Clinical Trials in Acute Heart Failure: Beginning of the End or End of the Beginning?

Despite numerous promising clinical development programmes, there have been relatively few major breakthroughs in the management of heart failure (HF) in the inpatient setting and the cornerstone of therapy remains intravenous (i.v.) diuretics, vasodilators, and less commonly inotropes.1, 2 However, recent developments, including the publication of TRUE-AHF3, 4 and the release of the topline results of RELAX-AHF-2 (https://www.novartis.com/news/media-releases/novartis-provides-update-phase-iii-study-rlx030-serelaxin-patients-acute-heart), challenge our understanding of the significance of cardiac and/or end-organ damage and in-hospital worsening HF (WHF) to the pathophysiological progression of this syndrome.5, 6 In addition, there remains no convincing evidence that early intervention with a short-term infusion of a novel vasoactive agent capable of relieving haemodynamic congestion has the potential to prevent and/or reverse cardiac, renal, and liver injury and facilitate improvement in clinical status and long-term outcomes. TRUE-AHF was a randomized, double-blinded, placebo-controlled trial which evaluated the short- and long-term efficacy and safety of a 48-hour infusion of ularitide, a novel recombinant natriuretic peptide with vasodilatory, diuretic, and natriuretic properties, in addition to standard of care.3 Patients presenting for an unplanned hospitalization or emergency room visit for WHF were eligible for enrollment if they had dyspnoea at rest despite receiving at least 40 mg of i.v. furosemide (or equivalent dosing of another loop diuretic), evidence of HF on chest X-ray, a B-type natriuretic peptide (BNP) >500 pg/mL or N-terminal proBNP (NT-proBNP) >2000 pg/mL, and a systolic blood pressure >115 mmHg. The co-primary endpoints were (i) a hierarchical clinical composite assessing signs and symptoms of HF at 6, 24, and 48 hours or death occurring within the first 48 hours, and (ii) cardiovascular mortality for the duration of the trial. A total of 2157 patients were randomized in a 1:1 fashion to either ularitide or matching placebo initiated a median of 6.1 hours after initial presentation.3, 4 Ularitide, compared to placebo, led to a significant improvement in cardiac distension and intravascular congestion as measured by a 47% greater decrease in NT-proBNP levels at 48 hours (P < 0.001). In addition, the incidence of in-hospital WHF at 48 hours was lower among patients treated with ularitide vs. placebo (5.1% vs. 8.1%, P = 0.005). However, there was no difference between treatment arms in cardiac injury as measured by high sensitivity cardiac troponin T (P = 0.70) or the clinical composite endpoint over the same timeframe (P = 0.82). Finally, cardiovascular mortality for ularitide vs. placebo was 22.1% vs. 20.7% (hazard ratio 1.03, 95% confidence interval 0.85–1.25; P = 0.75) over a median follow-up of 15.0 months. Cardiac injury, as evidenced by elevated or rising levels of cardiac troponin, is thought to play a central role in the pathophysiological progression of HF irrespective of underlying aetiology (i.e. ischaemic vs. non-ischaemic) or ejection fraction (i.e. reduced vs. preserved).7 For example, a recently published analysis of the RELAX-AHF trial found that high sensitivity cardiac troponins were abnormal (i.e. defined as greater than the 99th percentile of the upper reference limit) in upwards of 90% of patients admitted for a primary diagnosis of HF without clinical suspicion for acute coronary syndrome.8 In addition, detectable cardiac troponin levels have been linked to higher in-hospital and post-discharge mortality.8-11 However, it should be noted that troponin release and end-organ damage is known to occur before, during, and after hospitalization for HF.12, 13 Furthermore, the TRUE-AHF trial demonstrates that early intervention with a short-term infusion of a novel therapy capable of relieving haemodynamic congestion, as assessed by NT-proBNP levels, beyond standard treatment was unable to prevent cardiac injury and mediate tangible benefits on clinical status or long-term outcomes. Thus, in the case of ularitide, there is a clear dissociation between the resolution of cardiac distension and intravascular congestion and cardiac injury and end-organ damage in the inpatient setting. The TRUE-AHF trial also provides a unique perspective on the clinical relevance of in-hospital WHF events to the ‘patient journey’. Estimates of the prevalence of in-hospital WHF, generally defined as deteriorating signs and symptoms of HF requiring an intensification of therapy, have ranged from 5% to more than 40% based on retrospective analyses of registries and clinical trial databases.6, 14-19 In addition, in-hospital WHF has been shown to be associated with longer length of stay (LOS), greater healthcare expenditures, and higher post-discharge mortality and readmission rates.6, 14-19 However, when defined a priori and prospectively adjudicated, the prevalence of in-hospital WHF may be less than 10%.3, 4 In addition, more than 80% of cases of in-hospital WHF may require only low-intensity (i.e. defined as i.v. diuretics and/or low-dose dopamine) or moderate-intensity interventions (i.e. i.v. vasodilators and/or non-invasive ventilatory support), while only the minority of in-hospital WHF events require high-intensity interventions (i.e. i.v. inotropes/vasopressors, ultrafiltration, and/or mechanical ventilatory support). Furthermore, despite differences in the incidence of early in-hospital WHF in TRUE-AHF, these differences did not persist beyond the 48-hour infusion of ularitide and treatment with ularitide was not associated with longer LOS, higher 30-day readmissions, or increased cardiovascular mortality. Thus, given the overall low prevalence of in-hospital WHF and the absence of a clear association between treatment with ularitide and post-discharge cardiovascular morbidity and mortality, in-hospital WHF should be regarded as a supplementary endpoint and not as an independently validated outcome. Historically, there has been a conceptual divide between the diagnosis and management of acute vs. chronic HF both in the context of routine practice as well as in the conduct of clinical trials.20, 21 This traditional paradigm has been embraced by the major guideline-issuing professional societies, although this dichotomy is not supported by the available data nor has it led to major advancements in the field.22, 23 In fact, with the notable exception of a few specific aetiologies (i.e. hypertensive emergency, valvular pathology, myocardial infarction, etc.), hospitalization for HF is rarely precipitated by an acute event and cardiac filling pressures begin to rise in the days and weeks preceding hospital admission.24, 25 In addition, as a result of changes in public policy and concerted efforts to reduce early readmissions, there has been a proliferation of same-day access clinics and treatment infusion centers where patients can receive i.v. diuretics and/or other vasoactive medications in an outpatient setting.26, 27 Thus, the focus should be shifted from treating subacute chronic decompensation of HF with a short-term infusion of a novel vasoactive agent to optimizing care longitudinally and implementing established life-prolonging therapies across the continuum of HF. There are several examples of this conceptual framework in prior and more recent clinical development programmes that highlight this general principle. For instance, following the landmark β-blocker trials which clearly demonstrated a reduction in all-cause and cardiovascular-related morbidity and mortality in ambulatory HF patients with a reduced ejection fraction, there remained residual concern that β-blockers may not be safe or well-tolerated in patients currently hospitalized or recently experiencing a decompensation of chronic HF.28, 29 In contrast, the IMPACT-HF trial found that pre-discharge initiation of carvedilol, compared to usual care, in stabilized patients hospitalized for HF improved the use of β-blockers at 60 days without increasing side effects or LOS.30 This finding was further substantiated by data from the OPTIMIZE-HF registry, which found that β-blocker use at the time of hospital discharge was well tolerated, improved adherence, and was associated with an early survival benefit.31 In contrast, in a propensity-adjusted analysis, withdrawal of β-blocker therapy was associated with greater risk of mortality compared with continuation of β-blockers.32 Similarly, although sacubitril/valsartan, an angiotensin receptor-neprilysin inhibitor, was found to further reduce all-cause and cardiovascular-specific morbidity and mortality compared with enalapril in the landmark PARADIGM-HF trial,33 approximately 40% of patients had no prior history of hospitalization for HF and less than 1% of patients had New York Heart Association functional class IV symptoms at baseline. As a result, the phase IV PIONEER-HF trial (ClinicalTrials.gov Identifier: NCT02554890) is currently underway and aims to evaluate the efficacy, safety, and tolerability of in-hospital initiation of sacubitril/valsartan vs. enalapril in stabilized patients admitted for HF with a reduced ejection fraction. Future clinical development programmes should adopt similar models by broadly demonstrating safety and efficacy in patients with both stable and worsening chronic HF. In conclusion, the TRUE-AHF trial demonstrates that early intervention with a short-term infusion of a novel vasoactive agent capable of favorably modifying central haemodynamics did not prevent and/or reverse cardiac injury and other end-organ damage and translate into a benefit on clinical status and long-term outcomes. In addition, in-hospital WHF may have utility as a surrogate endpoint but should be cautiously interpreted as hypothesis-generating and not sufficient in isolation for establishing efficacy and/or regulatory approval. The future success of clinical trials in HF is dependent on reframing the traditional paradigm of acute vs. chronic HF and evaluating the safety, tolerability, and efficacy of existing guideline-directed medical therapies across the continuum of HF. Extending the evidence basis from stable to high-risk patients with worsening chronic HF may provide additional opportunities to improve short-term adherence and facilitate long-term persistence to life-prolonging therapies. Conflict of interest: A.P.A. has no conflicts of interest to declare. J.B. report research support from the National Institutes of Health, European Union, and Patient Centered Outcomes Research Institute, and consultant to Amgen, Astra-Zeneca, Bayer, Boehringer Ingelheim, BMS, CVRx, Janssen, Medtronic, Novartis, Relypsa, and ZS Pharma. M.G. has been a consultant for Abbott Laboratories, Astellas, AstraZeneca, Bayer HealthCare AG, CorThera, Cytokinetics, DebioPharm S.A., Errekappa Terapeutici, GlaxoSmithKline, Ikaria, Johnson & Johnson, Medtronic, Merck, Novartis Pharma AG, Otsuka Pharmaceuticals, Palatin Technologies, Pericor Therapeutics, Protein Design Laboratories, Sanofi-Aventis, Sigma Tau, Solvay Pharmaceuticals, Takeda Pharmaceutical and Trevena Therapeutics.