Cardiac glycosides

Digitalis is the oldest of the drugs used in the treatment of CHF today. The main action of the drug is thought to be exerted by inhibition of the plasma membrane Na⁺/K⁺-ATPase, increasing intracellular concentrations of Na⁺ and Ca⁺. A variety of autonomic effects have been shown in acute experimental studies and sympatholytic effects have been described.

Acute effects in CHF

Older uncontrolled studies have suggested that digitalis produces beneficial hemodynamic effects in patients with decompensated heart failure, expressed as a decrease in pulmonary capillary wedge pressure, an increase in cardiac output, and a fall in heart rate.^1,2 It appears that the effect of digitalis on hemodynamics is dependent on the hemo-dynamic state of the patient. Whereas positive effects have been observed in decompensated heart failure, the effects in normal subjects are largely negligible.^3,4 Although the slowing of heart rate would be of benefit in diastolic heart failure without systolic dysfunction, there are limited data about this effect.⁵ Acute administration of digitalis restores baroreceptor function and causes a decrease in sympathetic activity^6,7 though not all findings have been reproduced.⁸

Chronic digitalis therapy

The first double-blind placebo-controlled trial with chronic digoxin treatment was published in 1977.⁹ Subsequently a series of small short-term trials provided conflicting data about digoxin efficacy,^10–13 with some studies suggesting favorable effects of digoxin on clinical heart failure symptoms, echocardiographic parameters, and exercise capacity, in particular in patients with more advanced left ventricular dysfunction.

In the first large study, the Captopril-Digoxin Multi-center Research Group trial, 300 patients with relatively mild heart failure were compared using captopril, digoxin or placebo. Digoxin and captopril were equally effective in preventing hospitalization or an increase in diuretic dosages. Although digoxin-treated patients showed a significant increase in ejection fraction, in contrast to the captopril group, digoxin did not improve exercise capacity as much as captopril. The German and Austrian Xamoterol Study Group investigated the effect of digoxin together with xamoterol and placebo. Digoxin improved clinical indices of heart failure but not exercise capacity.¹⁴

Several trials have used a withdrawal design to evaluate digoxin efficacy. In the PROVED trial a randomized double-blind withdrawal of digoxin was investigated in 88 patients with stable NYHA class II – III maintained on digoxin and diuretics.¹⁵ More placebo patients had worsening heart failure with an increase in the need for diuretics and hospitalization, and an impairment in exercise capacity and LV function. In a similar study, the RADIANCE trial, 178 patients with CHF were investigated during digoxin withdrawal compared with maintained digoxin therapy;¹⁶ both groups were also on angiotensin-converting enzyme (ACE) inhibitors. The results were similar to those of the PROVED trial, with placebo patients showing a statistically significant deterioration in cardiac function, hospitalization, quality of life and exercise capacity.

It should be noted that the withdrawal study design is inferior to prospective treatment studies and provides only inferential data about efficacy. In contrast, the Digitalis Investigators Group (DIG) study, the only survival study of digoxin, was a multicenter, prospective, randomized, placebo -controlled, double-blind trial in 7788 patients with mild to moderate heart failure and sinus rhythm.⁵ Among the investigated patients, 6800 had signs of systolic dysfunction expressed as ejection fraction of < 45%. The remaining patients might be considered to have had only diastolic dysfunction. There was no effect on the primary endpoint of all-cause mortality but digoxin significantly reduced the number of hospitalizations from worsening heart failure.

In clinical practice, the use of digoxin appears to be declining¹⁷ as part of a secular trend as opposed to a reaction to the publication of the DIG trial results. However, other factors may be contributing. One substudy has garnered attention because of a putative increase in mortality in women assigned to receive digoxin.¹⁸ Subsequently, data from the DIG trial supporting the relationship between serum digoxin concentration (measured at 4 hours post dose) and outcomes were published. A survival benefit was observed for serum digoxin concentrations of 0.5–0.8ng/ dL compared to a neutral effect for concentrations between 0.9 and 1.1 ng/dL and an increase in mortality for higher concentrations (greater than 1.1 ng/dL).¹⁹ In an analysis that synthesized these two important observations, Adams et al determined that the increased mortality observed in women was likely an effect of higher serum digoxin concentrations, given that women tend to have smaller volumes of distribution of the drug than men.²⁰

Taken together, present data on digoxin suggest that this drug induces small but beneficial effects on cardiac function, morbidity, and symptoms. There is a neutral effect on all-cause mortality. However, the therapeutic window is narrow, and the potential risk for serious arrhythmias cannot be ignored. The need to monitor serum drug levels appears to be heightened by recent analyses and in that context practitioners must be vigilant about the possibility that drug – drug interactions may elevate digoxin concentration. Further, the applicability of the data derived from digoxin withdrawal trials to contemporary practice that includes beta-blockers is unclear; the efficacy of digoxin in stable patients with mild heart failure on combined ACE inhibition and beta -blocker has not been established.

Diuretics

Fluid retention is a consistent finding in patients with CHF. The need for reduction of blood volume in patients with edema was recognized several hundred years ago. Historically, drugs with mild diuretic effects, such as mercury salts, carbonic anhydrase inhibitors, and thiazides, have been used.

A more substantial and conventional way to achieve diuresis is the use of loop diuretics.²¹ Compensatory fluid retention, as a response to lower cardiac output and reduced kidney perfusion,²² might be of short-term benefit in restoring optimal preload in the earlier states of CHF. However, a further increase in intracavitary pressure increases wall stress in the myocardium with a parallel increase in oxygen consumption and energy expenditure. In the classic physiologic description, the elevation of venous pressure shifts the hydrostatic balance across the capillary wall toward a net filtration of fluid to the extracellular space, and finally to the formation of edema. The decrease in renal blood flow that commonly accompanies heart failure stimulates the renin system, which leads to secretion of angiotensin and aldosterone. Other neurohormones that promote retention of sodium and water include vasopressin, norepinephrine, and prostaglandins.²³ In contrast, endogenous sodium excretion is promoted by natri-uretic peptides and prostacyclin.

Neurohormonal and hemodynamic effects of loop diuretics

In patients with pulmonary edema, intravenous furosemide is normally followed by a prompt response and relief of symptoms. However, findings are not consistent regarding the mode of action of loop diuretics in the acute phase of decompensated heart failure. A reduction in filling pressures occurs even before diuresis is initiated^24–26 and has been attributed to vasodilation. However, acute arterial vasocon-striction has also been found, alone or in combination with venodilation,^27–29 potentially related, at least in part, to acute upregulation of neurohormones such as aldosterone. Similarly, cardiac output may increase, remain unchanged or decrease. Although furosemide is the most thoroughly tested loop diuretic, there are others available and commonly used, including bumetanide³⁰ and torsemide.³¹

Most long-term studies have involved a small number of patients and used a variety of drugs and doses; chronic neuroendocrine effects are less well studied. Oral furosemide treatment has been associated with a reduction in norepinephrine concentration and a profound increase in plasma renin activity, angiotensin, and plasma aldosterone concentration.^32,33

Effects on clinical outcomes and survival

No prospective randomized study has been performed in the ACE inhibitor era examining the effect of diuretics on long-term survival. One meta-analysis of (small) randomized trials with active controls suggests that diuretics improve exercise capacity and the risk of worsening disease.³⁴ Further, despite concerns regarding the potential neurohormonal activation by diuretics, it should be kept in mind that studies showing positive survival effects of ACE inhibitors, beta-blockers or vasodilators in heart failure have all permitted diuretics as background treatment.

Clinical management

Diuretics reduce symptoms in CHF. The effect on symptoms has been formally tested in trials with furosemide and torsemide.³⁵ Further, it has been observed that the effects of ACE inhibitors may require the co-administration of diuretics.^36,37 Diuretics are also more effective in relieving edema and congestive symptoms than ACE inhibitors when given as single therapy.³⁸

Through an increase in urinary excretion of electrolytes, diuretics are prone to induce metabolic abnormalities such as hypokalemia, hyponatremia, hypocalcemia, hypomag-nesemia, and metabolic alkalosis.³⁹ The need for potassium supplements can be diminished by using potassium -sparing diuretics, such as amiloride or triamterine. However, ACE inhibitors act synergistically with potassium -sparing diuretics, which may produce hyperkalemia; diabetic patients with proteinuria and renal tubular acidosis are at particularly high risk. The addition of a potassium-sparing diuretic to a loop diuretic may further increase diuresis but in the case of spironolactone, the therapeutic effects seen in heart failure are likely mediated by aldosterone blockade rather than direct diuresis. Additionally, the diuretic effect of a loop diuretic can be augmented by other diuretics acting at different sites in the nephron, especially if clinically apparent diuretic resistance has developed. In particular, the combination of a loop diuretic with a thiazide enhances the diuretic effect, when the former is administered slightly after (e.g. 30 minutes) the latter.

Alternative to diuretic therapy: ultrafiltration

It is difficult to foresee a future situation when diuretics are no longer needed in the treatment of CHF. However, recent data using ultrafiltration show promise, albeit with an invasive approach that is likely to be limited to patients with significant or refractory edema or diuretic resistance. Specifically, the UNLOAD trial demonstrated that ultrafil-tration led to greater weight loss (5.0 ± 3.1kg) than diuretics alone (3.1 ± 3.5, P = 0.001) in a cohort of 200 patients hospitalized with symptomatic heart failure. Reductions in patient-reported dyspnea were the same. There were fewer subsequent heart failure hospitalizations at 90 days though it is important to highlight the fact that the trial had an open-label design. Mitigating this confounder to some degree was the companion finding of fewer unscheduled visits for heart failure during extended follow-up.⁴⁰

Diuretics summary

The need for relief of edema and fluid retention and the generic status of most diuretics will prevent the initiation and completion of a classic placebo-controlled randomized long-term survival study. However, given concerns about the metabolic and neurohormonal effects of the diuretics and recognizing that peripheral edema is more a cosmetic nuisance than a life-threatening condition, it is advisable to keep the diuretic dosages as low as possible.

Aldosterone receptor blockers

Aldosterone plays an important role in the pathophysiol-ogy of heart failure, facilitating sodium retention and potassium loss. Further, it activates the sympathetic nervous system, stimulates myocardial and vascular fibro-sis, and is a component of the circulating renin – angioten-sin–aldosterone system.^41–44

Although aldosterone antagonists have diuretic effects, they differ from other diuretic agents in that they are neu-roendocrine antagonists and thereby have a potential to be effective in the long-term treatment of patients with CHF. In particular, spironolactone was evaluated in the RALES study in which 1663 patients in NYHA class III or IV were randomized to active drug or placebo.⁴⁵ Spironolactone was initiated at 25 mg/day with adjustments to 12.5 or 50 mg depending on serum potassium. Ninety-five percent of the patients were on ACE inhibitors but only 11% had a background therapy of beta-blockers. The trial was discontinued early after a mean follow-up period of 24 months because of beneficial effect of spironolactone. There were 386 (46%) deaths in the placebo group and 284 (35%) in the spironolactone group; the relative risk attributed to spi-ronolactone was 0.70 (95% confidence interval (CI) 0.60 – 0.82, P < 0.001) attributed mostly to a lower risk from sudden cardiac death. The RALES trial demonstrates that improved antagonism of the renin – angiotensin system by spironolactone reduces the risk of both morbidity and mortality in CHF. However, shortly after the publication of the RALES trial results, there was an apparent increase in admissions for hyperkalemia attributed to initiation of the drug,⁴⁶ highlighting the challenges of translating RCT data into sound clinical practice (e.g. appropriate patient selection and protocol-based monitoring of potassium levels).

Subsequently, in a different population (early post myocardial infarction (MI) with left ventricular dysfunction or heart failure), the selective aldosterone receptor blocker eplerenone was shown to reduce all-cause mortality and cardiovascular deaths, with effects most noticeable within the first 30 days following MI. This early effect suggests that the benefit may be mediated more by protection against hypokalemia than through inhibition of aldosterone. In fact, while cases of serious hyperkalemia were more common with study drug compared with placebo (5.5% versus 3.9%), the incidence of hypokalemia was decreased by more than half (from 8.4% to 3.1%).^47,48

Vasodilators: acute therapy

Vasodilation reduces left ventricular afterload and preload and these beneficial effects were observed as early as 1956,^49,50 but it was not until the 1970s that the concept was widely accepted.^51,52 The first drugs used were pure vasodilators, such as nitroprusside, nitroglycerin, and phentolamine. Later, agents with combined effects were developed. Examples of combination therapies are the inotropic drugs with simultaneous vasodilation, such as milrinone, and the ACE inhibitors, both of which are reviewed in other sections of this chapter.

Reduction of afterload and preload in CHF improves left ventricular performance according to the Frank – Starling relation with less myocardial oxygen demand and increased cardiac output.^53,54 Further, vasodilation might theoretically reduce valvular regurgitation by means of afterload reduction and may improve organ dysfunction by acting directly on selected vascular beds, such as the coronary and renal vasculature.

Nitroglycerin, nitroprusside and, in the United States, nesiritide have been used for acute short-term vasodilation therapy in heart failure. However, the clinical trials data for these drugs are either limited or controversial.

Nitroglycerin

Nitroglycerin causes smooth muscle cell relaxation and vasodilation of arterial and venous vessels through action on guanylate cyclase and the generation of cyclic guanosine monophosphate.⁵⁵ Nitrates are conventionally used topically or as an intravenous infusion; administration causes reduction in left ventricular filling pressures within 3 – 5 minutes, mainly by venodilation and lowering of preload.^56–61 Further, nitroglycerin reduces systemic vascular resistance and afterload, with ensuing improvement in cardiac output. It is also conceivable that nitrate therapy favorably affects myocardial perfusion and oxygen supply/ demand ratio.^62,63 However, tolerance occurs early and may limit effectiveness.

Nitroprusside

Nitroprusside generates nitric oxide and nitrosothiols, which stimulate guanylate cyclase to increase intracellular cGMP. Smooth muscle cell relaxation is rapidly induced after administration. Sodium nitroprusside is converted to cyanide and is metabolized to thiocyanate, which may accumulate and lead to thiocyanate toxicity during prolonged nitroprusside therapy. Toxicity is rare during short -term administration (<3μ g/kg/min for less than 72 hours) in the absence of renal failure. As compared to nitroglycerin, nitroprusside is more potent and causes a more pronounced arterial vasodilation; there are minor effects on the renal and hepatosplanchnic vasculature.⁵⁹ In contrast to nitroglycerin, nitroprusside may induce a coronary steal phenomenon.⁶⁴ Owing to its potent vasodilation property, nitroprusside may cause adverse hypotension, especially in cases of inadequate filling pressure.

Nesiritide

Brain natriuretic factor, or nesiritide, was shown in early studies to lower filling pressures and improve patient report of symptoms. Colucci and investigators demonstrated a dose-dependent reduction in pulmonary capillary wedge pressure at 6 hours in a double-blinded placebo-controlled study.⁶⁵ Subsequently, the VMAC trial demonstrated improvements in PCWP at 3 hours compared to placebo (and comparable to nitroglycerin), with greater reductions compared to nitroglycerin (TNG) seen at 24 hours, perhaps reflecting tachyphylaxis to the nitrate therapy. Patient self-report of dyspnea also favored nesiritide.⁶⁶

Following approval by the Food and Drug Administration in the United States, a series of controversial papers that included retrospectively reanalyzed published and unpublished data suggested that the drug might exacerbate renal dysfunction and worsen 30-day mortality rates.^67,68 The mechanism by which these effects occur was not clearly delineated but the impact on clinical practice was swift, leading to a marked decrease in use.⁶⁹ As a result, a panel of experts and the FDA reiterated that the major benefit is relief of dyspnea and the dose used to achieve this benefit is 0.01μ g/kg/min. In response to the concerns raised by the studies, a large randomized double-blinded outcomes clinical trial has been initiated. At the same time, other natriuretic peptides are either in clinical use (caperitide in Japan) or in clinical trials (ularitide, CD-NP).

It is important to note that data are lacking to support the use of nesiritide as an intermittent outpatient therapy⁷⁰ and in that context the data parallel what is known about the use of intermittent inotropic therapy.

Newer vasodilators

A number of other novel agents are in clinical trials for use in the acute phase including adenosine antagonists^71,72 and relaxin,⁷³ a naturally occurring peptide that modulates cardiovascular responses to pregnancy.⁷³ In addition, a direct myosin activator, CK1827452, will be studied both acutely using intravenous delivery and then, on conversion to oral formulation, chronically upon discharge, following the paradigm established in the EVEREST trial.

Effects of long-term vasodilator therapy

Nitrates and hydralazine

Oral nitroglycerin and hydralazine have been studied, either alone or in combination. The effects on left ventricular function and hemodynamics are similar to the acute effects of vasodilators described above.^74,75

Hydralazine was available as an antihypertensive agent when vasodilator therapy was adopted as a therapeutic strategy in CHF. Hydralazine acts as a dominant arterial vasodilator but has probably also mild inotropic properties, which might be due to a reflex activation of sympathetic activity.^76,77 This inotropic action might be responsible for a less favorable effect on myocardial oxygen consumption counteracting the unloading effects of vasodilation.^78,79

The addition of a nitrate to hydralazine causes a greater effect on the reduction in filling pressures than can be achieved by hydralazine alone.⁸⁰ In view of the beneficial action of nitrates on coronary dynamics, a nitrate should be added to hydralazine therapy in patients with significant coronary artery disease.⁸¹ However, despite the focus on direct venous and arterial vasodilation, current thinking about mechanism has focused on the nitrate component as a nitric oxide donor and hydralazine as an agent that mitigates nitrate tolerance through a complex mechanism mediated by NADH oxidase.⁸² This mechanism has been used to explain the possible race-based differential clinical effect seen among African Americans compared with Caucasians, inferred from the first two V-HeFT trials.

V-HeFT-I was the first placebo-controlled clinical trial to study the effect of any vasodilator on survival in patients with chronic heart failure. The study randomized 642 patients with mild to moderate heart failure to placebo, prazosin or the combination of hydralazine and isosorbide dinitrate. Two years after randomization, the survival in the hydralazine-isosorbide treated group was significantly better than the placebo group (P < 0.028); for the entire follow-up, the difference trended toward significance (P = 0.093). Of note, the mortality rate in the prazosin group was not different from the placebo group.⁸³

The second V-HeFT study compared the efficacy of hydralazine and isosorbide with that of enalapril. Two years after randomization, the all-cause mortality was 18% in the enalapril group as compared with 25% in the hydralazine-isosorbide group (P = 0.016). For the total follow-up, the difference was not significant (P = 0.08).⁸⁴

A retrospective analysis of both V-HeFT-1 and V-HeFT-2 suggested that African Americans derived benefit from the hydralazine-nitrate combination, whereas Caucasians, presumably most of European descent, did not.⁸⁵ Conceptually, this analysis was based on a series of observations that suggested a relative deficiency of nitric oxide in African Americans, though the exact nature of the defect(s) has not been established.

To test this hypothesis, 1050 self-identified African Americans with NYHA class III or IV heart failure were randomized in a survival trial with a three times daily formulation of hydralazine and nitrate.^86,87 A composite endpoint was used, combining mortality, quality of life as measured on the Minnesota Living with Heart Failure Questionnaire and time to first hospitalization; each component was statistically significant in favor of the combination therapy. Most pronounced was the impact on mortality, which declined from 10.2% in placebo-treated patients to 6.2% in patients on active therapy. The cohorts were well managed, with high percentages on conventional treatment with ACE inhibitor, angiotensin receptor blocker and beta-blocker. The magnitude of this change parallels or exceeds almost all other double -blinded placebo-controlled trials and as such represents a significant achievement. However, while providing an opportunity to advance medical therapy for CHF, the controversial approval based on race by the US Food and Drug Adminstration and other factors such as three times daily therapy have negatively impacted its adoption by practitioners.

Calcium channel blockers

The role for calcium channel blockers remains very limited in patients with CHF. With the first-generation calcium channel blocker nifedipine, vaso-dilatory effects are counterbalanced by negative inotropy and additional deleterious effects on hemodynamics, neurohormonal activation and, not surprisingly, disease progression.^88,89 The effects of diltiazem were unfavorable in patients with CHF in conjunction with myocardial infarction in a large placebo-controlled trial in 1237 patients.⁹⁰ Second-generation calcium channel blockers have not been extensively studied in patients with heart failure, but there are indications of a risk for clinical deterioration with drugs such as nisoldipine and nicardipine.^91,92 The second-generation calcium channel blocker felodipine caused vasodilation and an increase in cardiac output during 8 weeks of treatment in a placebo-controlled trial⁹³ but the effect on survival was neutral.⁹⁴

Amlodipine, a third-generation calcium channel blocker, was investigated in the PRAISE trial.⁹⁵ Among more than 1100 patients with NYHA III – IV heart failure, the overall effect on mortality as well as on the combined endpoint mortality and hospitalization was neutral but there were significantly fewer endpoints in the non-ischemic group treated with amlodipine as compared to patients on placebo (22% vs 35%, P < 0.001). As a consequence, patients with non-ischemic etiology in NYHA class IIIb or IV heart failure (n = 1652) were randomized to placebo or amlodipine 10 mg/ day.⁹⁶ There was no significant difference in all-cause or cardiac mortality and cardiac event rates between the two groups. The data from PRAISE and the felodipine trials suggest therapeutic neutrality; however, it is likely that both amlodipine and felodipine may be used safely to treat concomitant angina or hypertension in patients with CHF, if other proven drugs such as ACE inhibitors and beta-blockers are ineffective or not tolerated. All other drugs in this class are not indicated in any treatment algorithm for heart failure.

Other vasodilators

Other potent vasodilators, such as pra-zosin and minoxidil, are not indicated in the short-or long -term management of CHF. Several other drugs with multiple effects including vasodilation have failed in clinical trials. For example, flosequinan, a vasodilator with a combined venous and arterial effect and possible positive inotropic and chronotropic effects, was associated with increased mortality in a large multicenter trial (PROFILE).⁹⁷ Additionally, the prostacyclin epoprostenol, which might improve hemodynamics, has been shown to have an adverse effect on mortality in severe heart failure.⁹⁸

Drugs affecting the renin–angiotensin system

ACE inhibitors

ACE inhibitors were introduced for the treatment of heart failure within the last 25 years. Their potential value was suggested by studies showing attenuated LV remodeling after myocardial infarction^99,100 and improved symptoms,¹⁰¹ hemodynamics^102,103 and survival.¹⁰⁴ Multiple landmark studies have reported effects of ACE inhibitors on survival in patients at risk for heart failure and those with clinically manifest disease. In a similar way, these trials have demonstrated benefit across NYHA classes I – IV.¹⁰⁵ As a consequence, the use of ACE inhibitors is supported by all contemporary published clinical practice guidelines.

Effects on exercise capacity and hemodynamics

An extensive review of these data¹⁰⁶ suggested that ACE inhibitors improve exercise capacity, consistent with changes in symptoms. In addition, ACE inhibitors were documented in early studies to induce beneficial hemodynamic responses. These effects included a vasodilatory effect and an increased cardiac output, increased stroke volume, and reduced pulmonary wedge pressure.^102,103

Survival trials

The first major trial, CONSENSUS, included 253 patients in NYHA class IV randomized to receive placebo or enalapril. After a follow-up of 6 months (primary objective), the overall mortality was reduced by 27% (P = 0.003). The number of days of hospital care was reduced and NYHA classification significantly improved with enalapril.¹⁰⁴ In the Studies of Left Ventricular Treatment (SOLVD) trial, 2569 patients with symptomatic heart failure NYHA class II – III received placebo or enalapril besides conventional heart failure therapy.¹⁰⁷ After an average follow-up of 41.4 months, mortality was significantly reduced from 40% to 35% (P = 0.0036), most notably reducing deaths attributed to progressive heart failure. Hospitalizations for heart failure were also reduced and symptoms and quality of life assessed by questionnaire were improved.¹⁰⁸

The early post infarct cohort was examined in the Survival and Ventricular Enlargement (SAVE) trial, in which 2231 patients with ejection fraction of 40% or less, but without overt heart failure or symptoms of myocardial ischemia, were randomly assigned treatment with captopril or placebo.¹⁰⁹ Mortality from all causes was 20% in the captopril group and 25% in the placebo group (relative risk (RR) 19%, P = 0.019). In a similar evaluation (the TRACE study) with trandolapril, 1749 patients with left ventricular dysfunction were randomly assigned to treatment with placebo or active study drug.¹¹⁰ Treatment was initiated 3 – 7 days from the onset of myocardial infarction. All-cause mortality in the placebo group was 42.3% and 34.7% in the trandolapril group, a 22% relative reduction of mortality (P = 0.00065). Finally, in the AIRE study, 2006 patients with clinical evidence of heart failure any time after the index infarction were randomly allocated to treatment with ramipril or placebo on day 3 – 10 from the onset of infarction.¹¹¹ Mortality from all causes at the end of the study was 17% in the ramipril group and 23% in the placebo group (RR 27%, P = 0.002) with a mean follow-up of 15 months. Taken together, these studies provide incontrovertible support for the use of ACE inhibitors in post-MI patients with left ventricular dysfunction.

Dose and class effects: clinical questions remain

Uncertainties about the importance of dose of ACE inhibitor and class effects have stimulated debate and clinical evaluation. Two dose ranges of lisinopril were compared in the ATLAS trial. Patients with CHF (n = 3164) and ejection fraction < 30% were randomized to a low dose of lisinopril (2.5 – 5.0 mg/day) or a high dose (32.5 – 35 mg/day) for a median of 45.7 months.¹¹² There were 717 deaths in the low-dose group versus 666 in the high-dose group (hazard ratio (HR) 0.92; P = 0.128) for the high dose. The combined endpoint of all-cause mortality or hospitalization for any reason showed a hazard ratio of 0.88 (95% CI 0.82 – 0.96, P = 0.002). The side effects and tolerability were similar in the two groups.

These findings indicate that patients with heart failure should generally be titrated up from low doses of an ACE inhibitor, but suggest that a difference in efficacy between intermediate and high doses of an ACE inhibitor is likely to be small. Thus, patients should be titrated to dose levels achieved in the clinical trials. The value of additional dose levels such as greater than 20 mg per day of lisinopril remains uncertain but can be viewed as supported in part by the results of the ATLAS trial.

With regard to the relative benefits of different ACE inhibitors, there are no large comparative studies that provide definitive evidence of the superiority of one ACE inhibitor over another. Nevertheless, it is widely acknowledged that these drugs differ in a number of fundamental respects: the ability to bind to tissue ACE, chemical structure, pharmacodynamics, pharmacokinetics and the supporting clinical trials data.¹¹³ In an attempt to address the clinical relevance of different ACE inhibitors, Pilote and investigators compared 1-year survival among elderly post-MI patients using data from hospital discharge records and pharmacy databases in the province of Quebec, Canada.¹¹⁴ Attempts were made to correct for the propensity to receive a particular ACE inhibitor and to the extent accomplished, the data suggest a significant difference among the seven options for ACE inhibitors available on the provincial formulary, favoring ramipril. These data support the concept but do not provide absolute confirma-tion that the different ACE inhibitors have different long -term clinical benefits.

Trials on prevention

A reduced incidence of heart failure by ACE inhibitors has been demonstrated in several trials. In the prevention arm of the SOLVD study,¹⁰⁷ the incidence of heart failure and the number of hospitalizations were reduced and similar findings were reported in SAVE.¹⁰⁹ In an overview of ACE inhibitor trials,¹⁰⁸ the preventive potential of ACE inhibitors is clearly demonstrated.

Three landmark studies have been performed among patients at risk for heart failure: HOPE,¹¹⁵ EUROPA¹¹⁵ and PEACE.¹¹⁶ These trials, when considered together, suggest a role for ACE inhibitors in heart failure prevention,¹¹⁷ a topic covered in more detail elsewhere in this book.

Overall, it is clear that the number needed to treat to save one life in a year is significantly higher when heart failure prevention is the goal rather than treatment of established disease. Nevertheless, while the exact role for ACE inhibitors as a component of contemporary primary prevention is not absolutely clear, use of this class of drug in defined subgroups is highly justified. Those most at risk, including diabetics and hypertensives not at goal, should be strongly considered for ACE inhibitors as part of a comprehensive approach to risk factor modification.

Cost effectiveness

In asymptomatic patients with left ventricular dysfunction after an acute myocardial infarction (SAVE), captopril was cost-effective in patients aged 50 – 80 years compared to other interventions.¹¹⁹ Ramipril therapy for patients with clinical heart failure after acute myocardial infarction appears highly cost effective when assessed using data from the AIRE study.¹²⁰ ACE inhibitor treatment was also considered cost-effective in an economic evaluation of five independent studies.¹²¹ For example, enalapril therapy for patients with heart failure was cost-effective in SOLVD. Given the fact that many of the ACE inhibitors are available in lower cost generic formulations, the cost effectiveness now even more strongly favors this class of drug.

Clinical perspective

All patients with documented left ventricular systolic dysfunction (ejection fraction <35–40%) should be treated with an ACE inhibitor unless contraindications exist (including systolic blood pressure less than 80 mmHg, pronounced renal dysfunction, history of angio-neurotic edema and important valve stenosis). Treatment should be continued long term. The dosage to be used should be titrated from a low dose and increased to the levels employed in clinical trials. If no hypotension or renal dysfunction develops, titration up to enalapril 10 mg 2 ×/ day, captopril 50 mg 3 ×/day, ramipril 10mg/day, trandol-april 4 mg/day and quinapril 10 mg 2 × /day will be most effective.

Angiotensin II receptor (AT₁) antagonists

As ACE inhibition does not provide complete blockade from the synthesis of angiotensin II, a more effective blockade has been postulated by specific antagonism at the receptor (AT₁) level. One of the early ARB trials was ELITE -II, conducted in 3152 class II – IV patients with ejection fractions (EF) of < 40%, randomized to losartan 50 mg/day or captopril 50mg 3×/day.¹²² There was no significant difference in all-cause mortality or sudden death (hazard ratio 1.13; 95% CI 0.95 –1.35, P = 0.16). Significantly fewer patients in the losartan group discontinued study treatment because of adverse effects (9.7 vs 14.7%; P < 0.001). In the small RESOLVD trial, there were no differences between groups receiving candesartan and enalapril in exercise tolerance, ventricular function or symptomatic status over 43 weeks.¹²³ However, combined therapy with candasartan plus enalapril markedly reduced ventricular volumes and improved EF compared to either agent alone.

The major series of ARB trials began with Val-HeFT, a study of 5010 patients in class II – IV and EF of <40% randomized to placebo or valsartan.¹²⁴ Dose levels were increased to 160 mg twice a day. Background therapy with an ACE inhibitor was present in 93%. There was no effect on all-cause mortality (484 in the placebo group vs 495 in the valsartan group; RR 1.02, 95% CI 0.90 – 1.15, P = 0.8). In the other primary endpoint, mortality or hospitalization, there was a significant reduction from 801 (32.1%) to 723 (28.8%) (RR 0.87, 95% CI 0.79 –0.96, P = 0.009).

The CHARM Program was designed to study the effects of candesartan in a broad spectrum of patients with CHF. Three component trials included patients with reduced systolic LV function (EF <40%) in two (CHARM-Alternative and CHARM-Added) and patients with preserved LV (EF > 40%) in one (CHARM-Preserved). The primary outcome in each component trial was cardiovascular death or hospital admission for heart failure. In CHARM-Overall, the trials were combined and the effect on all-cause mortality was assessed.

Symptomatic patients with CHF intolerant to ACE inhibitors because of cough, symptomatic hypotension or renal dysfunction were included in CHARM-Alternative (n = 2028). Candesartan significantly reduced cardiovascular death or hospital admission for heart failure by 23% (P = 0.0004), whereas the rate of discontinuation of the study drug was similar to placebo.¹²⁵ In CHARM-Added (n = 2548), candesartan on top of ACE inhibitors significantly reduced the primary outcome of cardiovascular death or hospital admission for heart failure by 15% (P = 0.011). Hospitalizations for heart failure were also reduced significantly (P = 0.014).¹²⁶ In CHARM-Preserved (n = 3023), there was a non-significant effect on mortality. Hospitalizations for heart failure as reported by the investigators were reduced by 15% (P = 0.017).¹²⁷

In all patients with symptomatic heart failure (n = 7599), irrespective of background ACE inhibitor or beta -blocker therapy, candesartan reduced all-cause mortality by 9% (P = 0.055), particularly among those with left ventricular systolic dysfunction (HR 0.88; P = 0.018).¹²⁸ Among these patients, the effects on mortality were seen early; the HRs were 0.67 and 0.82 (both P < 0.001) at 1 and 2years respectively. Further, hospital admissions for heart failure were reduced significantly by 21% (P < 0.001).

ARB and background therapy

In the Val-HeFT trials, a post -hoc analysis suggested that the treatment effect with valsartan was attenuated by background beta-blocker therapy. This could not be confirmed in CHARM-Added where a similar effect of candesartan was observed regardless of background beta-blocker. In a meta-analysis, the effect of ARBs may be attenuated by background beta-blocker treatment.¹²⁹ In contrast, based on Val-HeFT and in particular CHARM-Added, the effect of valsartan and candesartan was additive on top of background ACE-inhibitor therapy with reduction of the composite primary outcome in both trials.

ARB summary

The trials with ARBs provide definitive proof that this class of drug can be used for treatment of patients with symptomatic heart failure who do not tolerate ACE inhibitors. The treatment effect is at least of similar magnitude as that achieved by ACE inhibitors. There is also an added, albeit modest effect on morbidity and mortality on top of ACE inhibitor therapy.

Non-digitalis inotropic drugs: short-term therapy but no long-term role

Vast efforts have been expended to develop drugs that might increase contractility or the state of inotropy. However, it has become increasingly evident that these drugs are associated with important negative effects in both the short and long term.

Inotropic agents differ according to their mode of action.¹³⁰ Several of these drugs increase the intracellular level of cyclic adenosine monophosphate (cAMP), either by receptor stimulation (beta-adrenergic agonists) or by decreasing cAMP breakdown (phosphodiesterase inhibitors). One class of agents affects intracellular calcium mechanisms by release of sarcoplasmic reticulum calcium or by increasing the sensitivity of contractile proteins to calcium. Further, there are inotropic drugs with multiple mechanisms of action.

Dobutamine

Drugs with beta-receptor agonist properties induce an increase in intracellular cAMP activity by stimulation of cellular receptors. Nearly 50 years ago, patients with cardiogenic shock were treated with beta-receptor agonists isoproterenol and norepinephrine.¹³¹ It was realized that both drugs had potential negative effects, such as an increased risk for arrhythmias or, in the case of norepineph-rine, untoward vasoconstriction. Dobutamine, a drug that is a modification of the isoproterenol molecule, has beta-1, beta-2 and alpha-1 adrenergic activity.¹³² Dobutamine induces vasodilation in combination with an increase in contractility, leading to an increase in stroke volume and cardiac output.^133–135 An enhancement of contractility is usually associated with an increase in myocardial oxygen consumption.¹³⁶ Side effects, such as arrhythmias or an unfavorable blood pressure response, are usually modest. Dobutamine can only be administered intravenously, in doses from 2μg/kg/min up to 20–25μg /kg/min.¹³⁷ It has been noticed that dobutamine may decrease beta -receptor sensitivity,^138,139 and prolonged infusion over 96 hours has been associated with a decrease in the hemodynamic effect by as much as 50%¹⁴⁰ (“tachyphylaxis”). The role of dobutamine is limited to hemodynamic support for patients in or near a state of cardiogenic shock. The use of the drug on an intermittent basis is not recommended though it has been used continuously for palliation in end-stage patients.¹⁴¹

Dopamine

Dopamine is an adrenergic agonist with predominantly beta-1 receptor activity.^142,143 This drug increases contractility with minor effects on heart rate or blood pressure. At low doses (0.5–2.0μ g /kg/min), dopamine acts on dopaminergic receptors, while at doses above 5.0 μ g /kg/min it has effects mediated through beta-1 receptors, and at higher doses also through alpha-receptors. Infusion at low doses causes dilation of smooth muscles in renal, mesenteric, and coronary arteries, leading to an increase in diuresis^144,145 though this therapy remains highly controversial.¹⁴⁶

Milrinone

Through inhibition of cAMP breakdown, the phosphodies-terase inhibitors bypass the beta-receptor pathway. Milrinone, related to amrinone (which fell into disfavor because of thrombocytopenia), enhances myocardial contractility and has potent vasodilatory effects,^147–149 without thrombo-cytopenia.¹⁵⁰ Whereas short-term administration may improve myocardial performance and clinical condition in CHF,^149,151 the long-term effects of an oral formulation were discouraging. Specifically, in the PROMISE trial, 1088 class III – IV patients were given milrinone or placebo. There was a 28% increase in mortality in patients treated with milri-none (95% CI 1 – 61%, P = 0.038).¹⁵²

Nevertheless, in those instances when an inotrope is required and selected, the choice of agent may be dictated, at least in part, by concomitant therapy with beta -blockers. In particular, use of milrinone should be strongly considered because the drug bypasses the beta-receptor. Indeed, Lowes and colleagues demonstrated that while dobutamine can bring about a comparable acute increase in cardiac index, it did so only at doses that are not usually used to treat heart failure: 15 – 20 μg /kg/minute).¹⁵³

Other inotropic agents: a history marked by failure

The list of inotropes (or drugs with both inotropic and other effects) that have failed in clinical trials is large. Interested readers are referred to an earlier iteration of this chapter (second edition) and various clinical reviews for more detailed descriptions.^154,155 The list of drugs includes ibopamine (an orally active dopaminergic agonist), xamoterol (beta-adrenergic blocking and high partial agonist activity), pimobendan (multiple actions including phosphodiesterase [PDE] inhibition), milrinone (oral formulation), vesnari-none (multiple effects) and flosequinon (multiple effects). More recently, levosimendan and enoximone (see below) were extensively studied but also failed in pivotal studies. Although drug development efforts continue, as for example with the novel drug CK-1827452, the goal of finding an inotropic drug that does not have proarrhythmic effects or increase energy expenditure may prove to be futile.

Levosimendan

A calcium-sensitizing agent with properties similar to pimobendan, levosimendan looked very promising from the standpoint of hemodynamics and clinical outcomes in a number of early moderately sized clinical trials.^156–158 The drug and one active metabolite with a prolonged half-life demonstrate a novel mechanism of action: calcium-dependent binding to troponin C, facilitating and prolonging actin-myosin protein cross-bridge formation without an increase in calcium flux or energy requirement. In addition, vasodilation is achieved through smooth muscle cell relaxation. However, in two large multicenter international trials, the outcomes were mixed. While the drug attained its primary novel composite endpoint in the short-term REVIVE study,¹⁵⁹ safety issues including hypotension, ventricular tachycardia and atrial fibrillation were more apparent in the active study arm compared with placebo. In the companion SURVIVE trial in which the drug was compared to dobutamine,¹⁶⁰ there was no difference in survival through 180 days with only a 2% absolute difference between the two treatment groups. In retrospect, it was likely not realistic to expect that an infusion of an intravenous drug during a heart failure hospitalization would result in improved outcomes at 6 months. Currently, the lack of demonstrable efficacy, combined with concerns over safety, have stalled the clinical development program of this novel drug.

Enoximone

In several Phase II/III trials, enoximone failed to achieve the primary endpoints. For example, in the EMOTE trial, enoximone did not significantly increase the likelihood of weaning from dobutamine versus placebo though there were favorable trends in time to death or reinitiation of intravenous inotrope. In the pivotal ESSENTIAL trial, the hazard ratio for time to first cardiovascular hospitalization or all-cause mortality was near unity and there were no differences in patient global assessments.¹⁶¹

Beta-adrenergic blockade

Clinicians have generally been cautious in using beta -blockers in patients with CHF, even though investigators in the early 1970s were already proposing a possible beneficial effect of beta-blockers in such cases.^162,163 Rigorous clinical trials data gathered during the last 15 years have incontrovertibly confirmed the beneficial role of beta -blockers in mild, moderate and severe heart failure.

Hemodynamic and neurohormonal effects

The short-term effects of beta-adrenergic blockade differ markedly from the long-term effects. After IV administration, there is a rapid reduction in heart rate, contractility, and blood pressure, with ensuing fall in cardiac output.^{164– 166} Intraventricular volumes, stroke volume, and ejection fraction are unaffected.¹⁶⁴ Beta-blockers with vasodilating properties cause an acute reduction in afterload with reduction in filling pressures.^164–167