Mortality in the Survival With ORal D-sotalol (SWORD) trial: why did patients die?

The Survival With ORal D-sotalol (SWORD) trial tested the hypothesis that the prophylactic administration of oral d-sotalol would reduce total mortality in patients surviving myocardial infarction (MI) with a left ventricular ejection fraction (LVEF) of < or = 40%. Two index MI groups were included: recent (6 to 42 days) and remote (> 42 days) with clinical heart failure (n = 915 and 2,206, respectively). The trial was discontinued when the statistical boundary for harm was crossed (RR = 1.65; p = 0.006). All baseline variables known to be associated with mortality risk (e.g., LVEF, heart failure class, age) as well as variables related to torsades de pointes (e.g., time from beginning of therapy, QTc, gender, potassium, renal function, dose of d-sotalol) were assessed for interaction of each variable with treatment assignment, computing RR and 95% confidence interval (CI) from Cox regression models. The d-sotalol-associated mortality was greatest in the group with remote MI and LVEFs of 31% to 40% (RR = 7.9; 95% CI 2.4 to 26.2). Most variables known to be associated with torsades de pointes were not differentially predictive of d-sotalol-associated risk, except female gender (RR = 4.7; 95% CI 1.4 to 16.5). These findings suggest that (1) most of the d-sotalol-associated risk was in patients remote from MI with a LVEF of 31% to 40%; comparable placebo patients had a very low mortality (0.5%); and (2) very little objective data supports torsades de pointes or any specific proarrhythmic mechanism as an explanation for d-sotalol-associated mortality risk.