In this phase II exploratory multi-center randomized placebo-controlled trial involving patients with LCOS treated with dobutamine after CABG surgery, we observed that i.v. ivabradine infusion was associated with (1) a quick, durable, and significant control of HR in the target range of 80 to 90 bpm, (2) a significant increase in SV and SBP, and (3) a higher incidence of POAF and bradycardia.
Hemodynamic effects of intravenous ivabradine
Like its oral counterpart, i.v. ivabradine was found to effectively reduce HR but with a shorter delay of action after treatment onset. This reduction was durable and allowed for a strict control of the HR in a pre-specified target range. In patients with LCOS, dobutamine is commonly used to improve oxygen delivery but induces tachycardia [
25,
26]. It has been argued that most of the improvement in CO after dobutamine was due to the increase in HR [
27]. However, the benefit of improving oxygen delivery may be counterbalanced by the deleterious effect of tachycardia on myocardial oxygen consumption. Indeed, some authors have observed that patients who received dobutamine after cardiac surgery had increased mortality in comparison with propensity-matched patients without dobutamine [
11,
12]. Therefore, there is a sound rationale for controlling HR in CABG patients with post-operative LCOS treated with dobutamine. In the present study, patients were closely monitored to assess the safety of the dobutamine/ivabradine combination. We observed that the decrease in HR was associated with a concomitant increase in SV and CO. This suggests that the prolonged diastolic time improved left ventricular diastolic filling. The resulting improvement in SV was sufficient to compensate the decrease in HR and even increased CO.
These results confirm preliminary animal studies in which dobutamine combined with ivabradine allowed for a simultaneous contractile enhancement and prolongation of diastole, allowing for optimal filling and enhanced CO, in spite of a decrease in HR [
22].
Because of the greater SV, SBP increased in the ivabradine group. MBP and DBP remained unaltered, indicating that systemic and left ventricular perfusion pressures were preserved. Overall, LVSWI increased and atrial pressures (RAP and PCWP) tended to decrease, suggesting a beneficial effect on biventricular end-diastolic congestion. The association of increased systemic blood flow, preserved mean arterial pressure, and reduced venous congestion in the systemic and pulmonary veins attests to an improved hemodynamic status. This corroborates the observations that renal function was maintained and that serum lactate did not increase significantly. We can assume that oxygen delivery was adequate and end-organ function was preserved, indicating that LCOS management was effective. These findings suggest that ivabradine did not interfere with a compensatory tachycardia but rather corrected the unwanted dobutamine-induced acceleration in HR.
These results were consistent and confirmed when assessing hemodynamic parameters at H2 and H3 after treatment initiation when most ivabradine patients reached target HR, emphasizing the causal role of ivabradine rather than the effect of time (as observed in the placebo group).
Indeed, ivabradine represents one of a few i.v. HR-decreasing agents without negative effects on inotropism, allowing its use in post-operative LCOS. Other existing therapeutic agents include digoxin and magnesium. Intravenous digoxin has the major drawback of presenting dose-dependent toxicity, including potential lethal arrhythmias, hyperkaliemias, and vasoplegic shock. Intravenous magnesium, though less deleterious, has a less significant effect on sinus tachycardia. Beta-blockers are part of the routine treatment for most patients with heart failure (all patients but one in our cohort) but, owing to their adverse effect on contractility and their antagonism with dobutamine, can rarely be administered in the post-operative period. Thus, ivabradine appears as a very promising tool offering an original alternative to control HR, even in patients with severely impaired left ventricular function.
Catecholamine administration in patients receiving ivabradine
Epinephrine and norepinephrine usage was similar in the two treatment arms. Dobutamine regimen was increased by attending physicians in six ivabradine patients who remained in LCOS. Although we cannot rule out that controlling tachycardia might have reduced CO in this subgroup, we believe that HR reduction facilitated dobutamine up-titration. The fact that changes in HR, MAP, and SV were similar in the subgroup of patients in whom dobutamine was not increased supports the role of i.v. ivabradine in hemodynamic improvement independently from other concomitant treatments.
Safety issues
As previously described for oral ivabradine in the SHIFT (Systolic Heart Failure Treatment with the I(f) Inhibitor Ivabradine Trial) [
28], the i.v. counterpart was also associated with episodes of arrhythmia and bradycardia. Incident arrhythmias were all supraventricular. Higher dobutamine cumulative dose may be involved in the higher incidence of atrial fibrillation in the ivabradine group, as suggested by the fact that most episodes occurred after ivabradine cessation. These arrhythmias were resolved within a few days of standard treatment. On the other hand, bradycardia, as defined by an HR below 75 bpm, was easily addressed by discontinuing ivabradine. This high threshold for defining “bradycardia” was chosen to avoid leaving patients with LCOS in a low perfusion state. This may have prompted early ivabradine discontinuation in some patients for safety reasons despite the absence of clinical signs of hypoperfusion.
Limitations of the study
The small number of patients is the main limitation to the validity of our observations. The inclusion criteria of this study were very restrictive in order to include a homogeneous population of patients and to avoid confounding factors. As a consequence, the number of inclusions was small but nevertheless appropriate to reach the primary objective: demonstrating the reduction of HR in the ivabradine-treated group. Inferential statistics were performed while taking into account the small sample size, allowing us to observe trends in the secondary hemodynamic endpoints. Consequently, adjusted statistical analysis could not be performed to rule out the confounding effects of concomitant catecholamine administration on hemodynamic changes. However, ad-hoc analyses showed that associated catecholamine administration was similar in the two treatment arms. However, even if subgroup comparison among ivabradine patients (a) with dobutamine increase and (b) without dobutamine increase showed similar results in hemodynamic variations, it is not possible to prove that the effect of ivabradine was independent from dobutamine dosage modifications.
Regarding the primary endpoint, we choose to make comparisons between baseline and the time of study drug cessation rather than at a fixed time point. The timing of study drug interruption differed among patients, attesting to the variability in LCOS severity. Indeed, many factors can influence LCOS (i.e., pre-existing cardiac dysfunction, CPB and aortic clamp durations, and quality of myocardial protection during bypass), which cannot be controlled for. Thus, we deliberately choose to make the comparisons between baseline and the time when LCOS was considered over in each patient (dobutamine weaning). This pragmatic approach allowed us to minimize the administration of study drug, which was important from a safety point of view, and to compare patients when they had reached a similar hemodynamic situation. However, it can be noted that changes between H0 and H2 (common to all patients) were similar to changes between H0 and the time of study drug cessation. Finally, because investigators were not always available when eligible patients underwent surgery, those who were included in this cohort are not strictly consecutive patients.