The Year in Cardiology

Heart Failure: the Year in Cardiology 2019

John G.F. Cleland; Alexander R. Lyon; Theresa McDonagh; John J.V. McMurray


Eur Heart J. 2020;41(12):1232-1248. 

In This Article

Angiotensin Receptor-neprilysin Inhibitors

Heart Failure With Reduced Ejection Fraction

As experience in the implementation of angiotensin receptor-neprilysin inhibitors (ARNIs) grows, both in clinical trials and in clinical practice, there is a strong argument to consider them as first-line agents, rather than angiotensin converting-enzyme inhibitors (ACEi) or angiotensin receptor blockers (ARB), for the treatment of HFrEF. In PIONEER-HF,[81] 881 patients with an LVEF ≤40% who were hospitalized for worsening heart failure were randomly assigned, without a run-in period, to sacubitril/valsartan or enalapril prior to discharge and followed for 8 weeks to determine the effect on plasma concentrations of NT-proBNP; about one-third had new-onset heart failure. Sacubitril-valsartan exerted a greater reduction in NT-proBNP. Reductions in markers of myocardial injury or stress, high-sensitivity cardiac troponin-T and soluble ST2, were also observed. These effects appeared early after randomization (within 1–4 weeks). Moreover, patients assigned to sacubitril/valsartan were less likely to experience adverse outcomes within the first 8 weeks. TRANSITION[82] randomly assigned 1002 patients to pre- or post-discharge initiation of sacubitril/valsartan, showing no adverse consequences to earlier administration.

EVALUATE[83] compared the effects of sacubitril/valsartan and enalapril on aortic stiffness in HFrEF most of whom were already chronically treated with an ACEi or ARB. After 24 weeks treatment, no differences in aortic stiffness were observed but slightly greater reductions in LV end-diastolic and systolic volumes were observed with sacubitril/valsartan compared with enalapril, although changes in LVEF were similar. Mitral E-velocity and left atrial volume declined, consistent with a fall in left atrial pressure. PROVE-HF,[84] an observational study, had similar findings and showed that most of the decline in NT-proBNP occurred within 14 days consistent with the rapid onset of clinical benefit observed with sacubitril/valsartan in trials and clinical practice. PRIME[85] was an RCT (n = 118) comparing the effects of sacubitril/valsartan or valsartan on functional mitral regurgitation in patients with an LVEF between 25% and 49% who were already receiving an ACEi or ARB. Those assigned to sacubitril/valsartan had greater reductions in mitral regurgitation and LV end-diastolic and left atrial volumes but LVEF increased by a similar small amount in each group (about 2.5%).

Further reports from PARADIGM-HF suggest that, compared with enalapril, sacubitril/valsartan may improve markers of collagen metabolism, in particular, decreasing synthesis of type-I collagen, which makes an important contribution to myocardial stiffness.[86] In I-PRESERVE, irbesartan (an ARB) did not affect collagen biomarkers compared with placebo.[87]

Heart Failure With Preserved Ejection Fraction

PARAGON-HF investigated the effect of sacubitril/valsartan compared to valsartan alone on morbidity and mortality in patients with HFpEF (defined as an LVEF >45%).[88] It was the first RCT since PEP-CHF[89] to require patients to be treated with diuretics, the first-line treatment for the relief of symptoms and signs of congestion, and to have echocardiographic evidence of cardiac dysfunction. It was also the first large trial of HFpEF to require all patients to have raised plasma concentrations of natriuretic peptides, the most powerful, widely available prognostic marker in HFpEF. Sacubitril/valsartan was compared with valsartan rather than placebo because many patients eligible for PARAGON-HF had indications for ACE inhibitors and ARBs such as hypertension and CAD. The only trial comparing valsartan to placebo in HFpEF was of modest size and neutral.[90] Previous RCTs of other ARBs, including candesartan (CHARM-Preserved) and irbesartan (I-PRESERVE) failed to show substantial benefit for HFpEF.[88] Patients had to tolerate, sequentially, both valsartan and sacubitril/valsartan at half the intended target dose before randomization. This simulates clinical practice (doctors do not usually prescribe medicines to patients unwilling or unable to take them) and reduces the risk of a neutral trial-outcome due to low adherence. Of 10 539 patients screened, 4822 were randomized.

PARAGON-HF was neutral for its primary endpoint (CV death or the total number of recurrent hospitalizations for heart failure;[91] Figure 4). Some have argued that the P-value was very close to 0.05 and that it was 'almost' positive. This misses the point. The trial shows that the size of the potential benefit of sacubitril/valsartan for HFpEF is modest, regardless of the P-value and that the treatment is, overall, unlikely to be cost-effective. Accordingly, we should look for more effective treatments or, more controversially, subgroups that obtain greater benefit. After a median follow-up of 35 months, 23% of patients experienced a primary event but the annual incidence of CV and all-cause mortality were, respectively, only about 3% and 5%, which is similar to those for previous trials of HFpEF and for elderly patients with resistant hypertension assigned to placebo in HYVET.[92] Although <3% of patients were reported to have heart failure in HYVET, a combination of indapamide and perindopril reduced all-cause mortality and cut the incidence of heart failure by >50%. Many of these patients probably had undiagnosed HFpEF prior to randomization. Higher rates of hospitalization for heart failure in trials of HFpEF compared to hypertension may well reflect ascertainment bias, as clinicians who are interested or expert in the management of heart failure are more likely to diagnose or report heart failure events. Overall, these trials suggest that the mortality rate and possibly the rates of cardiovascular and all-cause hospitalization may be similar in patients with and without a diagnosis of HFpEF, if they have a similar burden of co-morbidities. However, it is also likely that many patients with hypertension, CAD and T2DM have undiagnosed heart failure.

Figure 4.

Effect of sacubitril/valsartan compared with valsartan for heart failure with preserved ejection fraction in PARAGON-HF. Reproduced with permission from ref.91

Subgroup analysis suggested that the effect of sacubitril/valsartan on the primary endpoint was greater for patients with an LVEF below the median (57%), but this was driven almost entirely by an effect on hospitalization for heart failure rather than on CV death.[93] The effect of sacubitril/valsartan on the primary endpoint was also greater for women and this was true throughout the studied range of LVEF, but again this was driven by a difference in hospitalization for heart failure and not CV mortality.[94] Reductions in NT-proBNP were similar for each sex. Sacubitril/valsartan appeared to have a favourable effect on quality of life for men but not for women. Patients with a recent heart failure hospitalization may also have benefited more.[95] These observations should be interpreted in the light of a trial that was neutral for its primary endpoint. No effect was observed on mortality and the benefits of treatment on quality of life and hospitalizations for heart failure according to sex were inconsistent. In PARADIGM-HF, no difference in treatment effect according to sex was observed. A further sizeable RCT in HFpEF, PARALLAX-HF, investigating the effects of sacubitril/valsartan on quality of life and exercise capacity will provide more evidence in 2020 (