The PARAGON-Heart failure trial – another disappointment for heart failure patients with hypertension and preserved ejection fraction

Novel angiotensin receptor antagonist/neprilysin inhibitors (ARNIs) seek to exploit the clinical benefits of combining renin-angiotensin-aldosterone-system (RAAS) antagonism and neutral endopeptidase inhibition with neprilysin (NEP). ARNIs augment natriuretic peptide levels, while circumventing potential safety issues related to excess bradykinin and substance P, including angioedema, which was frequently observed with vasopeptidase inhibitors [1,2]. LCZ696, comprised of equal molecular moieties of the NEP inhibitor prodrug AHU377 (sacubitril) and the angiotensin receptor blocker (ARB) valsartan combined in one compound, is the first-in-class ARNI. After oral intake, sacubitril is metabolised by enzymatic cleavage to LBQ657, an active NEP inhibitor. In a clinical trial in patients with hypertension, LCZ696 had greater blood pressure lowering efficacy and similar tolerability compared with stand-alone valsartan treatment [3]. Supported by a strong body of preclinical evidence, recent clinical studies have demonstrated more favourable effects of LCZ696 on cardiovascular remodelling compared to the ARB olmesartan in hypertensive patients [4]. Despite these encouraging data, no adequately powered clinical outcome trial of LCZ696 for the treatment of uncomplicated hypertension is currently being planned or conducted. The medical community is left with a striking gap in evidence for the putative benefit and safety of LCZ696 in patients with hypertension.

LCZ696 lowers morbidity and mortality more effectively than enalapril in patients with heart failure with reduced ejection fraction (HFrEF) by mechanisms that appear to be independent of blood pressure lowering [5]. Moreover, in a phase-II trial in patients with heart failure with preserved ejection fraction (HFpEF), LCZ696 reduced the primary endpoint of NT-pro-BNP, a surrogate of cardiac wall stress, to a greater extent than valsartan [6]. Interestingly, in that study blood pressure reduction was greater in the LCZ696 group than in the valsartan group after 12 weeks of treatment. Whether or not these effects translate into improved cardiovascular (CV) disease outcomes in HFpEF has now been tested in the Efficacy and Safety of LCZ696 Compared to Valsartan, on Morbidity and Mortality in Heart Failure Patients With Preserved Ejection Fraction (PARAGON-HF) Trial [7].

The PARAGON-HF investigators tested the superiority of sacubitril/valsartan over valsartan in 4822 patients with HFpEF, of whom 96% had history of hypertension [7]. The PARAGON-HF participants were very similar to participants in previous HFpEF trials [8–11]. Comorbidities in PARAGON-HF included, beyond hypertension, coronary disease (43%), diabetes (43%), and atrial fibrillation (32%). All participants had ejection fraction ≥45%, on average 58%, and left atrial enlargement and/or left ventricular hypertrophy at baseline. All patients had to qualify by having elevated plasma NT-pro-BNP levels. This biomarker approach to ensure true heart failure had until now only been applied in the American part of the TOPCAT Study [11].

We have previously commented on the rationale, hypothesis, study design, patient population and the potential importance for further treatment of HFpEF, including treatment of severe hypertensive heart disease, as consequences of the final PARAGON-HF study results [12,13]. The PARAGON-HF investigators predicted that sacubitril/valsartan would be superior to valsartan in the reduction of mortality and/or prevention of heart failure hospitalization in patients with HFpEF [7]. If so, this finding could provide the evidence base for use of sacubitril/valsartan for treatment of this important half of the heart failure population. Importantly, HFpEF is the ultimate consequence of hypertensive heart disease, and it is likely that a subset of patients with severe hypertension would benefit from this compound that was originally proposed [3] but, unfortunately, bypassed as a treatment for patients with hypertension [12,13].

Unless PARAGON-HF could show superiority of the ARNI sacubitril/valsartan over valsartan, the results would be inconclusive since the study, like previous HFpEF outcome studies [8–11], did not have sufficient statistical power for non-inferiority analysis [7]. However, the primary endpoint in PARAGON-HF was a composite of CV death and hospitalization for heart failure and the study was supposed to have sufficient power to analyze both as co-primary endpoints [7].

On July 29^th the sponsor of the PARAGON-HF study [7] announced to the stock marked that statistical significance was missed for the composite endpoint of CV death and total heart failure hospitalizations. On September 1^st the main scientific results of the PARAGON-HF were presented at the European Society of Cardiology congress in Paris, France, and the full data-set was published [14]. The major finding was that the study missed the primary composite endpoint (894 events in 526 patients treated with sacubitril/valsartan vs. 1009 events in 557 patients treated with valsartan alone, rate ratio 0.87; 95% confidence interval [CI], 0.75 to 1.01; p = 0.06). There was no mortality benefit of sacubitril/valsartan compared to valsartan alone. The incidence of CV death was 8.5% in the sacubitril/valsartan group and 8.9% in the valsartan group (204 vs. 212 deaths, hazard ratio 0.95; 95% CI 0.79 to 1.16), and all-cause death was similar in the two groups, 342 (14.2%) vs. 349 (14.6%).

There were 690 and 797 total (both first time and recurrent) episodes of hospitalization for heart failure, respectively (rate ratio 0.85; 95% CI 0.72 to 1.00), in the sacubitril/valsartan compared to the valsartan patients. The number of patients with first time heart failure hospitalization was not specified, and this number would be needed in order to compare PARAGON-HF with previous HFpEF studies [8–11]. “Total hospitalizations for heart failure” appears to be a newly defined endpoint, as patients were counted more than once for the primary endpoint, which occurred in 526 and 557 patients in the two arms [14]. Including both first and recurrent heart failure hospitalization in the primary endpoint has never been done in outcome studies in HFpEF [8–11] and, to our knowledge, never in other fields of CV medicine.

Thus, what did the PARAGON-HF study show? In patients with HFpEF, sacubitril/valsartan was not superior to valsartan alone in the prevention of hospitalization for heart failure or death from CV causes [14]. Because of the lack of statistical power to show non-inferiority, it cannot be claimed that sacubitril/valsartan was as effective as valsartan in the treatment of patients with HFpEF [7,14]. Further, because treatment with an ARB for HFpEF has not been shown to be significantly better than placebo [8,10], and the ARB class is not an approved treatment for HFpEF [15,16], PARAGON-HF is inconclusive [7,14]. One is left wondering whether sacubitril/valsartan is an effective treatment in patients with HFpEF, in the same way as one wonders whether the ARBs candesartan [8] and irbesartan [10], the angiotensin converting enzyme inhibitor perindopril [9] and the mineralocorticoid receptor antagonist spironolactone are effective treatments in HFpEF.

This outcome is disappointing for patients with severe hypertension and/or hypertensive heart disease, but does not preclude a benefit of sacubitril/valsartan in the general population of hypertensive patients [3,4]. To achieve further progress in searching for the place of sacubitril/valsartan in the treatment and prevention of CV disease, we suggest that the compound should now be tested in an outcome trial in the general population of hypertensive patients. An additional informative approach would be to identify and characterize the various patient groups that hide within the term “HFpEF”.

HFpEF may include patients with longitudinal strain and long-axis systolic dysfunction caused by severe left ventricular hypertrophy [17], patients with combined moderate systolic dysfunction and diastolic dysfunction, patients with diastolic dysfunction only and patients with systolic and/or diastolic dysfunction who then become decompensated due to rapid atrial fibrillation [18]. To design effective therapies for HFpEF, the underlying mechanisms that cause the left ventricle to fail in the presence of a normal ejection fraction must be better delineated and new classes/combinations of heart protecting drugs must be developed and tested in randomized controlled outcome studies. For example, we all know that if HFpEF is provoked by sudden onset rapid atrial fibrillation, beta-blocker therapy may be effective in controlling the rhythm and patients may live without symptoms of heart failure despite remodeling of the left ventricle and mild to moderate systolic or diastolic dysfunction.

Disclosure statement

S.E.K., K.N., M.B. and S.O. are editors of Blood Pressure and report no relevant conflicts of interest to disclose related to this editorial.