange for the benefits of the medications. Angioedema of the oropharynx is an absolute contraindication to further use of ACE inhibitors.²¹ No particular ACE inhibitor is recommended over another. Other ACE inhibitors (lisinopril, quinipril) are available and improve exercise tolerance. However, there are no data on whether these other agents reduce mortality or what dose might be required to do so.

Diuretics are indicated in patients with pulmonary and peripheral vascular congestion.¹²³ There is no standard target dose. It is generally better to give loop diuretics such as furosemide as a single daily dose. If response to the initial dose is inadequate, more diuresis will usually be obtained by doubling the dose rather than by giving the same dose twice daily. At high doses furosemide can be given twice daily. Patients with severe volume overload may require more frequent intravenous doses to achieve the desired brisk diuresis, and some patients even require continuous administration. Doses as high as 240 mg twice per day may be required by patients with the most refractory volume overload, but doses greater than 160 mg/d usually require additional interventions to improve diuresis. These include administration of one or more doses of diuretic intravenously in the attempt to reduce intestinal edema that may impair absorption of orally administered diuretics or addition of metolazone at doses of 2.5 to 10.0 mg per day. Since the combination of metolazone with a loop diuretic is extremely kaliuretic and can produce marked volume losses, metolazone should not be used in the outpatient setting unless volume status, blood pressure, and electrolytes can be monitored very closely.¹²² Once adequate diuresis is achieved, it may be possible to reduce or discontinue metolazone. Spironolactone can be added to loop diuretics and ACE inhibitors only if renal function and serum potassium are normal. Potassium supplementation should be stopped and serum potassium levels should be measured every few days until stable. For subjects with higher systemic vascular resistance, in whom additional afterload reduction is desirable or possible, hydralazine and isosorbide dinitrate³² may be added to digoxin, diuretics, and ACE inhibitors. This combination of direct vasodilators is also a useful alternative in patients who cannot tolerate ACE inhibitors.

Digoxin has fared well in controlled clinical trials, most notably in withdrawal studies.^{124 125} Recommendations for the use of digoxin in subjects in sinus rhythm could change when the outcome of a large survival trial, the Digitalis Investigation Group Study (DIG), becomes available. It is reasonable to measure digoxin levels when heart failure worsens, renal function deteriorates, additional medications are added that could affect digoxin levels (quinidine, verapamil, amiodarone, antibiotics, and anticholinergic agents), or signs of toxicity develop. Digoxin should be temporarily discontinued or the dose reduced if the serum level exceeds 2.5 ng/mL, renal function deteriorates significantly, toxicity occurs, ventricular arrhythmias increase, or significant conduction abnormalities occur (second- or third-degree atrioventricular block, high-degree atrioventricular block in atrial fibrillation).¹²⁴

Two large clinical trials are examining the effects of the addition of highly vasoselective calcium antagonists to the standard therapy for heart failure. However, the routine use of these agents for additional afterload reduction awaits the demonstration of a favorable effect on natural history of heart failure. Caution should be taken when using afterload reduction in patients with evidence of ischemia, which could be exacerbated by excessive reduction of perfusion pressure.

In patients who continue to have heart failure symptoms despite optimal standard medical therapy, including digoxin, diuretics, and vasodilators, additional inotropic agents must and should be given. The long-term use of inotropic agents in ambulatory heart failure remains controversial, but their use is universally accepted as a bridge to heart transplantation. Two general strategies are available, experimental oral inotropic agents or intravenous β-adrenergic agents. The former strategy has been shown to be effective with lower doses of a phosphodiesterase inhibitor,¹²⁶ while the latter approach is currently being evaluated in a clinical trial in an outpatient population. Earlier studies of higher doses of dobutamine given intermittently to outpatients led to improved symptoms but resulted in an increased death rate several weeks into the study.¹²⁷ An ongoing dobutamine trial uses a continuous low-dose infusion that anecdotally has been well tolerated in pre–heart transplantation patients. Intermittent hospitalization for 48 to 96 hours for administration of dobutamine, or the so-called “dobutamine holiday,” remains common practice for patients who may be weaned from such therapy after a few days of stabilization. This strategy is based on observations that the beneficial effects of dobutamine last beyond the actual infusion.¹²⁸

In patients who remain hemodynamically unstable despite pharmacological circulatory support, mechanical assist devices may be used as a bridge to heart transplantation.⁶ Both the Novacor and the Thermo Cardiosystems, Inc. (TCI) left ventricular assist devices have been used in clinical trials. Approximately 280 Novacor left ventricular assist devices have been implanted with a maximum implant time of 370 days.¹²⁹ One hundred fifty-five patients (55%) subsequently received a heart transplant. Though the incidence of cerebral thromboembolic complications is presently unavailable, it is a risk, and patients who have the Novacor device are receiving anticoagulation therapy with heparin or warfarin.

The pneumatic TCI HeartMate has been implanted in 223 patients.¹³⁰ The longest time of implantation has been 344 days. One hundred forty-six patients (66%) subsequently underwent heart transplantation, and 88% of these were discharged from the hospital. Five thromboembolic complications have occurred in 15 patient-years of observation, an incidence of 0.2 episodes per patient-year of observation. Long-term use of heparin or warfarin has not been required with this type of device, and patients are treated with aspirin. The TCI pump was approved by the FDA as a bridge to heart transplantation in 1994.

Patients with left ventricular assist devices may be removed from heart transplant candidate lists for several weeks after device implantation to allow end-organ recovery before transplantation. These patients usually require only brief treatment in the intensive care unit. Outpatient treatment of persons with selected left ventricular assist devices will probably be feasible. Although the assist devices are used as a bridge to heart transplantation, mere insertion of a device does not guarantee acceptability for heart transplantation, because contraindications may develop or the device may be inadequate to reverse the effects of hypoperfusion.

Most patients with heart failure have arrhythmias. Before deciding to institute antiarrhythmic therapy, left ventricular function should be optimized and heart failure compensated. Electrolyte abnormalities must be corrected and active ischemia treated. The goal of antiarrhythmic therapy is to prevent sustained ventricular arrhythmias or sudden death.¹³¹ However, most antiarrhythmic agents have negative inotropic effects. Furthermore, the proarrhythmic effects of antiarrhythmic agents increase as ejection fraction decreases.¹³² Compelling reasons such as documented aborted sudden death should be present before antiarrhythmic therapy is considered in patients with heart failure. For patients who have had a sustained ventricular arrhythmia, therapy may prevent a recurrence and prolong life. Increasing numbers of such patients are treated with amiodarone, which has proved promising in recent heart failure clinical trials.^{70 133} These are briefly discussed in the section on prognostic significance of ventricular arrhythmias. If drugs are ineffective, poorly tolerated, or judged too risky, alternative therapies include the implantable cardioverter device, surgery, or heart transplantation.^{134 135 136} There is a growing belief that in view of the low success rates and high incidence of toxicity, antiarrhythmic drugs should not be the first preferred approach in patients with left ventricular dysfunction. The implantable cardioverter defibrillator is often considered the best option for such patients. It is recommended for the following groups of patients: (1) those who have experienced sudden cardiac death due to ventricular fibrillation or sustained ventricular tachycardia and who have not responded to drug therapy; (2) patients with recurrent arrhythmias after arrhythmia surgery; and (3) patients who have survived a cardiac arrest but do not have inducible ventricular arrhythmias. Many studies have shown that the implantable cardioverter defibrillator has been associated with an impressive reduction in mortality from ventricular arrhythmias in patients with poor left ventricular function who had experienced sudden death from tachyarrhythmias.¹³⁷ Among patients with an LVEF less than 30% who had undergone implantation with a cardioverter defibrillator, 3-year cumulative survival was 67%, compared with a substantially lower projected survival rate.¹³⁸ However, the implantable cardioverter defibrillator does not prevent death from bradyarrhythmias or progressive heart failure.¹³⁹ Furthermore, many patients with an implantable cardioverter defibrillator will still require antiarrhythmic therapy because of long runs of NSVT or supraventricular arrhythmias that may repeatedly activate the device.¹³⁵

It is clear that better strategies are needed for pharmacologically bridging patients to heart transplantation. In particular, orally effective inotropic agents that do not increase mortality must be developed. Inasmuch as heart transplantation is life-saving and life-extending, a neutral survival effect of an inotropic agent that stabilizes patients and increases their probability of receiving a transplant would obviously be a component of an overall mortality-reducing strategy in a population with advanced heart failure.

The key to successful management of heart failure is compulsory outpatient follow-up. In fact, in 1990 more than 11 million outpatient visits were required for patients with heart failure, and the total cost of management exceeded $34 billion.⁴

As a minimum, long-term management of congestive heart failure should include continuity of care by an experienced and knowledgeable physician, optimal dosing of conventional therapy, and periodic evaluation of left ventricular function with either echocardiography or radionuclide angiography.

New therapies for heart failure hold much promise but can only be provided to patients referred to regional centers. In an analogous situation, in the early 1970s patient access to promising chemotherapeutic agents for cancer was maximized with the development of the National Cancer Center Program sponsored by the National Cancer Act. Incorporation of regional specialized heart failure centers may improve accessibility to optimal treatment for patients with this condition, as well as stimulate the kind of broad-based clinical and basic research efforts required for progress in this critical area.