Introduction

Heart failure (HF) is a chronic, progressive disease process, marked by high hospitalization rates [1], substantial resource utilization [2], and adverse impacts on health-related quality of life [3]. The prevalence of HF has been estimated at 26 million worldwide and this global burden is expected to rise with time related to increasing age and comorbidities [4]. HF is the most frequent reason for hospitalization in older persons in the USA and Europe, contributing to over 1 million admissions annually [5]. Roughly, half of patients with chronic HF have reduced ejection fraction (≤40 %, HFrEF). Stepped pharmacologic and device therapy to improve outcomes in this population is well described by US [6••] and European [7••] treatment guidelines and emphasizes the central role of renin-angiotensin-aldosterone-system (RAAS) antagonists, β-adrenergic receptor blockers, and mineralocorticoid receptor antagonists (MRA).

Since the publication of the SOLVD (Studies of Left Ventricular Dysfunction) trial [8], treatment with angiotensin-converting enzyme (ACE)-inhibitors has been regarded as a cornerstone therapy for patients with heart failure with reduced ejection fraction (HFrEF) across the full spectrum of disease severity. However, the publication of the PARADIGM-HF (Prospective comparison of ARNi with ACEi to Determine Impact on Global Mortality and Morbidity in Heart Failure) trial, suggesting an incremental 16 % reduction in all-cause mortality with composite angiotensin receptor-neprilysin inhibition (ARNi) using sacubitril/valsartan compared to ACE-inhibition with enalapril, has raised the question of whether ACE-inhibition alone is adequate for many HFrEF patients resembling those enrolled in the study. Recent clinical guideline updates [6••, 7••] have accordingly suggested that clinicians should consider switching selected patients with HFrEF who are currently treated with ACEi or angiotensin II receptor blockers (ARB) to combination sacubitril/valsartan; however, ambiguity in the guidelines regarding precisely which patients to transition and discrepancies between published recommendations from the American College of Cardiology/American Heart Association (ACC/AHA) and European Society of Cardiology (ESC) in this regard leave considerable uncertainty for clinicians about whom to select for treatment with sacubitril/valsartan and when to make the transition in clinical practice. This uncertainty has fed a therapeutic inertia, with many electing to defer transition in eligible patients despite acknowledged benefits [9]. In this review, we discuss the evidence supporting use of ARNi in preference to ACEi/ARB in selected patients with HFrEF and identify a strategy for selection of appropriate patients for transition to ARNi in clinical practice.

The Road to NEP Inhibition

Neutral endopeptidase (NEP) is responsible of the degradation of a number of vasoactive peptides, including natriuretic peptides (NP), adrenomedullin, bradykinin, endothelin-1, and substance P [9]. Early experience with the NEP inhibitor, candoxatril, suggested a dose-dependent rise in circulating levels of NP, but no discernible effect on blood pressure or systemic vascular resistance [17]. This lack of hemodynamic effect, thought to be related in part to increased circulating levels of angiotensin II (also a substrate for NEP) fueled subsequent interest in blockade of NEP in tandem with the RAAS system. In the IMPRESS (Inhibition of Metalloprotease by Omapatrilat in a Randomized Exercise and Symptoms Study in Heart Failure) trial of 573 subjects with symptomatic HFrEF, treatment with omapatrilat, a dual ACE/NEP inhibitor, compared with enalapril was not associated with incremental benefit with regard to the primary efficacy endpoint of exercise duration; however, a beneficial trend with regard to the composite of death or HF hospitalization narrowly missed the margin for statistical significance (hazard ratio [HR] 0.53, 95 % confidence interval [CI] 0.27–1.02) [10]. These preliminary findings laid the foundation for the design of the OVERTURE (Omapatrilat Versus Enalapril Randomized Trial of Utility in Reducing Events) trial [11], which randomly assigned 5770 subjects with chronic HF, EF ≤ 30 %, New York Heart Association (NYHA) II–IV symptoms, and a prior HF hospitalization to treatment with omapatrilat or enalapril. In that study, no incremental benefit of omapatrilat with regard to the primary endpoint of death or hospitalization for HF requiring intravenous therapy was observed, and omapatrilat-treated patients experienced higher rates of symptomatic hypotension and angioedema compared with patients randomized to enalapril [11]. This heightened risk of angioedema (roughly three-fold that of enalapril) with composite ACE/NEP inhibition using ompatrilat was confirmed in a subsequent phase III study of over 25,000 patients with hypertension [12], precluding regulatory approval of the drug by the Food and Drug Administration. In retrospect, some speculated that the lack of benefit of omapatrilat in OVERTURE may have been related to incomplete inhibition of neprilysin due to once daily dosing; as well from a mechanistic standpoint, the heightened risk of angioedema was thought related to increased circulating levels of bradykinin due to simultaneous inhibition of three key proteases (ACE, aminopeptidase, and neprilysin) responsible for its degradation. These observations led to exploration of alternate strategies for inhibition of the RAAS system in tandem with NEP, culminating in the development of LCZ696 (sacubitril/valsartan), a first-in-class ARNi. This compound, a crystalline complex comprised of the neprilysin inhibitor prodrug, sacubitril, and the ARB, valsartan, dissociates into its component active moieties shortly after oral ingestion and was designed to have lower risk for angioedema due to inhibition of only one pathway responsible for bradykinin breakdown.

Clinical Trials of ARNi in Heart Failure

PARADIGM-HF

The PARADIGM-HF trial [13••] was a phase III double-blind randomized controlled trial designed to test the hypothesis that treatment with LCZ696 (sacubitril/valsartan) would be superior to enalapril in reducing the primary composite endpoint of cardiovascular death or HF hospitalization among patients with HFrEF. Eligible subjects included patients with chronic HF, NYHA class II–IV symptoms, and EF ≤ 40 % (revised to ≤35 % by a protocol amendment roughly 1 year into the trial) who were treated with guideline-directed medical therapy including an ACEi or ARB equivalent to enalapril 10 mg daily for ≥4 weeks. Subjects were also required to have elevated NP levels at the time of screening (B-type natriuretic peptide [BNP] ≥150 pg/mL [≥100 pg/mL if hospitalized in prior 12 months] or N-terminal pro-B-type natriuretic peptide [NT-proBNP] ≥600 pg/mL [≥400 pg/mL if hospitalized in prior 12 months]). Prior to randomization, all subjects underwent sequential, single-blind, active run-in periods to ensure tolerability of both enalapril and LCZ696 at target doses. Of 10,521 patients initially screened for entry, 8442 subjects with serum potassium ≤ 5.4 mmol/L, eGFR ≥ 30 mL/min/1.73 m2 of body surface area, and systolic blood pressure ≥ 95 mmHg and no symptomatic hypotension, postural symptoms, or other adverse events precluding continued participation at the conclusion of the run-in were randomly assigned to treatment with LCZ696 200 mg twice daily (sacubitril/valsartan 97/103 mg twice daily) or enalapril 10 mg twice daily [13••]. The baseline characteristics of the enrolled cohort were typical for a HFrEF population, with median age 63.8 years, systolic blood pressure 122 mmHg, median EF 29.5 ± 6.2 %, median BNP 253 pg/mL, and median NT-pro-BNP 1613 pg/mL. Seventy percent of the enrolled subjected reported NYHA II symptoms, with 26 % reporting NYHA III. The population was well-treated according to guidelines with high utilization of β-blockers (93 %) and MRA (56 %). Utilization of device therapy (∼15 % with implantable cardioverter-defibrillators or cardiac resynchronization therapy devices) was consistent with rates of utilization in other contemporary international trials of HFrEF [14]. After a median of 27 months, the trial was terminated early for efficacy by the Data Safety and Monitoring Committee due to overwhelming benefit of LCZ696 with regard to both the primary composite endpoint and cardiovascular death alone. Compared with enalapril, LCZ696 reduced the primary composite of cardiovascular death or HF hospitalization by 20 % (HR 0.80, 95 % CI 0.73–0.87), cardiovascular death by 20 % (HR 0.80, 95 % CI 0.71–0.89), and all-cause mortality by 16 % (HR 0.84, 95 % CI 0.76–0.93). LCZ696 was associated with an increased risk of symptomatic hypotension, but was associated with less frequent rates of renal dysfunction and hyperkalemia compared with enalapril. Rates of non-serious angioedema were also non-significantly increased by LCZ696 compared with enalapril (19 vs. 10 cases), but this did not reach statistical significance [13••].

The benefits of LCZ696 over enalapril with regard to cardiovascular mortality appeared to be driven by comparable impacts on reduction in both sudden death (35 % of deaths) and deaths due to worsening HF (21 % of deaths) and were apparent in patients with and without implantable cardioverter-defibrillators at baseline [15]. Benefits in PARADIGM-HF were consistent across all pre-specified subgroups examined and appear to be robust across the spectrum of age [16], EF [17], and major comorbidities such as diabetes/pre-diabetes [18]. Detailed analyses of HF hospitalizations highlight that the benefits of LCZ696 were apparent early in the study, with reductions in HF hospitalizations observed within 30 days of initiation of LCZ696 [19]. Substantial effects were also noted on HF progression, with reduction in the emergency department utilization, cumulative burden of HF hospitalizations, intensive care unit utilization, need for inotropic support, and necessity for implantation of HF device or cardiac transplantation [19]. Actuarial analysis of estimated life expectancy has suggested that treatment with LCZ696 in preference to enalapril for PARADIGM-like subjects would be expected to afford a survival benefit of 1 to 2 years across a range of age groups [20].

HFpEF and Future Clinical Studies

Limited data are available to support a role for ARNi outside of those with HFrEF. In a phase II study in 301 patients with HFpEF, Solomon and colleagues demonstrated that LCZ696 significantly improved NP levels, left ventricular filling pressures, and functional class compared with valsartan alone [21]. These promising initial findings have fueled the design of PARAGON-HF (Efficacy and Safety of LCZ696 Compared to Valsartan, on Morbidity and Mortality in Heart Failure Patients With Preserved Ejection Fraction; ClinicalTrials.gov NCT01920711) which will definitively evaluate the impact of sacubitril/valsartan on morbidity and mortality in the HFpEF population. Other large trials enrolling patients with acute HF and those with HF post-myocardial infarction (ClinicalTrials.gov NCT02226120 and ClinicalTrials.gov NCT02468232) are underway and promise to add to the clinical experience evaluating the safety and tolerability of the ARNi class in chronic HFrEF.

Drug Approval and Guideline Recommendations

The striking results from the PARADIGM-HF trial have support the approval for use of the combination of sacubitril/valsartan in the USA and Europe. Sacubitril/valsartan is available in three drug doses (24 mg/26 mg, 49 mg/51 mg, and 97 mg/103 mg). Guideline committees in the USA [6••] and Europe [7••] have recently released updated treatment recommendations for stage C chronic HFrEF, strongly supporting use of sacubitril/valsartan in this population as a novel mechanism to inhibit the RAAS axis (Table 1). In line with the PARADIGM-HF experience, the US guidelines recommend that ARNi therapy should be considered in preference to ACEi/ARB in patients with chronic HF, NYHA class II–III symptoms, and EF ≤ 40 %, who are on a stable dose of ACEi/ARB [6••]. The European guidelines apply more stringent criteria for ARNi consideration, requiring persistent symptoms and EF ≤ 35 % despite treatment with optimal or maximally tolerated doses of ACEi, β-blockers, and MRA [7••]. The European guidelines further require NP thresholds similar to those used in PARADIGM-HF [7••]. The US guidelines specify that sacubitril/valsartan should not be used concomitantly with an ACEi or within 36 h of last dose of ACEi and should be avoided in patients with a history of angioedema (class III) [6••].

Table 1 Summary of updated treatment guidelines for use of sacubitril/valsartan in stage C chronic HFrEF
Full size table

Perceived Barriers to Uptake and Clinical Implementation of Sacubitril/Valsartan

Despite early, robust, and consistent benefits of sacubitril/valsartan in PARADIGM-HF and regulatory approval in the USA and Europe, clinicians have faced challenges in translating these trial-based findings into clinical practice, and therapeutic uptake has been slower than anticipated. It is likely that therapeutic inertia in prescription of ARNI may be related in part to cost considerations, reluctance to transition “stable” patients, and concerns regarding the generalizability of the PARADIGM results to “real-world” practice [22]. We consider each of these in turn in the discussion that follows.

Cost Considerations

The estimated wholesale price of twice-daily dosing of sacubitril/valsartan is $12.50 per day, translating to approximately $4500 annually for each patient [23]. Out-of-pocket costs and co-payments vary widely by insurer, and prior authorization is frequently required. Recent data suggest that despite these costs, substitution of sacubitril/valsartan for enalapril in PARADIGM-like patients is likely to be cost-effective. In a recent formal cost-effectiveness analysis, Markov models were constructed from relative event rates derived from PARADIGM-HF [24]. Cost-effectiveness of sacubitril/valsartan combination therapy was $50,959 per quality-adjusted life year gained over a lifetime of therapy [24]. These data are consistent with the results of a cost-effectiveness analysis conducted by the PARADIGM-HF investigators themselves and suggest a value comparable to that of other commonly accepted cardiovascular therapies [25].

“Stability” on ACEi/ARB Therapy

Gaps in utilization of guideline-directed medical therapy in clinical practice may in part be related to reluctance (on the part of both clinicians and patients) to intensify medical treatment in the face of apparent clinical stability. High event rates in well-treated, NYHA II and III patients enrolled in the control arm of PARADIGM-HF, however, suggest that many minimally symptomatic HFrEF patients remain at high-risk for experiencing short- and long-term events. Enalapril-treated patients in PARADIGM-HF had a 15 % event rate for the primary composite and an 8 % rate of cardiovascular death at 1-year, highlighting substantial residual risk in this predominantly NYHA II population. Although most patients in PARADIGM-HF had NYHA class II functional class, the enrolled population was representative of a broad spectrum of risk as defined by the MAGGIC (Meta-Analysis Global Group in Chronic Heart Failure) and EMPHASIS-HF (Eplerenone in Mild Patients Hospitalization and Survival Study in Heart Failure) risk scores [26]; the benefits of LCZ696 over enalapril were consistent across all risk categories, including those at the extremes of the risk spectrum [26]. Apparent reductions in HF hospitalization within the first 30 days following randomization and early reductions in sudden death (which was the first clinical event for many enrolled subjects) underscore the potential costs of delaying the transition in eligible patients. Accordingly, a strategy of waiting for clinical deterioration may be inadvisable.

Concerns Related to Safety and Tolerability

The 8399 subjects in PARADIGM were a subset of 10,521 patients initially screened for randomization. This nearly 20 % rate of non-completion (∼10 % during each run-in phase) was predominantly attributed to adverse effects during titration of either enalapril or sacubitril/valsartan to the target doses required for randomization. The highly selected nature of the population randomized in PARADIGM-HF has raised concerns that the observed rates of study drug discontinuation in PARADIGM may overestimate the tolerability of the drug initiation in real-world clinical practice. In considering patients for transition to sacubitril/valsartan, it is important for clinicians to take heed of the detailed inclusion criteria for the study. Patients enrolled in PARADIGM-HF required screening serum potassium ≤ 5.4 mmol/L, eGFR ≥ 30 mL/min/1.73 m2 of body surface area, and systolic blood pressure ≥ 95 mmHg. Rates of symptomatic hypotension in LCZ696 compared with enalapril-treated patients in PARADIGM-HF (14 vs. 9.2 %) were observed in the context of a relatively sturdy blood pressure at the time of enrollment (average systolic blood pressure 122 mmHg) [27]. Accordingly, for patients with lower blood pressure or those treated with lower doses of ACEi/ARB (lower than doses equivalent to enalapril 10 mg daily) in whom tolerability is a question, initiation of sacubitril/valsartan at lower doses than those used in the PARADIGM may be appropriate. Indeed, the package insert recommends a lower starting dose (24 mg/26 mg twice daily) in patients who are not currently taking or who are on lower doses of ACEi/ARB (e.g., equivalent to less than enalapril 10 mg daily).

Generalizability of PARADIGM-HF

Although baseline characteristics of patients included in PARADIGM-HF are similar to those enrolled in other large, contemporary clinical trials of HFrEF, certain subsets of patients were not well represented in PARADIGM-HF including those with intracardiac devices (∼15 %) and black patients (5 %). Reassuringly, the clinical benefits of sacubitril/valsartan were robust in patients with and without intracardiac devices and were seen with regard to both sudden and HF-related deaths [15]. Although rates of life-threatening angioedema were very low in PARADIGM-HF and there was no clear variation in benefit of sacubitril/valsartan by race or ethnicity, clinicians should exercise caution during use in black patients who may be at higher background risk for angioedema. Notably, however, black patients with HFrEF are at higher overall risk [28], have potentially decreased therapeutic response to ACEi [29], and may stand to derive greater absolute benefit from transitioning to ARNi therapy.

Selecting Optimal Candidates for ARNi Initiation: Who and When?

Which patients?

Based on the accumulated evidence, we believe that many patients with chronic HFrEF (≤40 %) who remain symptomatic (NYHA II–III) despite treatment with ACEi/ARB and a β-blocker should be considered for transition to sacubitril/valsartan. Patients with prior history of intolerance to even low-dose ACEi/ARB, particularly those with prior angioedema, should not be transitioned, given the high potential for adverse effects. Ambiguity remains with regard to the optimal timing of transition to sacubitril/valsartan in relation to optimization of the residual HF regimen, and in particular, whether patients on low ACEi/ARB doses (≤equivalent doses to enalapril 10 mg daily) or low β-blocker doses should be uptitrated prior to the switch. Lack of variation in the benefits of sacubitril/valsartan in PARADIGM-HF according to background use of MRA suggests that initiation of sacubitril/valsartan need not follow sequentially on initiation of spironolactone or eplerenone. Moreover, at this juncture, there are limited data regarding safety and efficacy of ARNi in patients who are asymptomatic, those who are ACEi- or ARB-naïve, and those currently hospitalized for HF. The balance of safety and efficacy in HFpEF awaits the results of the PARAGON-HF trial. Consistent with the overall PARADIGM-HF exclusion criteria, patients are not appropriate ARNi candidates if they demonstrate symptomatic hypotension, baseline systolic blood pressures less than 95 mmHg, baseline-estimated glomerular filtration less than 30 mL/min/1.73 m2.

Practical Considerations in Transitioning Patients from ACEi to ARNi

Given concerns for angioedema in the overlap period, guidelines highlight the importance of maintaining a gap between discontinuation of ACEi and initiation of sacubitril/valsartan. Patients should be instructed to hold the ACEi for at least 36 h prior to initial dosing with ARNi to allow for sufficient washout to attenuate potential for angioedema [6••]. Since the risks of angioedema are lower in the transition from ARB to ARNi, a similar gap may not be necessary. After successful washout, most patients should be initiated on sacubitril/valsartan 49 mg/51 mg twice daily. After 2 to 4 weeks of therapy, the dose should be doubled to target doses of 97 mg/103 mg twice daily, as tolerated. A lower starting dose (24 mg/26 mg twice daily) is recommended in patients who are (1) ACEi/ARB-naïve; (2) on lower doses of ACEi or ARB; (3) have severe renal impairment; or (4) have moderate liver dysfunction. The recently completed double-blind TITRATION trial [30] tested the safety and tolerability of initiation of sacubitril/valsartan using two titration strategies (3 vs. 6 weeks) in 540 patients with HFrEF (≤35 %). The trial demonstrated that both titration approaches resulted in >75 % tolerance of target maintenance doses of sacubitril/valsartan through 12 weeks of therapy. A more gradual titration strategy may be more appropriate for patients on lower doses of ACEi at baseline based on these data [30]. Dose reduction of ARNi, or complete interruption and transition back to prior ACEi/ARB therapy, may be necessary in patients who develop unacceptable side effects during this initial titration phase.

Conclusion

Patients with HFrEF and NYHA II symptoms have substantial residual risk for HF progression and death. In patients with NYHA II-III symptoms and chronic, stable HFrEF who are treated with ACEi/ARB, a transition to sacubitril/valsartan should be considered even in the absence of apparent clinical deterioration. Initiation at low dose and more gradual titration to ensure tolerability, mimicking the run-in period in PARADIGM-HF, may be relevant for those on lower ACEi/ARB doses or those with lower blood pressures. Patients with advanced chronic kidney disease, those with advanced HF who cannot tolerate ACEi/ARB due to symptomatic hypotension or worsening renal function/hyperkalemia, and those with marginal systolic blood pressure less than 95 mmHg are likely not candidates for sacubitril/valsartan. Patients with acute decompensated HF, de novo or newly diagnosed HF, and patients who are ACEi/ARB-naïve represent groups where the balance of efficacy and safety are not well established. Data from the phase III PARAGON-HF trial are expected in 2017 to inform the utility of sacubitril/valsartan in HFpEF.