Not Found

Find information on medical topics, symptoms, drugs, procedures, news and more, written for the health care professional.

Drugs for Heart Failure

By Sanjiv J. Shah, MD, Professor of Medicine, Division of Cardiology, Department of Medicine, Northwestern University Feinberg School of Medicine

Click here for
Patient Education

Take our short survey

(See also Heart Failure.)

Drug treatment of heart failure involves

  • Symptom relief: Diuretics, nitrates, or digoxin

  • Long-term management and improved survival: ACE inhibitors, beta-blockers, aldosterone antagonists, angiotensin II receptor blockers (ARBs), or angiotensin receptor/neprilysin inhibitors (ARNIs)

Choice of drug depends on the type of heart failure along with individual patient characteristics. The most common classification of heart failure currently in use stratifies patients into

In heart failure with reduced ejection fraction (HFrEF), all these drug classes have been studied and have shown benefit for long-term management.

In heart failure with preserved ejection fraction (HFpEF), fewer drugs have been adequately studied. However, ACE inhibitors, ARBs, and beta-blockers are generally used to treat HFpEF. ARNIs are currently being studied. Randomized, controlled trials suggest aldosterone antagonists are beneficial, but nitrates probably are not. In patients with severe HFpEF, lowering the heart rate (eg, with a beta-blocker) can exacerbate symptoms because they have a relatively fixed stroke volume due to severe diastolic dysfunction; in these patients, cardiac output (CO) is heart rate dependent and lowering heart rate can thus lower CO at rest and/or with exertion.

In patients with hypertrophic cardiomyopathy, digoxin is not effective and may be harmful.

All patients should be given clear and explicit information about their drugs, including the importance of timely prescription renewal and adherence to therapy, how to recognize adverse effects, and when to contact their physician.


Diuretics are given to all patients with HF (regardless of underlying ejection fraction) who have current or previous volume overload; dose is adjusted to the lowest dose that stabilizes weight and relieves symptoms.

Loop diuretics should be used initially for control of volume overload, but their dose should be reduced when possible in favor of aldosterone antagonists.

Commonly used loop diuretics include furosemide, bumetanide, and torsemide. The starting dose of these drugs depends on whether the patient has previously received loop diuretics. Common starting doses are: furosemide 20 to 40 mg po once/day or bid, bumetanide 0.5 to 1.0 mg po once/day, and torsemide 10 to 20 mg po once/day. If needed, loop diuretics can be titrated up to doses of furosemide 120 mg po bid, bumetanide 2 mg po bid, and torsemide 40 mg po bid based on response and renal function. Bumetanide and torsemide have better bioavailability than furosemide. If patients are switched between different loop diuretics, they should be placed on equivalent doses. Furosemide 40 mg is equivalent to bumetanide 1 mg and both are equivalent to torsemide 20 mg.

In refractory cases, IV loop diuretics or metolazone 2.5 to 10 mg po can be used for additive effect. IV infusion of furosemide (5 to 10 mg/hour) or other loop diuretics may be helpful in selected patients with severe edema. A bolus dose of loop diuretic should be given before starting an IV infusion and before each increase in infusion rate.

Loop diuretics (particularly when used with metolazone) may cause hypovolemia with hypotension, hyponatremia, hypomagnesemia, and severe hypokalemia. The dose of diuretic required acutely can usually be gradually reduced; the target is the lowest dose that maintains stable weight and controls symptoms. When HF improves, the diuretic may be stopped if other drugs improve heart function and relieve HF symptoms. Using larger than required doses of diuretics lowers CO, impairs renal function, causes hypokalemia, and increases mortality. Serum electrolytes and renal function are monitored, initially daily (when diuretics are given IV) and subsequently as needed, particularly after a dose increase.

An aldosterone antagonist, either spironolactone or eplerenone, should be added early to offset the potassium-losing effects of higher-dose loop diuretics. Hyperkalemia may result, especially when ACE inhibitors or ARBs are also taken, so electrolytes must still be monitored, especially during a dehydrating illness that could cause renal dysfunction. These drugs may have particular benefit in chronic right ventricular failure, in which hepatic congestion results in elevated aldosterone levels as aldosterone metabolism is reduced. To reduce the risk of hyperkalemia, aldosterone antagonists should generally be given only to patients whose potassium level is < 5.0 mEq/L, serum creatinine is < 2.5 mg/dL, and GFR is > 30 ml/min/1.73 m2. Furthermore, it should be noted that the equivalent dose of eplerenone is twice that of spironolactone (ie, spironolactone 25 mg = eplerenone 50 mg).

Thiazide diuretics are not normally used alone unless being given as treatment of hypertension; however, a thiazide diuretic may be added to a loop diuretic for additional diuresis and to reduce the loop diuretic dose. Hydrochlorothiazide, metolazone, and chlorthalidone can be used in this manner.

Reliable patients are taught to take additional diuretic doses as needed when weight or peripheral edema increases. They should seek medical attention promptly if weight gain persists.

Vasopressin (antidiuretic hormone) receptor antagonists are not frequently used though they may be helpful in cases of severe refractory hyponatremia in patients with HF.

ACE inhibitors

All patients with HFrEF should be given oral ACE inhibitors unless contraindicated (eg, by plasma creatinine >2.8 mg/dL [> 250 μmol/L], bilateral renal artery stenosis, renal artery stenosis in a solitary kidney, or previous angioedema due to ACE inhibitors).

ACE inhibitors reduce production of angiotensin II and breakdown of bradykinin, mediators that affect the sympathetic nervous system, endothelial function, vascular tone, and myocardial performance. Hemodynamic effects include arterial and venous vasodilation, sustained decreases in LV filling pressure during rest and exercise, decreased systemic vascular resistance, and favorable effects on ventricular remodeling. ACE inhibitors prolong survival and reduce HF hospitalizations. For patients with atherosclerosis and a vascular disorder, these drugs reduce the risk of myocardial infarction and stroke. For patients with diabetes, they delay onset of nephropathy. Thus, ACE inhibitors may be used in patients with diastolic dysfunction and any of these disorders.

The starting dose typically should be low (usually one fourth to one half of the target dose depending on BP and renal function); the dose is gradually adjusted upward over 8 wk as tolerated, then continued indefinitely. Usual target doses of representative drugs include enalapril 10 to 20 mg bid, lisinopril 20 to 30 mg once/day, and ramipril 5 mg bid; there are many others.

If the hypotensive effect (more marked in patients with hyponatremia or volume depletion) is troublesome, it can often be minimized by separating administration of other BP-lowering drugs, reducing the dose of concomitant diuretics, using a longer acting ACE inhibitor (eg, perindopril), or giving the dose at bedtime. ACE inhibitors often cause mild to moderate reversible serum creatinine elevation due to vasodilation of the efferent glomerular arteriole. An initial 20 to 30% increase in creatinine is no reason to stop the drug but does require closer monitoring, slower increases in dose, reduction in diuretic dose, or avoidance of NSAIDs. Because aldosterone’s effect is reduced, potassium retention may result, especially in patients receiving potassium supplements. Cough occurs in 5 to 15% of patients, probably because bradykinin accumulates, but other causes of cough should also be considered. Occasionally, rash or dysgeusia occurs. Angioedema is rare but can be life threatening and is a contraindication to ACE inhibitors. Alternatively, ARBs can be used, although rarely cross-reactivity is reported. Both are contraindicated in pregnancy.

Serum electrolytes and renal function should be measured before an ACE inhibitor is started, at 1 mo, and after each significant increase in dose or change in clinical condition. If dehydration or poor renal function due to acute illness develops, the ACE inhibitor dose may need to be reduced or the drug may be temporarily stopped.

In HFpEF, a randomized controlled trial of the ACE inhibitor perindopril demonstrated improved exercise capacity. It did not improve survival, although there was a high rate of crossover from placebo to ACE inhibitor in this trial (1). Given the very high prevalence of hypertension in HFpEF, it is reasonable to use an ACE inhibitor to control hypertension in these patients as these drugs may have secondary beneficial effects on exercise capacity in these patients.

Angiotensin II receptor blockers (ARBs)

These drugs are not demonstrably superior to ACE inhibitors but are less likely to cause cough and angioedema; they may be used when these adverse effects prohibit ACE inhibitor use.

In chronic HFrEF, ACE inhibitors and ARBs are likely equally effective. Usual oral target doses are valsartan 160 mg bid, candesartan 32 mg once/day, and losartan 50 to 100 mg once/day. Introduction, upward titration, and monitoring of ARBs and ACE inhibitors are similar. Like ACE inhibitors, ARBs can cause reversible renal dysfunction, and the dose may need to be reduced or stopped temporarily during an acute dehydrating illness.

Adding an ARB to a regimen of an ACE inhibitor, beta-blocker, and aldosterone antagonist is unlikely to be helpful and should be avoided given the risk of hyperkalemia. If a patient who is taking an ACE inhibitor or ARB is still symptomatic, an aldosterone antagonist should be started and/or an angiotensin receptor/neprilysin inhibitor (ARNI) should be used.

In HFpEF, a large randomized controlled trial of candesartan (2) demonstrated reduced number of hospitalizations for recurrent HF; however, hospitalization was a secondary endpoint. In another trial, (3) irbesartan was not associated with any improvement in outcomes in HFpEF. Therefore, ARBs should be used in HFpEF only if they are already being used to treat hypertension, diabetic kidney disease, or microalbuminuria.

ARBs are contraindicated in pregnancy.

Angiotensin receptor/neprilysin inhibitor (ARNI)

ARNIs are a new combination drug for the treatment of heart failure. They include an ARB and a new class of drug, neprilysin inhibitors (sacubitril is the only member currently available). Neprilysin is an enzyme involved in the breakdown of vasoactive substances such as BNP and other peptides. By inhibiting the breakdown of BNP and other beneficial vasoactive peptides, these drugs lower blood pressure, decrease afterload, and enhance natriuresis. Because neprilysin inhibitors increase BNP levels, NTproBNP levels (which are not increased by the drug) should be used instead to help diagnose and manage HF.

A large randomized, controlled trial (4) recently compared sacubitril/valsartan to enalapril in patients with NYHA class II through IV HFrEF. Sacubitril/valsartan reduced the primary endpoints of combined cardiovascular mortality or hospitalizations for HF; the number needed to treat was 21. Sacubitril/valsartan also reduced all-cause mortality. Thus, the ARNI sacubitril/valsartan should be considered in all patients with stable HFrEF, particularly those with NYHA class II or III symptoms on optimal guideline-directed medical therapy .

There are 3 strengths of sacubitril/valsartan: 24/26 mg, 49/51 mg, and 97/103 mg, all are taken po twice/day. The starting dose is 49/51 mg po bid for patients previously taking an ACE inhibitor or ARB, and 24/26 mg for patients previously taking a low dose of an ACE inhibitor or ARB (eg, ≤ 10 mg enalapril daily) or in those patients who are ACE inhibitor/ARB naive or who have low/borderline blood pressure. ACE inhibitors must be discontinued 36 h before initiation of sacubitril/valsartan. Patients previously taking an ARB can simply switch to sacubitril/valsartan without a washout period.

Complications associated with use of ARNI include hypotension, hyperkalemia, renal insufficiency, and angioedema. Sacubitril must be coupled with valsartan (an ARB) because of the increased risk of angioedema with the use of sacubitril alone or in combination with an ACE inhibitor. For this reason, combined ACE/ARNI therapy is absolutely contraindicated.

In HFpEF, a phase 2 trial showed that the ARNI valsartan/sacubitril reduced NTproBNP levels at 12 wk and left atrial volume at 36 wk; a large randomized controlled phase 3 trial of this drug is ongoing and should determine whether it improves outcomes in HFpEF.

Aldosterone antagonists

Because aldosterone can be produced independently of the renin-angiotensin system, its adverse effects are not inhibited completely even by maximal use of ACE inhibitors and ARBs. Thus, the aldosterone antagonists (also termed mineralocorticoid receptor antagonists) are often used, particularly for patients with moderate to severe symptoms or signs of heart failure. Typical drugs include spironolactone 25 to 50 mg po once/day and eplerenone 25 to 100 mg po once/day (does not cause gynecomastia in males). Aldosterone antagonists can reduce mortality, including from sudden death, in patients with LVEF <30% and chronic HF, or acute HF complicating acute myocardial infarction.

Potassium supplements should be stopped. Serum potassium and creatinine should be checked every 1 to 2 wk for the first 4 to 6 wk and after dose changes. Dose is lowered if potassium is between 5.0 and 5.5 mEq/L and stopped if potassium is > 5.5 mEq/L, if creatinine increases above 2.5 mg/dL (220 μmol/L), or if ECG changes of hyperkalemia are present. Aldosterone antagonists should not be used in patients receiving both an ACE inhibitor and an ARB because of the high risk of hyperkalemia and renal dysfunction.

In patients with HFrEF, an aldosterone antagonist plus either an ACE inhibitor or ARB is preferred over the combination of an ACE inhibitor and ARB.

In patients with HFpEF, spironolactone reduces hospitalization for HF and likely reduces cardiovascular mortality (5). Thus, aldosterone antagonists should be used in patients with HFpEF, particularly if they are volume overloaded and/or have a history of HF hospitalization. Loop diuretics can be minimized if necessary to accommodate the use of aldosterone antagonists.


In patients with HFrEF, beta-blockers, unless otherwise contraindicated (by asthma, 2nd- or 3rd-degree atrioventricular block, or previous significant intolerance), are critical for the treatment, and an important addition to ACE inhibitors in these patients. In HFrEF, beta-blockers are best started when the patient has no evidence of pulmonary congestion. Specific beta-blockers such as carvedilol and metoprolol succinate (ie, long-acting metoprolol) improve LVEF, survival, and other major cardiovascular outcomes in patients with chronic HFrEF, including those with severe symptoms.

In patients with HFpEF, beta-blockers have not been adequately tested in clinical trials , but large registries have shown that beta-blocker use is associated with improved outcomes in HFpEF despite the relatively high prevalence of chronotropic incompetence in HFpEF.

The starting dose should be low (one fourth of the target daily dose), then the dose is gradually increased over 8 wk as tolerated. The acute negative inotropic effects of beta-blockade may initially cause cardiac depression and fluid retention. In such cases, a temporary increase in diuretic dose and slower upward titration of the beta-blocker dose is warranted. Tolerance may improve over time, and efforts should be made to reach target doses. Usual oral target doses are carvedilol 25 mg bid (50 mg bid for patients 85 kg), bisoprolol 10 mg once/day, and metoprolol 50 to 75 mg bid (tartrate) or 200 mg once/day (succinate extended-release). Carvedilol, a 3rd-generation nonselective beta-blocker, is also a vasodilator with alpha-blocking and antioxidant effects; it is the preferred and most widely studied beta-blocker but is more expensive in many countries. Some beta-blockers (eg, bucindolol, xamoterol) do not appear beneficial and may be harmful.

During a severe, acute decompensation, beta-blockers should not be started until patients are stabilized and have little evidence of fluid retention. For HFrEF patients with acute HF exacerbation already taking a beta-blocker, the dose should not be decreased or stopped unless absolutely necessary. Often the beta-blocker dose can be continued in patients with an acute HF exacerbation if the diuretic dose is temporarily increased.

In HFrEF, after initial treatment, heart rate and myocardial oxygen consumption decrease, and stroke volume and filling pressure are unchanged. With the slower heart rate, diastolic function improves. Ventricular filling returns to a more normal pattern (increasing in early diastole), which appears less restrictive. Improved myocardial function is measurable in some patients after 6 to 12 mo but may take longer; EF and CO increase, and LV filling pressure decreases. Exercise capacity improves.


Hydralazine plus isosorbide dinitrate may help patients truly intolerant of ACE inhibitors or ARBs (usually because of significant renal dysfunction), although long-term benefit of this combination is limited. In black patients, when added to standard therapy, this combination has been shown to reduce mortality and hospitalization, and improve quality of life. As vasodilators, these drugs improve hemodynamics, reduce valvular regurgitation, and increase exercise capacity without causing significant renal impairment.

Hydralazine is started at 25 mg po qid and increased every 3 to 5 days to a target total dose of 300 mg/day, although many patients cannot tolerate > 200 mg/day because of hypotension. Isosorbide dinitrate is started at 20 mg po tid (with a 12-h nitrate-free interval) and increased to a target of 40 to 50 mg tid. Whether lower doses (frequently used in clinical practice) provide long-term benefit is unknown. In general, vasodilators have been replaced by ACE inhibitors, which are easier to use, are usually better tolerated, and have greater proven benefit.

Nitrates alone can relieve HF symptoms in patients with HFrEF; patients can be taught to use sublingual nitroglycerin spray as needed for acute dyspnea and a transdermal patch for nocturnal or exertional dyspnea. In HFrEF, nitrates are safe, effective, and well tolerated and are particularly helpful in patients with HF and angina. Adverse effects include hypotension and headache. Isosorbide mononitrate has been tested in HFpEF (6), where it was shown to be associated with increased adverse effects (eg, headache) and reduced physical activity. Thus, routine use of long-acting nitrates should be avoided in HFpEF.

Other vasodilators such as calcium channel blockers are not used to treat LV systolic dysfunction. Short-acting dihydropyridines (eg, nifedipine) and nondihydropyridines (eg, diltiazem, verapamil) may be deleterious. However, amlodipine and felodipine are better tolerated and may be useful for patients with HF and associated angina or hypertension. Both drugs may cause peripheral edema; rarely, amlodipine causes pulmonary edema. Felodipine should not be taken with grapefruit juice, which significantly increases plasma levels and adverse effects by inhibiting cytochrome P-450 metabolism. In patients with HFpEF, dihydropyridine calcium channel blockers such as amlodipine may be used as needed to treat hypertension or ischemia; nondihydropyridines such as diltiazem or verapamil may be used to control ventricular rate in atrial fibrillation. Verapamil is often used in hypertrophic cardiomyopathy.


Digoxin inhibits the sodium-potassium pump (Na+, K+-ATPase). As a result, it causes weak positive inotropy, reduces sympathetic activity, blocks the atrioventricular node (slowing the ventricular rate in atrial fibrillation or prolonging the PR interval in sinus rhythm), reduces vasoconstriction, and improves renal blood flow. Digoxin is excreted by the kidneys; elimination half-life is 36 to 40 h in patients with normal renal function.

Digoxin has no proven survival benefit but, when used with diuretics and an ACE inhibitor, may help control symptoms and reduce the likelihood of hospitalization in patients with HFrEF. However, because of the availability of a large number of evidence-based treatments for HFrEF, digoxin use has dropped significantly and is reserved for patients with significant symptoms despite optimal treatment with other mortality lowering medications. Digoxin should not be used in HFpEF unless it is being used to control heart rate in concomitant atrial fibrillation or to augment RV function in patients with RV failure. Digoxin is most effective in patients with large LV end-diastolic volumes and an S3. Acute withdrawal of digoxin may increase the hospitalization rate and worsen symptoms.

In patients with normal renal function, digoxin, 0.125 to 0.25 mg po once/day depending on age, sex, and body size, achieves full digitalization in about 1 wk (5 half-lives). More rapid digitalization can be achieved with digoxin 0.5 mg IV over 15 min followed by 0.25 mg IV at 8 and 16 h or with 0.5 mg po followed by 0.25 mg po at 8, 16, and 24 h. Prescription patterns vary widely by physician and by country, but in general, doses are lower than those used in the past, and a trough (8- to 12-h post-dose) digoxin level of 0.8 to 1.2 ng/mL is preferable. In addition, unlike in the treatment of atrial fibrillation, there is typically little reason to rapidly digitalize (ie, digoxin load) patients with HF. Thus, simply starting digoxin at 0.125 mg po once/day (in patients with normal renal function) or digoxin 0.125 mg po every Monday, Wednesday, and Friday (in patients with abnormal renal function) is sufficient in patients with heart failure.

Digoxin toxicity is a concern, especially in patients with renal dysfunction and perhaps in women. These patients may need a lower oral dose, as may elderly patients, patients with a low lean body mass, and patients also taking amiodarone. Digoxin has a narrow therapeutic window. The most important toxic effects are life-threatening arrhythmias (eg, ventricular fibrillation, ventricular tachycardia, complete atrioventricular block). Bidirectional ventricular tachycardia, nonparoxysmal junctional tachycardia in the presence of atrial fibrillation, and hyperkalemia are serious signs of digitalis toxicity. Nausea, vomiting, anorexia, diarrhea, confusion, amblyopia, and, rarely, xerophthalmia may occur. If hypokalemia or hypomagnesemia (often due to diuretic use) is present, lower doses and serum levels can still cause toxicity. Electrolyte levels should be monitored in patients taking diuretics and digoxin, so that abnormalities can be prevented if possible; potassium-sparing diuretics may be helpful.

When digoxin toxicity occurs, the drug should be stopped; electrolyte abnormalities should be corrected (IV if abnormalities are severe and toxicity is acute). Patients with severe toxicity are admitted to a monitored unit, and digoxin immune Fab (ovine antidigoxin antibody fragments) is given if arrhythmias are present or if significant overingestion is accompanied by a serum potassium of > 5 mEq/L. Digoxin immune Fab is also useful for glycoside toxicity due to plant ingestion. Dose is based on the steady-state serum digoxin level or total amount ingested. Ventricular arrhythmias are treated with lidocaine or phenytoin. Atrioventricular block with a slow ventricular rate may require a temporary transvenous pacemaker. Isoproterenol is contraindicated because it increases risk of ventricular arrhythmia.

Other drugs

Ivabradine is an inward "funny" (If) channel blocker that acts at the AV node to slow the heart rate. It is currently approved for use in HFrEF patients who have symptomatic HF, normal sinus rhythm, and heart rate > 70 beats/minute despite guideline-directed medical therapy (which should include beta-blockers). Typically, patients who may benefit from ivabradine are those with HFrEF who have NYHA class II or class III symptoms and heart rate > 70 beats/min who are at target beta-blocker dose or cannot tolerate a further increase in beta-blocker dose.

Various positive inotropic drugs have been evaluated in heart failure but, except for digoxin, they increase mortality risk. These drugs can be grouped as adrenergic mode of action (norepinephrine, epinephrine, dobutamine, dopamine) or nonadrenergic (enoximone, milrinone, levosimendan [calcium sensitizers]). Regular outpatient IV infusions of inotropes (eg, dobutamine) increase mortality and are not recommended. However, outpatient continuous infusions of inotropes such as dobutamine or milrinone can be used for palliative purposes in patients with severe HFrEF.

Anticoagulant drugs may be considered in patients with very large ventricles at risk of mural thrombus.

Drug treatment references