Clinical Use of Digitalis: A State of the Art Review

Thomas F. Whayne Jr


Am J Cardiovasc Drugs. 2018;18(6):427-440. 

In This Article

Clinical Role of Digitalis Glycosides in Modern day Medicine

Endogenous Cardiac Glycosides, Including Ouabain

Endogenous cardiac glycosides (compounds formed from a simple sugar and another compound resulting from replacement of a hydroxyl group in the sugar molecule) are present in plants, animals, and insects.[38] The widespread presence of these compounds complicates our understanding of digitalis derivatives and their place as medications. The compound, ouabain, is actually an endogenous hormone with a function in renal tubular cells of modulation of the Na+/K+-ATPase (sodium pump) activity[39] and thereby control of natriuresis.[39,40] Cardiac glycosides are also classified as cardiotonic steroids.[39,41] Ouabain is a cardiac glycoside/cardiotonic steroid, as is digoxin; both ouabain and digoxin have been identified in mammalian adrenal glands, the hypothalamus, and blood plasma.[42] Biosynthesis of both digoxin and ouabain appears to occur in the adrenals with control by angiotensin II, endothelin, and epinephrine released from the midbrain by stimulation of cerebral areas that sense sodium concentration in the cerebrospinal fluid and potassium in the plasma.[42] With exercise, rapid changes in endogenous ouabain may facilitate cardiac function by increasing the intracellular calcium level in atrial and myocardial cells.[42] Increased concentrations of endogenous ouabain are also associated with cardiac dysfunction, cardiac hypertrophy, and arterial hypertension; 40% of Europeans with uncomplicated hypertension have been found to have increased endogenous ouabain in association with cardiac hypertrophy and a decreased heart rate, whereas, in advanced hypertension, increased endogenous ouabain has a correlation with blood pressure and total peripheral resistance.[42] Increased endogenous ouabain is also associated with idiopathic dilated cardiomyopathy patients whose CHF is more likely to progress.[42] Rostafuroxin, which antagonizes endogenous ouabain, has been found to decrease blood pressure in hypertensive animals and could represent a new class of antihypertensive medication that works by interfering with endogenous ouabain at the Na+/K+-ATPase pump.[42]

An interesting observation regarding endogenous cardiac glycosides, with an association with hypertension, is that digitalis-like factors have been shown to be elevated in women with preeclampsia. Of special note is that the use of digoxin-binding antibodies in women with severe preeclampsia remote from term has been shown to decrease the occurrence of pulmonary edema and deterioration in renal function.[43]

Herbal Remedy Mix-ups Leading to Digitalis Toxicity

An interesting toxicity problem is that boiled comfrey leaves are used by some individuals as a herbal treatment for insomnia.[44] Apparently, the leaves of comfrey (Symphytum officinale) and foxglove (D. purpurea) are easily confused when the foxglove is not in bloom.[45] Outbreaks of foxglove poisoning have been reported when the intention was to imbibe a comfrey herbal tea,[45,46] as in the case of nine patients presenting with nausea, vomiting, diarrhea, and dizziness, with peak serum digoxin levels ranging from 4.4 to 139.5 ng/ml and treated with digoxin-binding antibodies.[45] The processing and ingestion of the leaves of the foxglove leads to digitalis toxicity from the cardiac glycosides present in the plant, such as digoxin and digitoxin.[44] Such a case was reported where an elevated serum digoxin level confirmed the mistaken use of foxglove leaves, and treatment with digoxin-binding antibodies was successful.[44] Another case involved three patients who consumed potato dumplings flavored with leaves of Borago officinalis that had been erroneously confused and mixed with leaves of the foxglove, D. purpurea.[47]

Worldwide, cardiac glycosides are also present in other plants beside the foxglove. For example, a case report involved digitalis toxicity from consumption of an extract of the Indian rubber vine plant (Cryptostegia grandiflora).[48] Another example involves a case report of two Italian individuals who ingested leaves of a plant they thought were "donkey ears" and subsequently presented to an emergency room with nausea, vomiting, diarrhea, fatigue, visual hallucinations, marked sinus bradycardia, hypotension, and repolarization abnormalities on the electrocardiogram, and 1:1 atrioventricular block in one patient.[49] To complicate the diagnosis, serum digoxin levels were not detectable, but subsequent digitoxin levels were elevated. Of note, Digifab was used as treatment because it interacts with multiple cardiac glycosides, including digitoxin, whereas monoclonal antibodies are more specific to a particular cardiac glycoside, such as digoxin.[49] Day lilies (genus Lilium and genus Hemerocallis) are ornamental plants widely used for various celebrations but can be confused with lily of the valley (genus Convallaria), which contains potent cardiac glycosides with a similarity to digitalis.[50] These lilies in general can be especially toxic to house cats, with reports of renal failure with day lilies and digitalis toxicity with lily of the valley.[50] The leaves of oleander (Nerium oleander) contain digitalis, as evidenced by a case of ingestion in a suicide attempt where serum digoxin on emergency room arrival was 5.69 nmol/l (4.44 ng/ml).[51] Oleander is an ornamental plant whose toxicity is mainly due to a mixture of non-digitalis cardiac glycosides; unfortunately, it is occasionally ingested as an herbal medicine in the form of a tea made from the leaves.[52] The yellow oleander results in thousands of poisoning cases in South Asia every year.[53]

The conversion of digitalis from foxglove as a folk medicine to a modern pharmaceutical shows the importance of regulation, since such a preparation cannot be listed as a drug until the pharmaceutical company proves effectiveness and safety. When the product is still a dietary supplement or used as an herbal remedy, both of which can be subject to abuse, it is up to the US FDA to show the supplement is unsafe.[54]

Herbal Interactions With Digoxin

Unanticipated interactions between a medicine such as digoxin administered for a specific reason and another medication are always of concern. What is especially relevant is that anywhere between 18 and 45% of patients may also be taking an alternative, supplemental, or herbal medicine; medical practitioners must be aware of and document such usage.[55] For example, St John's wort (SJW; Hypericum perforatum) can decrease the plasma or serum concentration of digoxin because of a cytochrome P450 (CYP) and P-glycoprotein (PgP) association between digoxin and SJW.[56] In fact, it is CYP3A4 that is so potently induced by SJW;[57] both SJW and digoxin are substrates for CYP3A4 and PgP.[58] Other commonly used supplements such as ginkgo, ginseng, kava, and saw palmetto do not affect the serum digoxin level.[56] Cessation of an SJW herbal tea has been reported to result in digoxin toxicity in a patient taking digoxin before stopping the tea.[59]

Clinical Pharmacology

Cardiac glycosides/cardiotonic steroids have been thought to function as inhibitors of the ion transport activity (sodium pump) of Na+/K+-ATPase in myocardium cells,[60] but this thinking is now challenged, and the inotropic effect of cardiotonic steroids is thought not to be due to inhibition of Na+/K+-ATPase.[61] The digitalis glycosides increase intracellular calcium via action on intracellular ryanodine receptors and the subsequent formation of transmembrane calcium channels.[61] On the other hand, ouabain triggers the release of calcium from the cell through signal transduction paths, with activation of myocardial metabolism, started by ouabain's activation of extracellular Na/K-ATPase.[61] Accumulation of digoxin in the body without a loading dose can arrive at a steady-state serum concentration in 5–7 days, whereas digitoxin takes 35 days to achieve a steady state without a loading dose.[62] For elimination, digitoxin depends much less on the kidney and renal function than does digoxin.[62] Regarding the inotropic effect of digitalis, it has been suggested that digitalis interacts with Na+/K+-ATPase to decrease sodium transport capacity, which results in a transient increase in Na+ during the early part of the cardiac cycle that could increase the subsarcolemmal Na+ concentration, followed by increased Ca2+ through Na+/Ca2+ exchange, resulting in a positive inotropic effect.[63] The alpha-subunit of sarcolemmal Na+/K+-ATPase present on most eukaryotic cell membranes has long been recognized, but whether this is the cause of positive cardiac inotropy remains uncertain.[13] Therefore, in summary, the Na+/K+-ATPase present in most eukaryotic cell membranes has been known for many decades as the molecular target of digitalis glycosides, but whether symptomatic relief of CHF symptoms in dilated cardiomyopathy is due to sympatholytic effects or positive inotropic effects remains controversial.[13]


It appears that much of the controversy over the safety and benefits of digitalis began at the end of the last millennium. The DIG trial showed a trend toward a decrease in deaths from CHF (RR 0.88; 95% CI 0.77–1.01); when combined with the statistically significantly decreased risk for CHF hospitalization with digoxin compared with placebo, statistical significance remained (RR 0.72; 95% CI 0.66–0.79).[64] The reduction in relative risk was also greater for patients with EF ≤ 25%, more advanced symptoms, and an enlarged heart.[64] A 2015 editorial summarized that clinicians using medications with a very narrow therapeutic index, such as digoxin, must understand the pharmacokinetics, advantages, disadvantages, and any other issues and carefully educate the patient on the use of digoxin.[65]

Cancer Risk

There is interest in digitalis as a possible anticancer treatment. A study of prostate cancer cell lines presented evidence that digitalis cardiac glycosides, including digoxin, digitoxin, and ouabain, inhibited the proliferation of prostate cancer lines, with some variable sensitivity among the three cell lines. Two possible mechanisms for the inhibition of the prostate cancer cells are a sustained increase in Ca2+ concentration in the cells and increased apoptosis.[66] It has also been hypothesized that activation of Na+/K+-ATPase, a known effect of digitalis preparations, may be a new target for cancer therapies to improve clinical outcomes.[67] Unfortunately, as discussed in a 2017 meta-analysis, the antitumor effect of digitalis noted in preclinical studies required high digitalis concentrations that would not be tolerated in the typical clinical situation; the estrogen-like activity of cardiac glycosides might also increase the risk of certain cancers.[68] Digitalis glycosides have also been shown to have benefit in a renal adenocarcinoma cancer cell line, in particular, as well as others, with the cytotoxicity of digitalis most likely due to apoptosis.[69]

Sex Differences, age, Body Weight, and Renal Insufficiency With Digitalis use

Sex differences in clinical medicine have frequently been neglected, and a look at this in relation to digitalis glycosides is appropriate. Sex-based differences have been reported, with an increased risk of death in women compared with men when taking digoxin, but specific information is sparse.[70] The general comment can be made that many more women need to be enrolled in clinical trials using digitalis glycosides. The elderly have an increased risk of digoxin toxicity, so extra care in prescribing is indicated.[71]

Consideration and understanding of the effect of body weight is indicated because a digoxin dosage calculated based on total body weight in obese patients may be excessive, as fat-free body weight appears to be more important in determining the serum digoxin concentration.[72]

In renal insufficiency with decreased creatinine clearance, essential total body and renal clearance of digoxin are decreased, and the dose must be adjusted accordingly.[73] On the other hand, digitoxin dosage does not need to be adjusted for renal insufficiency, since any decreased renal excretion is compensated for by increased fecal excretion.[73]

Regarding racial differences that might be associated with digitalis, a search of the literature yielded very little other than increased morbidity and mortality for underrepresented racial and ethnic groups with AF, without elucidation of specific mechanisms.[74] A secondary analysis of 272 adults with chronic CHF found that, in the group with heart failure with preserved EF (HFpEF), African Americans were more likely to receive digoxin and a diuretic when symptomatic than were Caucasians.[75] Another study regarding possible racial bias in the treatment of CHF found no multivariate association between race and the administration of digoxin or diuretics.[76] No specific information was readily available regarding any specific physiologic racial differences in association with digitalis, and any difference appears most likely relegated to patient background, education, and cultural factors that may affect medication adherence.

Myocardial Infarction

Any recommendation for the use of digitalis in acute myocardial infarction would appear problematic and currently lacking in clinical evidence. Although digitalis glycosides increase the contractility of a heart with global ischemia, there is complex physiology with regional functional impairment in an ischemic heart.[77] It appears that deterioration of segmental function occurs with digitalis in necrotic and ischemic segments, with increased contractility involving the normal adjacent myocardial segments.[77] With the invasive and noninvasive management of ischemic heart disease now available, it would be difficult to justify the use of digitalis without proof from a randomized clinical trial; support for and performance of such a trial would appear to be most unlikely.

Congestive Heart Failure

Standards of care are well-established for dilated cardiomyopathy with reduced EF (frequently referred to as heart failure with reduced EF [HFrEF]). The acronym CHF is usually associated with HFrEF; HFpEF is lacking in wellestablished pharmacologic management.[78] There does not appear to be any proven indication for digitalis in HFpEF, which is, in most cases, associated with hypertension,[78] unless a rate control issue with AF is also present. Therefore, in this review, CHF is associated with HFrEF, and standards for pharmacologic management include ACE-Is, beta blockers, angiotensin-receptor blockers (ARBs), and aldosterone antagonists. Isosorbide dinitrate with hydralazine can be of value in patients intolerant of an ACE-I or an ARB.[79] A neprilysin inhibitor can offer additional benefit in some patients with CHF not responding to standard management.[80]

For a long time in the history of digitalis, its primary clinical effect was thought to be an ability to slow the heart rate. In NSR, the effect of digitalis on slowing the heart rate is via increased vagal tone, so the effect is only noted at rest.[65] It is now established that digitalis glycosides increase myocardial contractility of the failing heart;[77] increased cardiac output and cardiac index are then associated, and digoxin is also associated with a significant decrease in the pulmonary capillary wedge pressure.[81] However, despite these potentially advantageous physiologic effects in CHF, digoxin has now generally been relegated to symptom relief.[82,83] A 2011 analysis of 80 review articles, randomized controlled trials, or observational studies found that digoxin, compared with placebo, decreased the progression of CHF but did not appear to decrease mortality.[84] A 2012 metaanalysis of ten observational studies, with 1841 patients with CHF or AF, found a very high prevalence (38.7%) of noncompliance with digoxin.[85] Such noncompliance may contribute to the clinical problems reported with digoxin usage, which further highlights that only experienced and skillful clinicians should use cardiotonic steroids and that part of that skill must involve careful patient instruction and compliance, with medication withdrawal if compliance does not occur. Without an analysis of compliance, lower but adequate digoxin levels (there appears to be little relation between the serum concentration and therapeutic effects of digoxin[86]), verification of practicing clinician skill level, and placement of digoxin in context with currently accepted primary CHF therapies, it appears unacceptable to throw out an old CHF standby as advocated because of some negative meta-analyses.[87] In contrast and for controversy, others have concluded that the DIG trial established clinical efficacy and safety[13,86] and that, in general, digitalis glycosides, specifically digoxin, definitely have a role in CHF, especially to decrease CHF deterioration and increase exercise tolerance.[88] Knowledge of other cardiac medications that increase digoxin levels (e.g., verapamil, amiodarone, spironolactone, and flecainide) and so require a reduced digoxin dosage is essential, and toxicity risk can be decreased by not pushing digoxin for AF rate control but rather avoiding increased doses and adding another medication such as a beta blocker to attain ideal rate control.[86] A 1997 study of 19 patients with CHF with EF < 45% found that a low dose of digoxin (0.125 mg/d over 2 weeks) resulted in a significant increase in ventricular performance, with no further increase with a moderate dose of digoxin (0.25 mg/d). This study in CHF adds support for a lower dose of digoxin with no advantage from higher doses or levels.[89] A 2004 analysis of 11 articles of ≥ 20 adult patients with symptomatic CHF found no difference in mortality with digitalis versus controls, but digitalis therapy was associated with reduced hospitalizations and decreased clinical deterioration,[90] consistent with current concepts of the drug.

Atrial Fibrillation

There are two major strategies for managing AF—rhythm control (attempt to return AF to NSR) and rate control (use of medication to slow ventricular rate response or in extreme cases, creation of complete heart block with ventricular pacing). Digitalis glycosides are a major consideration for rate control. In terms of clinical outcomes, it is generally considered that there is little overall difference in outcomes with rhythm or rate control.[91] Pushing digoxin or digitoxin to very high levels, such as 9–18 ng/ml, has been reported with successful cardioversion.[92] However, despite such reports, this can be legitimately classified as beyond consideration, especially in this era of concern regarding the place of digitalis glycosides, especially from younger clinicians. The 2016 focused update of the Canadian Cardiovascular Society Atrial Fibrillation Guidelines Committee pronounced digoxin as indicated for rate control when other medications such as beta blockers or calcium channel blockers are inadequate, contraindicated, or not tolerated (e.g. hypotension).[93] Despite this, the Canadian Cardiovascular Society considered digoxin a second-line medication for rate control with AF because of various published retrospective, cohort, subgroup, and other studies suggesting possible harm at times from digoxin use, even in rate control situations.[93] A cost-effectiveness analysis from 2011 concluded that, in patients with CHF, rate control is less expensive and at the same time more effective than rhythm control and recommended digoxin as initial management for patients with AF with concomitant CHF.[94] However, opposing studies are available that are against any use of digoxin for rate control. A 2015 meta-analysis involving 17 studies encompassing 408,660 patients reported that the use of digoxin resulted in a 14% increase in all-cause mortality in patients with AF with CHF (RR 1.14; 95% CI 1.04–1.24) and a 36% increase in mortality in patients without CHF (RR 1.36; 95% CI 1.18–1.56).[95]

However, much support remains for the use of digoxin in rate control protocols for AF. A 2009 systematic review of multiple databases, which concluded that a descriptive synthesis of available data provided little evidence to show that beta blockers or calcium antagonists improved either exercise capacity or symptoms in patients with chronic AF, showed that digoxin with either a beta blocker or calcium antagonist should be offered as first-line therapy for chronic AF.[96] The main indications for digoxin in AF were previously considered to be restoration of sinus rhythm and prevention of AF recurrence, but this is certainly not the case, and the only potential benefit in AF is rate control via slowing of the ventricular response rate.[97] Studies from 2011 and 2015, both published in BMJ Clinical Evidence with systematic reviews, supported the addition of digoxin for rate control when a rate-limiting calcium channel blocker failed to result in adequate rate control in AF.[98,99] The key to safe usage of digoxin in AF appears to be the serum digoxin level. As already described in Sect. 4.1, the ARISTOTLE trial found that the baseline use of digoxin was not associated with increased mortality when the digoxin level was < 0.9 ng/ml compared with those not receiving digoxin (HR 1.00; 95% CI 0.85–1.16; p = 0.96).[24] However, for each subsequent 0.5 ng/ml increase in serum digoxin, there was an adjusted HR increase of 19%; at a serum digoxin level ≥ 1.2 ng/ml, there was a 56% increase in HR (HR 1.56; 95% CI 1.20–2.04).[24] For new digoxin users, the mortality risk was even higher (HR 1.78; 95% CI 1.37–2.31).[24] An editorial comment following the 2018 ARISTOTLE article de-emphasized any advantage for digoxin and cavalierly dismissed any use of digoxin, suggesting that perhaps the only place for it was the garden and the history of medicine.[100] In addition, polypharmacy may have been a major confounding factor in the interpretation of ARISTOTLE, since each participating patient was taking a median of six drugs.[101] However, careful assessment of such an opinion leads to support from the ARISTOTLE trial for safe usage of digoxin, predicated on careful attention to lower serum digoxin levels,[24] with management by a skilled and experienced clinician who carefully considers when digoxin is a favorable option.