Update on ivabradine for heart failure

Abstract

Despite dramatic advances in therapy for heart failure (HF) during the past 3 decades, hospitalization and mortality rates remain relatively high. In recent decades, it has become apparent that HF is divisible into two equally lethal but pathophysiologically different sub-classes, the first comprising patients with LV systolic dysfunction [heart failure with reduced ejection fraction (HFrEF)] and the other, approximately equal in size, involving patients with “preserved” systolic function [heart failure with preserved ejection fraction (HFpEF)]. Evidence-based event reducing therapy currently is available only for HFrEF. With the completion of seminal trials of beta blockers, now part of standard therapy for HFrEF, it was apparent that heart rate slowing is an underlying basis of clinical effectiveness of HFrEF therapy. With the discovery of the “f current” that modulates the slope of spontaneous diastolic depolarization of the sino-atrial node, a non-beta blockade approach to heart rate slowing became available. Ivabradine, the first FDA-approved f-current blocker for HFrEF, markedly reduces hospitalizations for worsening heart failure, while also progressively reducing mortality as pre-therapy heart rate increases, and also promotes beneficial left ventricular remodeling, improves health-related quality of life and is effective despite a wide range of comorbidities. The drug is well tolerated and adverse effects are relatively few. Ivabradine represents an important addition to the armamentarium for mitigation of HFrEF.

Introduction

Despite dramatic advances in therapy for heart failure (HF) during the past 3 decades, hospitalization rates remain relatively high and mortality, though significantly mitigated by drugs and devices during the 30-year interval, nonetheless is approximately 50% at 5 years after the diagnosis is established. Therefore, considerable research continues in efforts to develop new therapies to improve clinical outcomes.

The history of drug therapy for patients with HF began more than 2 centuries ago with the application of digitalis for relief of symptoms. The development of diuretics during the past century added to the capacity for symptom relief. However, though digoxin therapy has reduced hospitalizations for worsening HF, neither digoxin nor diuretics has been shown to decrease mortality. The search for therapies to improve natural history moved forward importantly almost 50 years ago with the demonstration that peripheral vasodilatation (by nitroprusside) is effective in improving cardiac output and reducing LV filling pressures in patients with HF, reducing symptoms and disordered hemodynamics. Moreover, these data, together with both experimental and small clinical studies, supported the concept that not only the heart but also the circulatory system and other regulatory systems of the body are involved in the pathophysiology of HF and contribute to its chronicity and outcome. A summary of the pathophysiological and therapeutic, which form the context within which the new drug, ivabradine, was developed and must be understood is reported in detail in recent guidelines on chronic HF [1], [2].

The first therapeutic target identified based on the developing understanding of pathophysiology during the past 50 years was the renin–angiotesin–aldosterone system (RAAS) [3]. Modulation of this system first was attempted with angiotensin converting enzyme inhibitors (ACEI), initially developed for treatment of patients with hypertension and soon adapted for HF as effects of angiotensin on myocytes were recognized. There followed a groundbreaking randomized, placebo-controlled clinical trial (RCT) assessing the effect of captopril and then enalapril (both ACEI) for HF. Survival was improved among those treated with the ACEI rather than placebo. Subsequently, a series of RCTs extended the benefits of ACEI to patients with progressively clinically less severe HF.

Building upon the efficacy of ACEI, another approach to modulating the effect of the renin–angiotensin–aldosterone system, direct angiotensin receptor blockade, was studied when relevant drugs, angiotensin receptor blockers (ARBs), were developed. Several trials demonstrated that these drugs, also, provide survival and hospitalization reduction benefits. Extrapolating further from the benefits of RAAS blockade, aldosterone blockade with mineralocorticoid receptor antagonists (MRAs), such as spironolactone, next were assessed and further reduced mortality and hospitalization risk in patients with mild or severe HF. Finally studies focused on renin, itself, mostly released by the kidneys, which starts the functional cascade of the RAAS axis. However, trials testing the effects of a drug directly counteracting renin activity did not show clear benefits in chronic HF [4], [5]. The result of this long and highly successful trail of research has been that all major components of the RAAS can be modulated pharmacologically. This approach has become standard for treatment of chronic HF, specifically when left ventricular (LV) systolic dysfunction exists, as most recent trials have demonstrated benefit in this population but not among patients with HF without systolic dysfunction (see below).

While RAAS manipulation was under study, exploration of beta blockade for HF also began, based on the observation that excessive beta adrenergic stimulation of myocytes leads to progression of myocyte dysfunction and death, which can be mitigated by adrenergic blockade, and that these effects overcome the adverse effects of negative inotropy with these agents [1], [2], [3]. The negative chronotropic effect of beta blockers, now considered the primary basis for the benefits of this therapy, was not at first recognized as beneficial. Initial studies suggested benefit among patients with HF, leading to RCTs employing a series of beta blocking drugs (metoprolol succinate, bisoprolol, and carvedilol) administered on a background of the already accepted ACEI, as well as diuretics. These beta blocker trials uniformly demonstrated further improvement of survival over that achievable with ACEI. Despite these clear benefits, none of these beta blocker trials demonstrated improvement in quality of life (QoL).

Importantly, during the last decades, it has become apparent that HF is divisible into two equally lethal but pathophysiologically different subclasses, the first comprising patients primarily manifesting LV systolic dysfunction [now known as heart failure with reduced ejection fraction (HFrEF)]. The other [now known as heart failure with preserved ejection fraction (HFpEF)], approximately equal in size, comprises patients with complex cardiovascular alterations not primarely affecting LV systolic function, and specifically not affecting LVEF. The clinical presentation of these two subclasses is quite similar as are hospitalization rates and mortality risk. However, pharmacological therapies are effective in reducing hospitalization and mortality rates for those with HFrEF but, as yet, no drug therapy has been shown to reduce hospitalizations or mortality for those with HFpEF. Thus, the trail of research described above applies to HFrEF, but not to HFpEF. Moreover, though several devices are available or in development to mitigate HF, the predominant (and USFDA approved) devices now include implantable cardiac defibrillators (ICD) and biventricular pacemakers for cardiac resynchronization therapy (CRT). Beside other criteria (QRS duration and QRS morphology for the CRT) the indication for the implantation of these devices for primary mortality prevention requires a LVEF ≤35%. Accordingly, though ICDs would be applicable for HFpEF if a lethal arrhythmia had already occurred (i.e., for secondary prevention) and might be indicated for primary prevention if certain rhythm and electrophysiological criteria were met on formal testing, ICDs would be indicated for primary prevention only in patients with HFrEF with LVEF ≤35%. The distinction between HFrEF and HFpEF is relevant for the discussion of ivabradine, which follows, because this drug has been evaluated and is recommended only for patients with HFrEF.

Beta adrenergic receptors are found in many organs and tissues. Beta blocking drugs have many potential effects. The benefits of beta blockers have been attributed to several properties of these drugs. However, with the publication of a cross-sectional study including survival trials available at the time, it became clear that the magnitude of survival benefit (beta blocker versus placebo) was directly related to the magnitude of heart rate slowing by the beta blocker in comparison with placebo [6]. This initial insight was followed by formal meta-analyses, including patient level as well as study level data, which demonstrated that survival benefits were indeed related to heart rate lowering effects and specifically were unrelated to beta blocker dose [7]. Thus, while other effects of beta blockers may contribute to benefit in HFrEF (possibly including prevention of arrhythmic death), heart rate slowing now is recognized to be among the most important bases of benefit.

Upon this background, within the past 15 years, another possible treatment modality emerged and recently has been added to the standard armamentarium for HFrEF (in addition to ACEI or ARB, beta blockers, MRAs, diuretics for mitigation of fluid overload and symptoms, as well as ICDs, and CRT in appropriate patients) [1], [2]. In 1979, a previously unidentified hyperpolarization activated cyclic nucleotide-gated (HCN) ion channel was discovered in the sinoatrial node. This channel mediates a small late sodium current (I_f), which modulates the slope of spontaneous diastolic depolarization in the sinoatrial node [8,9]. Blockade of this current diminishes the slope and, thus, slows the heart rate. Less than 15 years after discovery of the channel, a drug, ivabradine, highly specific for this channel, had been synthesized in Europe. This agent blocks the f current and slows the heart rate [10]. Ivabradine first was developed for angina prevention, drawing upon the experience with beta blockers and other heart rate slowing modalities [11]. This development was successful and led to approval of the drug for angina prevention in Europe in 2005. With the success for angina prevention, a RCT was mounted to study the utility of heart rate reduction with ivabradine for improvement of natural history in patients with chronic stable coronary artery disease, a benefit that was believed to be associated with beta blockade, but never had been studied. The lack of such assessment in part was attributable to the difficulty in disentangling the clinically measurable bradycardic action of beta blockade from the other less obvious and less easily documented effects. The trial, MorBidity-mortality EvAlUation of The I_f inhibitor ivabradine in patients with coronary disease and left ventricULar dysfunction (BEAUTIFUL), randomized almost 11,000 patients but revealed that heart rate slowing provides no outcome benefit in this patient population [12]. Of note, lacking prior data on which to base a selection, the BEAUTIFUL investigators selected a pre-randomization heart rate of ≥60 bpm as a criterion for entry of patients into the trial. However, it was prespecified that another analysis should be performed for patients entering at heart rate ≥70 bpm. Though the primary outcome was not significantly improved in this subset, a secondary outcome, non-fatal myocardial infarction, was significantly reduced by ivabradine in this subset. Since the primary outcome was not supported, the secondary outcome result must be assessed with great caution and could not be used as a basis for regulatory approval. However, this finding informed the selection of heart rate for subsequent large outcome trials (below), both of which selected pre-therapy heart rate ≥70 bpm as an entry criterion. While BEAUTIFUL was ongoing, and drawing on the by then clearly demonstrated benefit of heart rate slowing in HFrEF with beta blockers, in 2005 a group of investigators, of which we were a part, designed a trial of pure heart rate slowing with ivabradine in patients with HFrEF [13]. The trial, the Systolic Heart failure treatment with the I_f inhibitor ivabradine Trial (SHIFT), began randomization in 2006 and was completed in 2010 [13], [14]. Data from SHIFT were the basis for approval of ivabradine for HFrEF in Europe in 2012 and in the United States in April 2015, and have provided much of what we know about the impact of heart rate slowing in HFrEF. Of note, the drug has not been studied for natural history effects in HFpEF, though the theoretical benefit for this condition is similar to the theoretical benefit of beta blockade.

In total, 6505 analyzable patients were randomized in SHIFT, the largest RCT of HFrEF ever mounted by the time of its completion. Inclusion required clinical stability for at least 4 weeks, LVEF ≤35%, sinus rhythm (since, except when conduction system disease already is present, ivabradine acts virtually only in the sinus node, it does not affect heart rate if atrial fibrillation or other non-sinus rhythm is operative) with heart rate ≥70 bpm. A hospitalization for worsening HF within the 12 months prior to randomization also was required to enrich the population for subjects likely to have outcome events during the trial and most likely to benefit from new drugs to prevent further recurrences. SHIFT was a parallel arm trial in which patients received either ivabradine or placebo on a background of optimized HF therapy, necessarily including beta blockers administered at guidelines-based target doses or maximally tolerated doses if below target. Very importantly, if patients were not receiving the guidelines-based target doses of beta blocker or were not receiving beta blocker at all, the investigator needed to complete a separate case report form to indicate the reason for this deficiency. The guidelines-based target doses of ACEI/ARB and MRA also were expected to be achieved. Based on epidemiological data plus the intuition of the protocol design team, the heart rate goal in SHIFT was 50–60 bpm. To achieve this, ivabradine was begun at 5 mg bid and heart rate was assessed 2 weeks later. If heart rate was higher than 60 bpm, dose was increased to 7.5 mg bid and heart rate was assessed again 2 weeks later, and reduced if heart rate then was <50 bpm. If heart rate was <50 bpm at the initial post-dosing assessment, the dose was reduced to 2.5 mg bid. If, at the next 2 week assessment, rate remained <50 bpm the drug was stopped (an occurrence in only 3% of patients). If heart rate was 50–60 bpm at the first post-dosing assessment, 5 mg bid was maintained. At completion of titration, almost two-thirds of patients were receiving 7.5 mg bid and approximately one-fourth were receiving 5 mg bid. After establishment of the “final” ivabradine dose, patients were followed regularly for clinical assessment and to assure appropriate drug alteration if needed. The trial continued until all subjects had been followed for at least 6 months; average follow-up was approximately 2 years and maximum follow-up was 3.5 years. During this time several pre-specified substudies (echocardiography [15], QoL [16], rhythm monitoring [17], and biomarkers (not yet reported) were performed, and a far larger number of post-hoc analyses were undertaken.

SHIFT demonstrated a highly significant 18% impact of heart rate with ivabradine on the primary outcome, cardiovascular death or first hospitalization for worsening HF [13]. This primary result was largely driven by the 26% reduction in first hospitalization for worsening heart failure. The 9% reduction in cardiovascular mortality did not reach statistical significance. However, a prespecified analysis of mortality segregated at the median heart rate, which was 77 bpm, revealed a highly significant reduction in mortality for those above the median, but no significant result below the median. Based on this finding, the European Medicines Agency ordered a post-hoc assessment of the effect of treatment above heart rate 75 bpm, found significant reduction in cardiovascular mortality and approved the drug for reduction in cardiovascular mortality or hospitalizations for patients with HFrEF and heart rate ≥75 bpm in sinus rhythm. The FDA determined to consider the data only as tabulated according to the protocol (i.e., with heart rate ≥70 bpm), and approved ivabradine for reduction in HF hospitalizations alone.

The primary analysis of SHIFT revealed that the predetermined heart rate goal was highly effective in achieving outcome benefits. Reduction in outcome events was achieved to a progressively greater extent as a function of the final absolute heart rate and also the magnitude of the reduction from pre- to on-therapy [14]. Thus, the maximal event reduction was achieved with a heart rate of 50–60 bpm, which now can be considered the target heart rate with ivabradine (on background of guidelines-based standard therapy, including beta blockers, as in SHIFT). In clinical practice, dose needed to reach the target heart rate may vary among patients but generally follows the distribution pattern in the trial. The recommended titration schedule for clinical application is as performed in SHIFT (described above). SHIFT also demonstrated that relatively little variation in heart rate (and, thus, in ivabradine dose) is observed over time after the initial titration goal is reached [16].

The use of time to first event analyses is routine in clinical trials but, for potentially recurrent events like hospitalizations, this approach commonly precludes more than half the events from inclusion in an efficacy analysis. Therefore, post-hoc, another analysis was performed including all hospitalizations. This revealed that ivabradine use reduced total hospitalizations for worsening heart failure throughout the trial by 25% and, when hospitalizations did recur, significantly lengthened the time interval between successive hospitalizations [18], [19], [20].

The prespecified echocardiographic substudy found that, in the ivabradine group, LV end-systolic volume was reduced compared with baseline and that the ivabradine-associated reduction was sigificantly greater than the change (which was minimal to non-existent) in the placebo group. Similarly, LVEF rose modestly but significantly in the ivabradine group but not at all in the placebo group, a significant difference between the groups [15].

The prespecified QoL substudy, using the Kansas City Cardiomyopathy Questionnaire (KCCQ), as well as the European Quality of Life Scale and the New York Heart Association Functional Classification scheme, revealed a significant improvement in QoL with ivabradine by KCCQ but not with placebo, a significant difference between the groups, and a magnitude of change (>5 units) with ivabradine (but not with placebo) that generally is accepted as being clinically meaningful [16]. Post-hoc analyses have revealed no difference in the benefits of ivabradine among patients of different age [21], with or without left bundle branch block [22], chronic obstructive pulmonary disease [23], diabetes mellitus [24], multiple comorbidities [25], hypertension [26], renal dysfunction [27] or with particularly severe HFrEF (defined as LVEF < 15% or <20% with New York Heart Association Functional Class IV HF) [28], or among those receiving or not receiving MRA [29]. Moreover, there was no significant interaction of ivabradine-related benefits and the extent to which target doses of beta blockers were employed [30]. No tendency toward ventricular arrhythmias or particular conduction abnormalities was seen in SHIFT substudies (ivabradine has little or no effect on the normal AV node or the Bundle of His or the bundle branches) [17], though in small studies unrelated to SHIFT and involving patients who already had conduction abnormalities (first- or second-degree heart block or left or right bundle branch block), rare cases of heart block have occurred, leading to a USFDA recommendation for back-up pacemaker implantation in such patients, as well as in those with sino-atrial block or sick sinus syndrome. [These data have not appeared in peer-reviewed publications but are the basis of the Corlanor (ivabradine) Prescribing Information (Amgen) approved by the FDA.]

Since the completion of SHIFT, another trial, study assessing the morbidity–mortality benefits of the I_f inhibitor ivabradine in patients with coronary artery disease (SIGNIFY) [31], has revisited the issue of heart rate slowing for chronic stable CAD. This study randomized more than 19,000 subjects and differed from BEAUTIFUL and SHIFT in that LVEF needed to be >40% for inclusion (the average LVEF was 55%) and HF needed to be absent. Like BEAUTIFUL, this trial found no benefit of heart rate slowing for survival or myocardial infarction frequency. Since the trial did not show a significant improvement in the primary hypothesized benefit, by consensus secondary analyses are not appropriate. However, almost two-thirds of the patients in SIGNIFY had angina at entry and were prespecified to be analyzed separately. In that subgroup there was an 18% increase in the primary outcome event among patients who received ivabradine versus those who received placebo. In BEAUTIFUL, a similar assessment was performed in patients who had entered the trial with angina and found a significant reduction in primary outcome events with ivabradine in this subgroup, but the analysis was post-hoc and thus cannot be a basis for firm inferences. Interestingly, investigating the antianginal effect of ivabradine in the SIGNIFY angina subgroup, the results were consistent with the symptom-reducing benefit of ivabradine in patients with stable angina pectoris. There were improvements in CCS angina class versus placebo (p = 0.01) and a trend toward lower incidence of elective coronary revascularisations (p = 0.058) [31]. In addition, a substudy of SIGNIFY evaluated the effects of ivabradine on angina-related quality of life in 4187 patients. The results showed a consistent improvement of self-reported quality of life parameters related to angina, in terms of angina frequency and disease perception [32].

However, despite the difference in the primary outcome results of the angina subgroup in SIGNIFY from that of post-hoc analysis of the angina subgroup in BEAUTIFUL, the results of SIGNIFY caused concern for the use of ivabradine in patients with HFrEF, since, in general, approximately two-thirds of such patients have coronary artery/ischemic heart disease as the etiology underlying their HF. Therefore, a post-hoc study has been completed among patients who entered SHIFT with angina, a group which comprised approximately one-third of the total study population. That study revealed no differences in outcome among those with angina as compared with those without angina and as compared with the overall SHIFT population [33]. In addition, when analyzed for the SIGNIFY outcomes, the SHIFT patients with angina tended to have better outcomes on ivabradine than on placebo, emphasizing the differences in the effects of heart rate slowing as a function of underlying myocardial performance, i.e., providing greater benefit in patients with dysfunctional hearts and less or no benefit in those with normal cardiac function.

In summary, the benefit of heart rate slowing with ivabradine on mortality and major morbidity has been proven unequivocally for patients with HFrEF. However, as with all drugs, application in individual patients requires consideration of possible adverse responses. These primarily involve bradycardia (not surprising for a heart rate slowing drug) though, in SHIFT, symptomatic bradycardia (defined by the investigator, not by any protocol specification) occurred in only 3% of patients. Additionally, atrial fibrillation, common in HFrEF (and occurring in 6% of patients on placebo) occurs more commonly on ivabradine (8% of patients in SHIFT). The reason for this discrepancy is unclear and, fortunately, among those with atrial fibrillation in SHIFT, strokes were not encountered. In SHIFT, blood pressure rose more on ivabradine than on placebo (by approximately 2 mmHg). Given the improvement in cardiac function with the drug, the relatively small average increase in blood pressure may reflect a benefit rather than a risk. However, among patients who entered SHIFT with a history of hypertension, the absolute blood pressure reached on drug was deemed “inadequately controlled” by individual investigators in 7.1% of patients receving ivabradine and 6.1% of patients receiving placebo; ivabradine-associated outcome benefits were maintained in these patients. (Once again, these data do not appear in peer-reviewed publications but were part of the regulatory submissions to the EMA and FDA.) An unusual adverse effect, “phosphenes” (flashing scotomata most commonly induced by changing light intensity), occur more frequently on ivabradine than on placebo. They often disappear spontaneously, are not associated with permanent eye damage and do not impair driving. In SHIFT, phosphenes led to withdrawal of the drug in less than 1% of patients receiving ivabradine. These events are believed to relate to the effect of ivabradine on HCN channels, known as “h channels,” in the retina that are similar to the HCN channels in the SA node. Finally, animal studies have revealed that, if given during pregnancy, resulting fetuses may have cardiac malformations. Therefore, ivabradine should not be used in pregnant women.

Recent experience with ivabradine-mediated heart rate slowing for HFrEF indicates that this is highly effective in reducing hospitalizations for worsening HF when administered as an adjunct to beta blockade together with other standard HFrEF therapies in patients with LVEF ≤35% who are in sinus rhythm with heart rate ≥70 bpm. Indeed, to prevent one initial hospitalization for worsening heart failure, ivabradine needs to be administered to only 27 eligible patients for 1 year; to prevent recurrent hospitalizations, the number needed to treat (NNT) for 1 year is 14 [34]. The data also indicate that addition of ivabradine reduces cardiovascular mortality risk among those with HFrEF and heart rates ≥75 bpm in sinus rhythm, improves health-related QoL, induces cardiac remodeling to improve myocardial function, and does not induce ventricular dysrhythmias.

Several opportunities exist for future research. In the HF domain ivabradine has been studied only for chronic HFrEF. Its value for acute HF is not known. This has led to some uncertainty about whether the drug should be continued in patients admitted to hospital for worsening HF who already are receiving the drug. In SHIFT, ivabradine was not stopped in most patients admitted to hospital and results do not seem to have been adversely affected but, since randomized withdrawal was not performed, it is not rigorously known whether any adverse effect of continuation may occur. Post-hoc analyses were also performed in patients with severe HF (NYHA IV, LVEF ≤ 20% or 15%) showing the same benefit of ivabradine as observed in the overall study population [28]. Data do not yet exist regarding outcomes with ivabradine in HFpEF, a subject of particular interest. The basis of the relation of the drug to atrial fibrillation requires further exploration. Other issues exist for which further research also would be of interest. However, overall, the addition of pure heart rate slowing to the armamentarium of those managing patients with HFrEF thus far has provided distinct advantage.