Background. The Vasodilator-Heart Failure Trial (V-HeFT) data base pro
vides information on the mechanism of death of male veterans entered i
nto two trials that evaluated the effect of vasodilator therapy on sur
vival in heart failure. Methods and Results. Men aged 18-75 years with
heart failure were recruited at 13 Department of Veterans Affairs Med
ical Centers. In V-HeFT I, 283 of 642 patients (44%) died during follo
w-up (average, 2.3 years), and in V-HeFT II, 285 of 804 randomized pat
ients (35.5%) died during follow-up (average, 2.5 years). Mechanism of
death was established centrally using a standardized classification.
In V-HeFT I, 124 of the 283 deaths (43.8%) were sudden with no worseni
ng of symptoms; in V-HeFT II, 104 of the 285 deaths (36.5%) were sudde
n. An average of 31.5% of the deaths (31.4% and 31.6%, respectively) i
n the two trials was due to pump failure. The proportion of sudden dea
ths that occurred without worsening of symptoms was similar in patient
s with and without ischemic heart failure. Sudden deaths tended to occ
ur earlier and pump failure deaths later in both V-HeFT studies. There
was a trend for a lower percentage of cardiac deaths from pump failur
e and a higher percentage from sudden death in subgroups with higher p
eak exercise oxygen consumption (VO2), higher ejection fraction, and l
ower plasma norepinephrine levels. The proportion of deaths that occur
red suddenly was similar in placebo, prazosin, and hydralazine plus is
osorbide dinitrate treatment groups but was significantly lower in the
enalapril treatment group. In V-HeFT I, measures of cardiac function
and VO2 predicted pump failure death and sudden death. In V-HeFT II, V
O2 and cardiothoracic ratio were independent predictors of all-cause d
eaths and pump failure deaths; only ejection fraction was an independe
nt predictor of both pump failure and sudden death. Conclusion. Althou
gh mechanistically distinct terminal events can be identified in patie
nts with heart failure and physiological measurements can provide some
insight into the risk of these disparate events, sudden death and pum
p failure death both appear largely to be linked to the severity of ca
rdiac dysfunction and symptoms. Strategies to identify individuals for
selective preventive therapy are not yet practical.