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

Published Data on Digitalis

Relevant Clinical Trials

In assessing and analyzing the current clinical status of the digitalis glycosides, mainly in the form of digoxin, a review of various clinical trials is essential (Table 2). The 1997 randomized double-blind placebo-controlled trial by the Digitalis Investigation Group (DIG) was a significant clinical trial. In this study, patients with CHF in normal sinus rhythm (NSR) were studied over 3–5 years to evaluate the effects of digoxin ( Lanoxin®, Glaxo Wellcome) on anycause mortality (primary endpoint) and hospitalization for CHF (secondary endpoint).[15] In total, 3397 patients were enrolled in the digoxin group and 3403 in the placebo group. The DIG trial found no decrease in any-cause mortality or significant difference between digoxin and placebo groups.[15] However, digoxin decreased both recurrent hospitalizations for CHF and worsening of CHF versus placebo (26.8 vs. 34.7%, risk ratio [RR] 0.72; 95% confidence interval [CI] 0.66–0.79; p < 0.001). The investigators considered that their large clinical trial defined a benefit with digoxin in decreased CHF hospitalizations and CHF deterioration but that the drug neither decreased nor increased all-cause mortality.[15] The ancillary DIG trial results, published in 2006, involved 988 ambulatory patients with chronic mild to moderate diastolic failure, in NSR, and on an angiotensinconverting enzyme inhibitor (ACE-I) plus a diuretic who were randomized to digoxin (n = 492) or placebo (n = 496).[16] In these ambulatory patients, digoxin had no significant effect on endpoints that included mortality, all-cause hospitalizations, or cardiovascular hospitalizations.[16] Atrial fibrillation (AF) has been a major focus of digitalis use. The 2007 SPORTIF (Stroke Prevention using an Oral Thrombin Inhibitor in Atrial Fibrillation) III and V studies investigated 7329 patients receiving either warfarin or the direct oral thrombin inhibitor ximelagatran. SPORTIF analyzed the survival of digitalis users and nonusers and found that patients receiving digitalis had higher mortality than nonusers (255/3911 [6.5%] vs. 141/3418 [4.1%]; hazard ratio [HR] 1.58; 95% CI 1.29–1.94; p < 0.001).[17] After adjusting for the greater number of baseline risk factors in digitalis users, the increased mortality persisted. The authors postulated that any increased mortality from digitalis possibly concealed in CHF may be revealed when using the drug for rate reduction.[17] On the other hand, a 2008 Russian analysis of international multicenter studies reported that digoxin appeared highly effective in the treatment of chronic CHF due to compromised myocardial contractility, AF, and paroxysmal supraventricular tachycardia.[18]

Turning to studies from the most recent decade, an additional analysis of the DIG study in 2013 showed that digoxin decreased all-cause 30-day hospital admissions in older patients (mean age 72 years) with chronic systolic CHF.[19] The 2014 study TREAT-AF (The Retrospective Evaluation and Assessment of Therapies in Atrial Fibrillation) found that digoxin was associated with increased mortality in patients with newly diagnosed AF.[20] After multivariate adjustment, digoxin was associated with increased mortality (HR 1.26; 95% CI 1.23–1.29; p < 0.001). This increased association was independent of adherence to digoxin, renal function, other therapies, and cardiovascular comorbidities; therefore, the authors considered their results a challenge to the current recommendations of some cardiovascular societies that digoxin should be used in AF.[20] In a 2014 study based on the Alabama Heart Failure Project involving 5153 Medicare beneficiaries in hospital for acute CHF and not receiving digoxin, 1054 (20%) were given a new digoxin prescription at discharge.[21] In this study, a prescription for digoxin at discharge for patients with ejection fraction (EF) < 45% resulted in a decreased 30-day rate of all-cause readmission to hospital; this was still in effect at 12 months, with no associated increase in mortality. The authors commented on the need for further randomized controlled trials.[21] The 2016 MAGIC (MAGnesium in Coronaries) trial assessed the value of early magnesium use in ST-elevation myocardial infarction (STEMI).[22] The authors reported on 5574 patients, with 852 (15.3%) deaths during the first month of follow-up. The 1-month follow-up also included 170 (3.0%) patients receiving digoxin at the same time; 42 (24.7%) of these patients died.[22] However, any association between digoxin and 1-month mortality in STEMI disappeared after correction for comorbidities and clinical characteristics in a multivariable analysis (HR 0.86; 95% CI 0.62–1.19; p = 0.372).[22] Three subsequent multivariable analyses found similar results, and the authors concluded that, after correction for comorbidities and clinical characteristics, there was no significant increase in mortality in the MAGIC trial when digoxin was used in the acute phase of a STEMI.[22] In 2017, the RATE-AF (Rate control Therapy Evaluation in permanent Atrial Fibrillation) trial, the first head-to-head randomized clinical trial of a beta blocker versus digoxin in AF, was designed and described. This is a prospective, randomized, open-label study with a blinded design at endpoint comparing initial rate control with digoxin or bisoprolol; the study plans to enroll 160 participants.[23] Follow-up is planned for 6 and 12 months, with a primary outcome of patient-reported quality of life and a secondary outcome to include EF determined by echocardiography, capacity for exercise, and various biomarkers to reflect clinical and cellular responses.[23] Completion is projected for 2019. The trial also has a feasibility component to help design a future trial to evaluate differences in hospital admission rates between digoxin and beta blockade.[23]

The just-reported 2018 ARISTOTLE (Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation) clinical trial assessed 17,897 patients with AF regarding mortality, modification of any association with CHF, and any modification by serum digoxin concentration.[24] At baseline, 5824 patients (32.5%) were receiving digoxin. A propensity score analysis was used for patients receiving digoxin at the beginning, patients placed on digoxin during the trial, and control participants. The use of digoxin at baseline was not associated with increased mortality (HR 1.09; 95% CI 0.96–1.23). However, patients with serum digoxin ≥ 1.2 ng/ml showed a 56% mortality (HR 1.56; 95% CI 1.20–2.04) when compared with no digoxin.[24] In addition, ARISTOTLE found that the risk of mortality (HR 1.78; 95% CI 1.37–2.31) or sudden death (HR 2.14; 95% CI 1.11–4.12) increased significantly with a new prescription for digoxin. In addition, the investigators reported a 19% increase in the hazard of death for each 0.5 ng/ml increase in serum digoxin, regardless of the presence of CHF (p = 0.001).[24] ARISTOTLE indicates that the clinician should consider whether a justifiable reason exists to prescribe digoxin in AF, such as a rate control approach with low blood pressure; if so, serum digoxin level should be careful monitored.

Meta-analyses, Observational Studies, and Registries

A 1990 meta-analysis studied whether patients with CHF in sinus rhythm benefit from digoxin. The meta-analysis assessed seven randomized controlled trials and reported that the chance of CHF deteriorating while receiving digoxin versus placebo had an odds ratio (OR) of 0.28 (95% CI 0.16–0.49).[25] The authors concluded that an average of one in nine patients with CHF and sinus rhythm received clinical benefit from digoxin.[25] Another 1990 meta-analysis found that, in 14 studies, patients receiving therapeutic drug monitoring, including for digoxin, had fewer toxic drug reactions than nonmonitored patients (OR 0.35; 95% CI 0.13–0.89), emphasizing the importance of determining drug levels in the blood.[26] These results certainly emphasize the importance of clinical skill and follow-up when using digoxin and how, when used carefully, digoxin appears to have good clinical value in CHF.

In the current decade, the 2014 European Society of Cardiology report of baseline results from the EORP-AF (EURObservational Research Programme Atrial Fibrillation) pilot general registry to assess management of AF found that amiodarone was the most commonly used antiarrhythmic medication, whereas beta blockers and digoxin were the most commonly used medications for rate control.[27] In 2015, the ORBIT-AF (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation) study evaluated a registry of 9619 patients with AF who were followed serially every 6 months for 3 years, grouped as follows: group I included 2267 patients (23.6%) who were given digoxin at study enrollment, group II included 681 patients (7.1%) who started on digoxin during follow-up, and group III included 6671 patients (69.4%) who never received digoxin.[28] Group I was not associated with subsequent symptoms, hospitalizations, or mortality: the adjusted HR for death was 1.04 in those with CHF and 1.22 in those without CHF. In group II, patients with CHF had a propensity adjusted HR of 1.05, but those without CHF had a significant association with mortality, with a propensity adjusted HR of 1.99.[28] The authors concluded that, under most circumstances, digoxin use in AF has a neutral outcome. In 2015, a meta-analysis of eight selected studies, including 302,738 patients, found that digoxin use was associated with an increased risk of allcause mortality in AF (HR 1.375; 95% CI 1.201–1.574; p = 0.0001).[29] Subgroup analysis found that digoxin was associated with increased all-cause mortality, both in AF with CHF (HR 1.201; 95% CI 1.074–1.344; p = 0.001) and in AF without CHF (HR 1.172; 95% CI 1.148–1.198; p = 0.0001).[29] The authors claimed that sensitivity analyses were significant and suggested that digoxin should be preferred over other rate control medications in AF. Another 2015 meta-analysis also reported an increased risk for mortality in AF with the use of digoxin, with the presence or absence of CHF resulting in no significant difference.[30] In this meta-analysis of 11 observational studies in which 318,191 patients were followed for a mean of 2.8 years, the overall use of digoxin in AF was associated with a 21% increased mortality risk (HR 1.21; 95% CI 1.12–1.30). Patients with AF receiving digoxin, matched for propensity score, showed a 17% mortality increase (HR 1.17; 95% CI 1.13–1.22).[30] In another 2015 meta-analysis, involving 235,047 patients with AF, digoxin use had a 29% association with increased mortality risk (HR 1.29; 95% CI 1.21–1.39); in the same article, in 91,379 patients with CHF, digoxin increased mortality risk by 14% (HR 1.14; 95% CI 1.06–1.22).[31] On the other hand, an additional 2015 meta-analysis of 52 studies comprising 621,845 patients found that patients receiving digoxin were 2.4 years older than controls, had a higher incidence of diabetes mellitus (DM), and showed increased use of both diuretics and antiarrhythmic medications.[32] In this meta-analysis, the unadjusted pooled risk for mortality with digoxin was increased (RR 1.76; 95% CI 1.57–1.97), but a randomized controlled trial found no association between digoxin and mortality (RR 0.99; 95% CI 0.93–1.05); across all types of studies, digoxin resulted in a small but significant decrease in all-cause hospitalizations (RR 0.92; 95% CI 0.89–0.95; p < 0.001; n = 29,525).[32] Another 2015 review of 18 studies reported that, in CHF with sinus rhythm, digoxin appeared to have a neutral effect on mortality, especially when close monitoring of the digoxin blood level was carried out.[33] On the other hand, these authors found that digoxin use in AF without systolic CHF resulted in an overall increase in mortality and concluded that the role for digoxin in the treatment of patients with CHF in sinus rhythm was limited.[33] In contrast, a 2016 meta-analysis of 16 studies involving 111,978 patients found that digoxin had a stronger association with all-cause mortality when used in patients with AF without CHF (HR 1.47; 95% CI 1.25–1.73) than in those with AF with CHF (HR 1.21; 95% CI 1.07–1.36).[34]

In terms of the most recently published reviews and meta-analyses, a 2016 meta-analysis of 47 randomized controlled trials used in the European Society of Cardiology (ESC) and American College of Cardiology Foundation (ACCF)/American Heart Association (AHA) guidelines found that the following all decreased the risk of CHF hospitalization: medications affecting the renin–angiotensin–aldosterone system (24–37% decrease), digoxin (60% decrease), beta blockers (22% decrease), and cardiac resynchronization therapy (36% decrease).[35] Digoxin was not mentioned in the therapies that decreased mortality. One caveat by the authors was that several included studies were from the 1990s and earlier.[35] A 2017 systematic review of risk factors for QTc prolongation implicated hypokalemia, some diuretics, some antiarrhythmic medications, and some other medications but found no association between digoxin or statins and QTc prolongation.[36] In 2017, an extensive systematic review protocol was planned that involved a meta-analysis and a trial sequential analysis of digoxin versus placebo, no intervention, or another medical intervention for AF and atrial flutter.[37] Hopefully, the results, when available, will have significant benefit for millions of patients and the economics of healthcare and will help further clarify the precise place of digoxin in the management of these prevalent arrhythmic diseases.