What is the background to this?
Digoxin reduces ventricular heart rate when the patient is at rest, but is less effective at controlling the heart rate during exercise. For this reason, NICE guidance on the management of atrial fibrillation (AF) recommends that, in patients with permanent AF who need treatment for rate control, beta blockers or rate-limiting calcium channel blockers should be the preferred monotherapy. Digoxin should generally only be used as monotherapy in predominantly sedentary adults or when the other drugs are contraindicated. Digoxin may be used in combination with beta blockers and calcium channel blockers when monotherapy is inadequate. NICE guidance on the management of chronic heart failure recommends that digoxin should be used in patients with atrial fibrillation and any degree of heart failure, or in those with worsening or severe heart failure due to left ventricular systolic dysfunction despite ACE inhibitor, beta-blocker and diuretic therapy
What does this analysis claim?
Data from over 7,000 patients in the SPORTIF III and V studies was analysed. The results suggest that digoxin, like other inotropic drugs (i.e. those which affect the muscular contraction of the heart), may increase mortality in patients with AF at moderate to high risk of stroke (see NPC key slide 3 to calculate this). Significantly more patients who were taking digoxin died compared with non-users (6.5% vs. 4.1%, respectively; hazard ratio [HR] 1.58, 95% confidence interval [CI] 1.29 to 1.94; P<0.001). After adjusting for baseline risk factors such as age, body mass index, blood pressure, gender, race, smoking status, and co-morbidities, the risk remained unchanged (HR 1.53, 95%CI 1.22 to 1.92; P<0.001)
As the study authors point out, the results should be interpreted cautiously because the patients were not randomised with respect to digoxin use. Also, one of the studies that was analysed (SPORTIF III) was not blinded. This retrospective analysis only demonstrates an association between digoxin and increased mortality. There may be confounding factors which have not been controlled for, and it is unclear what possible pathophysiological mechanisms may be responsible for the observed increase in mortality with digoxin. A prospective randomised controlled trial is needed to demonstrate whether or not digoxin is the cause of the increase in mortality.
Action – Clinicians can be reassured that, if they are following NICE guidance, there is no need to change practice. As outlined earlier, beta blockers or rate-limiting calcium channel blockers are the preferred monotherapy in patients with permanent AF who need treatment for rate control. Digoxin should only be used as monotherapy for AF in patients who are predominantly sedentary, or who have co-existing heart failure, or in whom beta blockers and calcium channel blockers are contra-indicated.
Further information on atrial fibrillation is available on the cardiovascular section of NPC.
The Stroke Prevention using an ORal Thrombin Inhibitor in atrial Fibrillation (SPORTIF) III and V studies compared the efficacy of ximelagatran (a thrombin inhibitor subsequently withdrawn by its manufacturer for safety reasons) and warfarin for the prevention of thromboembolism. This substudy to compare survival in users and non-users of digoxin and to determine its potential effect on mortality was planned before the SPORTIF studies were closed and results analysed.
7,329 patients with AF at moderate-to-high risk of stroke were randomised to receive treatment with warfarin or ximelagatran. SPORTIF III was open-labelled. SPORTIF V was double-blind. Patients were not randomised with respect to digoxin use.
53.4% of patients used digoxin at baseline, and these patients had a higher mortality than non-users. Significantly more patients who were taking digoxin died compared with non-users (6.5% vs. 4.1%, respectively; hazard ratio [HR] 1.58, 95% confidence interval [CI] 1.29 to 1.94; P<0.001). After adjusting for baseline risk factors such as age, body mass index, blood pressure, gender, race, smoking status, and co-morbidities, the risk remained unchanged (HR 1.53, 95%CI 1.22 to 1.92; P<0.001)