Digoxin: a drug whose time has gone?

The cardiac glycosides, of which digoxin is a member, have been used in medicine for more than 200 years, but even today digoxin remains a controversial drug. Arguments continue over its use for various cardiovascular indications; nonetheless, continued widespread use of this drug could partly be due to convention, but its acceptance for over two centuries probably reflects a perception that its beneficial effects exceed any harmful ones. It is in the era of ‘evidence-based practice’ that randomised controlled trials have finally examined the efficacy and safety of digoxin in various contexts. Indeed, the inappropriate use of digoxin could lead to an increase in all-cause and cardiovascular mortality.1 Furthermore, a high proportion of (mainly elderly) patients still continue to be treated with digoxin for inappropriate indications1, 2 that could be detrimental, given the drug's narrow therapeutic index. It has even been commented upon that the cardiac glycosides would have a difficult ride through the drug regulatory bodies by modern criteria for safety and efficacy. Digoxin is most commonly used for ventricular rate control in chronic atrial fibrillation. However, we now recognise that it is only useful in controlling the resting heart rate and is less effective in rate control during exercise or in conditions of high sympathetic tone, whereas a beta-blocker or rate-limiting calcium antagonist, such as verapamil or diltiazem, is far more effective. Digoxin is also an ineffective drug for the cardioversion of atrial fibrillation to sinus rhythm, or the long-term maintenance of sinus rhythm postcardioversion. Furthermore, there is some evidence that digoxin may make paroxysmal atrial fibrillation worse by increasing paroxysms, and when these occur, they do so at higher heart rates than if digoxin were not used.3 In view of these observations, some authorities, especially in North America, have even described digoxin as a drug ‘whose time has gone’.4 Nevertheless, digoxin has recently been revisited for its role in heart failure;5 indeed, while its value in patients with heart failure who have atrial fibrillation has long been recognised, its efficacy in those who have sinus rhythm has often been questioned, at least until recently. The evidence supporting the efficacy of digoxin in patients with heart failure who are in sinus rhythm is also increasing, at least as a second- or third-line agent in reducing hospital admissions and some heart failure deaths.6, 7 A question remains as to whether we are using the right dose. Indeed, the dose used could have a varying effect on survival among patients with coronary artery disease. In the Digitalis Investigation Group (DIG) study,6 there was an excess of supraventricular and life-threatening cardiac arrhythmias in the digoxin group (average dose 0.250 mg), leading to an increase in arrhythmic deaths. An excess mortality associated with digoxin, demonstrated in some trials of acute myocardial infarction (MI), appears to be related to the higher dose.8 For example, in the SPRINT study, the use of digoxin doses >250 μg was an independent determinant associated with increased death during the first year after MI, although the confidence intervals were wide.8 In heart failure, a low dose of digoxin may even be as effective as a higher dose, with a lower incidence of side-effects.9 In addition, there is some concern about the use of digoxin in patients with ischaemic heart disease, although some differences are apparent between patients with acute and chronic cardiac ischaemia.10 Administration of digoxin may be potentially harmful in the setting of acute MI.10-13 Reicher-Reiss et al14 reported that digoxin was associated with a 30–50% increase in the risk of overall and cardiac mortality in chronic ischaemic heart disease, but this is more likely to be related to concomitant use of diuretics and other antiarrhythmic drugs, as well as differences in underlying cardiac impairment between the groups. There was no increase in mortality due to digoxin in the large randomised DIG study,6 where 70% of the patients had chronic ischaemic heart disease. Indeed, digoxin still remains the only inotrope that is fairly safe in severe heart failure, while other inotropes have been shown to increase mortality.15 In this issue of the IJCP (pp 108–114), Abarquez16 presents a robust defence of the ‘old but reliable digitalis’. Quoting the extensive literature (albeit relying mainly on meta-analyses and post hoc subgroup analyses), he argues for a re-examination of digoxin as a cost-effective agent for use in heart failure and an expansion of indications for its usage. This raises the argument whether the scientific community should believe retrospective post hoc non-randomised comparisons, cohort studies, meta-analyses of small (and often heterogeneous) trials, or large, prospective randomised placebo-controlled trials that have been specifically designed to answer different hypotheses. Evidence-based medicine diehards clearly favour the latter – but how much of everyday clinical practice is evidence-based? An examination of the latest issue of Clinical Evidence, published by the BMJ Publishing Group,17 clearly illustrates the lack of evidence in everyday clinical medicine. However, we have been caught out before, with the observations from small trials and meta-analyses, as well as one large trial (LIMIT-2),18, 19 that magnesium was beneficial in acute MI. It was the large mega-trial ISIS-4 that did not demonstrate such a benefit, and even showed a trend towards an adverse effect with magnesium use in acute MI.20 Perhaps the role of digoxin needs to be better defined in light of new evidence and questions raised from recent well-conducted randomised trials. The main reason for the lack of mortality advantage for digoxin in the DIG study was probably an increase in deaths which were presumed to be arrhythmic (p=0.04), perhaps related to concomitant electrolyte abnormalities. In the Randomised Aldactone Evaluation Study (RALES),21 there was a significant survival and symptomatic benefit, as well as decreased hospitalisation with the use of spironolactone in patients with NYHA II-IV heart failure, who were already on ACE inhibitors and a loop diuretic. Importantly, most of these patients were on digoxin as well and there was no increase in arrhythmogenic deaths. This could partly be due to the avoidance of hypokalaemia with spironolactone use. Furthermore, beta-blockers have been shown to decrease overall mortality, as well as mortality due to pump failure and sudden cardiac death among patients with heart failure.22 Using beta-blockers in heart failure could potentially decrease the arrhythmogenic risk of digoxin. Finally, neurohumoral activation contributes to the worsening of heart failure, and its therapeutic modulation has long been recognised to have a role in treating the disease. Digoxin reduces cardiac sympathetic activity in severe congestive heart failure;23 thus, further trial evidence is needed on whether digoxin is additive to beta-blockers in heart failure. Even among patients with heart failure and atrial fibrillation, where digoxin has been traditionally the drug of choice, we need to know whether its role should be revised, because the beta blockers, which now have good prospective trial survival data,7 could well prolong life as well as control the ventricular rate more efficiently. It is apparent that we are still learning about a drug which has been in clinical use for the past 200 years. There may still be much more to learn. The days of digoxin are far from gone.