The therapeutic and prognostic paradigm of heart failure with reduced ejection fraction (HFrEF) has changed since the introduction of sacubitril/valsartan (SV). This dual angiotensin receptor-neprilysin inhibitor (ARNI) has shown efficacy and tolerability in HFrEF outpatients1 and in those with acute decompensations and haemodynamic stability.2,3 However, the recent use of inotropes was an exclusion criterion in PIONEER-HF and TRANSITION trials2,3 and to date, no randomised trial has assessed the efficacy and safety of the use of SV combined with inotropes. The only published data in this setting comes from a case report and a small retrospective series of six patients,4,5 in which this combination was started in the context of continuous invasive monitoring during admission in an intensive care unit. Inotropic support might enhance end-organ perfusion and optimise haemodynamic conditions for SV initiation and up-titration, minimizing the risk of developing symptomatic hypotension and renal dysfunction.

We retrospectively analysed our experience with the use of inotropic support to facilitate the initiation and/or up-titration of SV in patients admitted with decompensated HFrEF in a conventional cardiology ward. From September 2018 to February 2020, 105 patients admitted to our ward for acute HFrEF were treated with inotropes. In 12 of them, SV was initiated or up-titrated under inotropic support. In all cases, inotropes were started due to the presence of signs of a low cardiac output status or refractory congestion not responding to intensive diuretic therapy, or with the purpose to provide haemodynamic support to initiate SV (defined by a systolic blood pressure lower than 95mmHg). Similarly, a contemporary cohort of 12 patients admitted for acute HFrEF and treated with inotropes but in which SV was not initiated nor titrated, and matched to cases by age, sex and aetiology of left ventricular dysfunction and with similar comorbidity, was used as a control group (Fig. 1A). Heart failure (HF) and HFrEF were defined according to the current ESC Guidelines.6 Baseline and follow-up variables were recorded in a database.

Fig. 1.

(A) Individual and median (interquartile range) values of systolic blood pressure (SBP), mean arterial pressure (MAP), heart rate, N-terminal pro-B-type natriuretic peptide (NT-proBNP) plasmatic levels, serum creatinine and potassium at initiation of inotropes and of sacubitril/valsartan (SV) (or up-titration) and at discharge. (B) Flowchart of the study. ICCU: intensive coronary care unit; ns, not significant.

(0.35MB).

In the ARNI group, the inotropic agent, the time between inotrope therapy instauration and SV initiation (or up-titration) and the initial and tolerated dose at discharge and at follow-up were characterised. Blood pressure, heart rate, serum creatinine, potassium and N-terminal pro-B-type natriuretic peptide (NT-proBNP) plasmatic levels at initiation of both inotropes and SV and at discharge were reported. Incidence of adverse events during hospitalisation in both groups were registered, including acute kidney injury, symptomatic hypotension, hyperkalaemia, angioedema or death, as defined in the PIONEER-HF trial.3 In the long-term, HF readmissions and the need of mechanical circulatory support or eventual heart transplant were also analysed. This study was approved by the local Research Ethics Committee.

Categorical data were presented as frequencies (percentages) and continuous variables were expressed as mean (±standard deviation), or median (interquartile range), as appropriate. Comparisons were performed for continuous variables using Student's t or Mann–Whitney U tests, or χ2 or Fisher's exact tests in the case of categorical variables, as appropriate. Wilcoxon's paired test was used to compare median values before and after treatment with SV. Statistical analyses were performed using SPSS, version 23.

Table 1 summarises baseline clinical characteristics and events during hospitalisation and follow-up. Mean age was 62.9±10.4 years and 10 patients (83.3%) were male in each group. In the SV group, four patients (33.3%) were admitted due to a first HF episode and all of the remaining patients with previous HFrEF diagnosis were receiving optimal medical therapy, including renin–angiotensin system antagonists (100%), beta-blockers (100%) and mineralocorticoid receptor antagonists (100%). Three of the former (37.5%) were already receiving SV and 5 (62.5%) were carriers of an implanted cardioverter defibrillator (ICD), all of them for primary prevention.

In the SV group, levosimendan (single perfusion of 12.5mg, without bolus) was used in 9 patients (75%), with a median duration of infusion of 3.0 (2–4) days. There was a concomitant use with dobutamine in 3 patients (25%). The median dose of dobutamine at initiation or up-titration of SV was 7.6 (5–8)μg/kg/min. The median time lapse between inotrope therapy initiation and SV initiation or up-titration was 10 (8–12) days.

In 9 patients (75%), the initial SV dose was 24/26mg. At discharge, SV was withdrawn in 1 patient (8.3%), and 5 (41.7%) and 4 (33.3%) received the maximum and the intermediate dose, respectively. During follow-up, a higher proportion of patients tolerated 97/103mg (58.3%) and three (33.3%) remained with the intermediate dose. In the 3 patients that had previously received SV, treatment was not interrupted while on inotropes, and up-titration was possible (two were discharged on 49/51mg and one on 97/103mg).

Worsening renal function occurred in 3 patients (25%) in the SV group and in 4 patients (33.3%) in the control group (P=.65). In the ARNI group, acute kidney injury was transient and resolved in 24h in 2 cases; in the third patient SV was withdrawn. Between SV initiation and hospital discharge (Fig. 1B), a reduction in NT-proBNP levels was observed (3016.5 [949.5–8743.8] vs 1212.5, [485.8–2583.5]pg/mL; P=.005), without significant variations in systolic (112.5 [102.3–118.3] vs 105 [101.5–112.3]mmHg; P=.20) and mean arterial pressure (81.7 [75.4–84.7] vs 76.8 [74.7–81]mmHg; P=.18), serum creatinine (1.1 [0.8–1.3] vs 1.03 [0.9–1.4]mg/dL; P=.94) or potassium (4.2 [3.9–4.4] vs 4.2 [4.1–4.4]mmol/L; P=.84). During inotropic support, arrhythmias were recorded in 5 patients (41.7%) in the ARNI group (1 de novo atrial fibrillation, 3 cases of non-sustained ventricular tachycardia, and 1 case of sustained ventricular tachycardia in the context of severe hypokalaemia, without recurrence after potassium correction) whereas in the control group, non-sustained ventricular tachycardias were recorded in 4 patients (25%), with no cases of sustained or de novo atrial fibrillation. There were no episodes of symptomatic hypotension, hyperkalaemia, angioedema, or death during hospitalisation in both groups.

During a median follow-up of 10.9 (9.7–13.7) months in the ARNI group, 2 patients (16.7%) were readmitted for HFrEF (mean time to readmission of 76.5±2.1 days). Additionally, 1 patient (8.3%) received intermittent intravenous outpatient administration of levosimendan and another patient (8.3%) died presumably due to sudden cardiac death (elderly patient with end-stage HFrEF who did not tolerate SV at discharge and refused an ICD). Due to partial recovery of EF (≥35%) after SV initiation, ICD was avoided in two cases. In the control group (median follow-up of 11.5 [5.9–11.1] months), 3 patients (25%) were readmitted for HF (mean time to readmission of 35.5±3.6 days), 2 patients (16.7%) were included in heart transplant waiting list, 1 was enrolled in intermittent intravenous ambulatory administration of levosimendan and 1 died due to end-stage heart disease.

Our data suggest that SV combined with inotropes might be safe in patients admitted for acute HFrEF. This combination may allow SV up-titration and might be routinely attempted when inotropes are needed. Ours is a single centre series and might not be representative of clinical practice in less complex scenarios. Also, although the ARNI group showed a trend towards a reduction in clinical events in the follow-up, the cohort size lacks an adequate power to detect statistical significance. Additionally, as right heart catheterisation was not performed systematically, invasive haemodynamic response to SV could not be evaluated. Finally, this investigational strategy was limited to a highly selected patient on inotrope treatment and could not be extrapolated to most of them.

We report, for the first time, the use of SV and inotropes in patients hospitalised for acute HFrEF in a cardiology ward. The rate of adverse events was low and similar to the reported in the pivotal trials. Similarly, SV withdrawal was low, and after hospital discharge, a high proportion of patients reached an optimal dose, suggesting that early introduction of SV under inotropes may increase the odds to reach target doses after discharge. Moreover, the reduction in NT-proBNP levels we found seems to correlate well with left ventricle reverse remodelling and improvement in EF. However, our findings should be confirmed in randomised prospective studies.