Background: Flosequinan is a direct-acting vasodilator that exerts beneficial hemodynamic effects and improves the exercise tolerance of patients with heart failure. However, a multicenter trial has demonstrated that long-term administration of flosequinan is associated with increased mortality rate. To explore a possible role of neurohormonal activation on this adverse outcome, we conducted a substudy to examine the plasma levels of 3 neurohormonal systems known to have prognostic implications in heart failure.
Methods: At 20 participating Canadian centers, paired plasma samples at baseline and 1 month after randomization for the measurement of N-terminal atrial natriuretic peptide (N-ANP), angiotensin II, and norepinephrine were obtained in 234 patients (114 receiving flosequinan and 120 receiving placebo).
Results: Treatment with flosequinan was associated with a decline in median plasma N-ANP levels (2139 pmol/L at baseline to 1625 pmol/L at 1 month [P =. 0001]), unchanged plasma angiotensin II levels (40 to 50 pmol/L [P =. 2700]), and a modest increase in plasma norepinephrine levels (391 to 439 pg/mL [P =.002]). These changes were not observed in the placebo group. Multivariate analysis of baseline variables revealed that plasma norepinephrine level predicted patients' death whereas analysis incorporating both baseline and 1-month variables indicated that plasma N-ANP level predicted patients' death. Furthermore, in the flosequinan group, a significant decline in plasma N-ANP level was observed in the survivors only. On multivariate analysis of baseline and 1-month data, the increase in plasma norepinephrine level did not predict the increase in heart rate associated with the use of flosequinan, suggesting that the 2 effects might be mediated by separate mechanisms.
Conclusions: Results of our study demonstrate that in patients with severe heart failure, baseline norepinephrine level predicts death. Flosequinan increases plasma norepinephrine level and heart rate in these patients through mechanisms that override its beneficial hemodynamic effects. Our study reinforces the concept that the direct actions of a pharmacologic agent may have a more profound impact on the prognosis of these patients than the hemodynamic effects.