Heart failure with reduced ejection fraction (HFrEF) is related to bad prognosis at midterm1: 3 drug classes – betablockers, angiotensin converting enzyme inhibitors (ACE-i), or angiotensin receptor blockers (ARB) and mineralcorticoid receptor antagonists (MRA) – and recently, angiotensin receptor neprilysin inhibitors (ARNI), have demonstrated to improve survival and quality of life in this condition.2–6

Current European guidelines recommend starting betablockers and ACE-i/ARB together in debut HFrEF, although there are no prospective trials evaluating this strategy versus the sequential initiation of drugs. Adding MRA is recommended if after dose optimization the patients is symptomatic and left ventricular ejection fraction (LVEF) is still ≤35%,2 which in the real-world setting takes several months. European guidelines also recommend starting the neurohormonal blockade during hospitalization in the stability phase after an acute heart failure episode. This recommendation is based in a relatively weak evidence.7,8 More recently, the TRANSITION and PIOONER-HF trials have demonstrated that starting ARNI in this scenario is safe and improves heart failure parameters, even in debut heart failure patients.9,10

A recent update of the Canadian Society of Cardiology heart failure guidelines, support the initial triple neurohormonal blockade (TNHB) with betablockers, ACE-i/ARB and MRA in debut HFrEF11 in opposition to the sequential strategy of the European and American heart failure guidelines. This recommendation is not based in any study, but, despite controversial, might represent a not uncommon real life approach. Initial TNHB could be seen as an attractive idea as theoretically allows more renin-angiotensin-aldosterone system blockade from the diagnosis, which might allow an earlier and stronger modification of the course of the disease, with the side effects of MRA as the “yang” of this strategy.

In this paper we perform an exploratory analysis of the frequency and the prognostic impact of TNHB in a cohort of patients discharged from our centre with new onset HFrEF. This is, to the best of our knowledge, the first time this kind of analysis is reported in the literature.

From 665 patients with debut HF, 317 had HFrEF. A total of 37 were excluded of the analysis for not meeting the additional inclusion criteria. Finally, 280 discharged patients fulfilled the study criteria. Of them, 162 (57.9%) were discharged on TNHB. The remaining 118 patients (42.1%, the non-TNHB group) were discharged on any combination of betablockers and ACE-i/ARB, betablockers and MRA or ACE-I/ARB and MRA (83 patients, 29.6% of the whole cohort and 70% of the non-TNHB group) or on just one heart failure drug (35 patients, 12.5% of the whole cohort and 30% of the non-TNHB group) (Fig. 1). In general, 85.4% of the patients were discharged on betablockers, 85.7% on ACE-i or ARB, 70% on MRA, 77.1% on furosemide, 53.9% on statins and 8.6% on ivabradine.

Baseline characteristics of the whole cohort and stratified depending on TNHB are shown in Table 1; 36.8% of patients had ischaemic heart disease as the cause of left ventricular dysfunction; 63.2% of patients had a dilated cardiomyopathy of different aetiologies. Mean age was 65.4±12.5 years, mean LVEF was 29.8±7.3% and tricuspid annular plane systolic excursion was 16.8±4.4mm. Discharge creatinine was 1.14±0.3mg/dL, discharge serum potassium was 4.2±0.6mEq/L and median N-terminal prohormone of brain natriuretic peptide (NT-ProBNP) at admission 2917pg/mL. After a mean follow-up of 43.4±26.6 months, 72 patients died (25.7% of the cohort), with a mean survival of 76.1 months (Table 2).

TNHB group had some significant baseline differences compared to non-TNHB group. TNHB group had less ischaemic heart disease (31.5% vs. 44.1%; P=.03), was younger (62.9±12.7 vs. 69.0±11.4; P=.0001), more frequently smoker (27.2% vs. 14.4%; P=.03), with poorer LVEF (28.7±7.4% vs. 31.4±6.7%; P=.002) but better creatinine (1.08±0.3 vs. 1.21±0.4md/dL; P=.02) and lower estimated pulmonary pressures (43.26±9.38 vs. 48.85±14.61; P=.02). Mean survival on TNHB was better (months, 82.1±3.0 vs. 68.6±3.8; P=.007).

Age > 70 years, diabetes mellitus, previous stroke, weight, heart failure aetiology, TNHB and discharge creatinine were entered in a Cox regression survival model. The discriminative accuracy of the model evaluated by the area under the ROC curve was 0.75 (95% confidence interval [95%CI], 0.68–0.82). In this multivariate analysis, TNHB was still an independent predictor of better survival (HR, 1.79; 95%CI, 1.05–3.05; P=.03) (Fig. 2), meanwhile age > 70 years (HR, 0.39; 95%CI, 0.21–0.72; P=.003), previous stroke (HR 0.39; 95%CI, 0.18–0.84; P=.01) and weight (HR, 0.98; 95%CI, 0.96–0.99; P=.04) were independent predictors of worse survival.

Fig. 2.

Cox Regression Analysis. Kaplan–Meier survival curves regarding triple neurohormonal blockade.

(0.06MB).

To investigate factors independently related to TNHB prescription, previous myocardial infarction, heart failure aetiology, discharge creatinine, age, potassium, LVEF, right ventricular end-systolic pressure and hypertension were entered in a logistic regression model. Increasing age was the only factor independently related to less TNHB prescription (HR, 0.92; 95%CI, 0.88–0.97; P=.002).

In this retrospective analysis of a single centre cohort of patients with new onset HFrEF requiring hospital admission and that were successfully discharged, we show that treating patients from the diagnosis with THNB is related to better survival, even after adjusting for possible confounders. This is, to the best of our knowledge, the first time that the outcomes and frequency of TNHB as the initial strategy for debut HFrEF has been studied in the literature.

It is also of note that more than 50% of the patients of our cohort were discharged on TNHB, which probably resembles clinical practice in some European centres despite anticipating the current guideline algorithm, although there is no information in the literature. In our cohort, TNHB was less used mainly in elderly patients despite not having contraindication for MRA prescription, which it is in consonance with previously published works.13 TNHB in our cohort was mainly reserved to younger patients but with lower LVEF, thus clinician perception of disease severity was probably higher because of the LVEF, although, in fact, those not treated with TNHB had a more unfavourable clinical profile (worse renal function, higher pulmonary pressures and older age).

Anti-remodelling and anti-fibrotic effects of MRA have been related to improved clinical endpoints in the HFrEF pivotal trials.5,8,14 It is probable that these effects are taken into consideration when thinking about initial TNHB, added to betablockers and ACE-i/ARB, especially taking into account that in some MRA trials8 survival curves separate from the study beginning. Nevertheless, this practise has not been included in the latest ESC guidelines. Furthermore, it is also probable that MRA are added to patients with persistent congestion to increase the effect of loop diuretics, especially if coexists with hypokalaemia, instead of other diuretics and potassium supplements.

Current guidelines recommend initial dual neurohormonal blockade with betablockers and ACE-i/ARB although this strategy has not been proven in randomized prospective trials; thus prospective evidence of dual and triple neurohormonal blockade as the initial approach of patients with debut HFrEF could be considered at the same level. To this respect, it is considered that the potential benefit of initial dual neurohormonal blockade is related to the additional benefit of one drug over the other, after the results of the pivotal clinical trials, a rationale that might be correct in the same way adding MRA to this equation. The guidelines greatly encourage drug titration and algorithm-based new drug initiation during the follow-up in the heart failure or general cardiology clinics, but despite of this hospital discharge treatment is mainly unchanged in a great proportion of patients,15,16 which could also be involved in the benefit observed in our study with initial TNHB. It has been shown that a significant proportion of patients candidates to MRA do not receive the treatment, maybe because of therapeutic inertia.13 This is probably even truer in functional class II patients, as they are usually seen as “clinically stable”, but paradoxically get more benefit of drug optimization as it is probably easier to modify the HF progression from an earlier stage of the disease where ventricular remodelling is more probably reversible.6

MRA in HFrEF have demonstrated to improve the prognosis of HFrEF based on the results of 3 big trials: RALES, EPHESUS and EMPHASIS.5,8,14 The results of these trials have been translated into current 2016 ESC recommendations. Although also involving MRA, the rationale, design and objectives of other trials like ATHENA-HF17 are not really comparable with the results of the previous trials. The EPHESUS trial could also be considered considerably different to RALES and EMPHASIS as it was conducted only in patients with a recent acute coronary syndrome and only 15% had HF on the background, but in fact tested initial TNHB versus the standard treatment. In the EPHESUS trial more than 85% of patients were on ACE-i/ARB and 75% were on betablockers. Moreover, in the EPHESUS trial the benefit with eplerenone was only found in the group of patients already on betablockers and ACE-i/ARB. These pivotal MRA trials demonstrated a significant reduction of mortality on top of optimized betablockers and ACE-i/ARB in RALES5 and EMPHASIS.14 Whether this reflects the way clinical trials were conducted, waiting months on previous therapy to add a new drug that decreases mortality even more, with the evidence from trials in which survival curves separate from each other from the initial weeks on treatment and, taking into account that this rationale is not considered necessary when recommending adding betablockers to an ACE-i/ARB, might be the reasons why the Canadian update have recommended initial TNHB. Furthermore, recent results of the PIONEER-HF trial10 with sacubitril-valsartan come to challenge even more current recommendations, as it demonstrates, with more than 50% of naïve patients after an acute heart failure episode, further reductions of NT-ProBNP and HF hospitalizations compared to the standard therapy of ACE-i/ARB, which was also true in those with new onset HFrEF.

In debut HFrEF requiring hospital admission, being discharged on TNHB might be related to a better prognosis. This finding should be tested in prospective randomized placebo-controlled trials.

None declared.