Introduction

Heart failure (HF) is the final stage of many cardiovascular diseases. Despite optimal medical therapy (OMT), HF with reduced ejection fraction (HFrEF) remains a central issue of clinical management, with high social and economic impact, worldwide [1,2,3]. Among the latest technological and pharmacological innovations, the sacubitril/valsartan drug combination, also known as angiotensin receptor neprilysin inhibitor (ARNI), has been demonstrated to get relevant benefits in this clinical setting [4,5,6,7,8]. Hence, neurohormonal modulation by ARNI has gradually become a landmark of HF treatment, but questions on its dose-dependent effectiveness are open yet [7,8,9].

We sought to evaluate clinical and functional effects in low- vs high-dose medication use in real-life HFrEF patients.

Methods

This was a prospective observational study conducted at Messina and Palermo University Hospitals (Italy). Among all outpatients admitted from October 2017 to June 2019 due to HFrEF from any aetiology, we enrolled those fulfilling the same inclusion criteria as in the PARADIGM-HF study [5]. Patients should have been New York Heart Association (NYHA) functional class II or III, clinically stable since their last hospitalization and with left ventricular ejection fraction (LVEF) ≤ 0.35. Patients who had received an implanted cardioverter defibrillator (ICD) at least 3 months prior to admission to study, whether in combination with cardiac resynchronization therapy (CRT-D), were also included.

Conversely, refractory HF, impaired glomerular filtration rate (< 60 mL/min), severe hepatic dysfunction, sustained hypotension, permanent atrial fibrillation, aortic or mitral stenosis, history of angioedema, hospitalization within the last 3 months due to destabilization, acute coronary syndrome and recent surgery were exclusion criteria.

According to the Italian Heath Care Regulatory System, the ARNI administration was authorized after 6-month treatment with ACE inhibitor or angiotensin-receptor blocker (ARB). Sacubitril-valsartan was given on top of optimal medical therapy (OMT) at the starting dose of 50 (24/26) mg twice a day (bis in die, BID), in an open-label fashion. According to 2017 guidelines [10], uptitration to 100 (49/51) mg and 200 (97/103) mg BID was established within the first 2-month period, upon clinical condition, drug tolerance, home-reported blood pressure (BP), renal function and kaliemia. Patient had to accomplish at least 10-month follow-up treatment. The observation period ended July 2020. The median medication dose was calculated on the annual dose taken by each patient and served for clustering the low-mid- (group A) and mid-high-dose (group B) recipients.

Physical examination, quality of life (QoL) assessment by Kansas City Cardiomyopathy Questionnaire (KCCQ) [11], 12-lead ECG, transthoracic echocardiography, 6-min walk test (6MWT), ICD check and laboratory samples were scheduled every 3 months. Changes in NT-proBNP levels and estimated glomerular filtration rate (eGFR) from serum creatinine (Cockroft-Gault equation) were monitored all through the study length. The protocol was conducted according to the Helsinky declarations and international guidelines, so it was just approved by the local advisory boards, with no trial number registration required.

All patients gave informed consent.

Six-minute walk test

An indoor 6MWT was performed along a flat straight corridor with a hard surface. The walking course was 30 m in length, with turnaround points safely marked. Patients were asked to walk as fast as possible with no restriction in resting for a while, if necessary. Heart rate (HR) and symptoms were monitored during the walk. Absolute and predicted walk distance (WD), and walk speed, were calculated in everyone. Predicted WD was calculated according to individual characteristics (weight, height and age), as follows [12, 13]:

  • Men: [(7.57 × height in cm) − (5.02 × age) − (1.76 × weight in kg) − 309 m]

  • Women: [(2.11 × height in cm) − (2.29 × weight in kg) − (5.78 × age) + 667 m]

Echocardiography

Colour Doppler ultrasound examination with standard measurements was performed in each participant. Examiners were blinded on previous achievements from the patient file. Studies were stored in Digital Imaging and Communications in Medicine (DICOM) format and then analysed offline. Left ventricular end-diastolic/end-systolic volumes and LVEF were calculated using the modified biplane (4- and 2-chamber apical views) Simpson rule method [14].

In 77/90 patients (85.5%), LV diastolic function was assessed by mitral valve early Doppler velocity (E)/early tissue velocity (E′) ratio (normal value < 12) and global longitudinal strain (GLS) deformation (normal value: lower than −18%) was measured from the 4-, 3- and 2-chamber apical views.

Statistical methods

Continuous variables are expressed as mean ± SD and categorical variables as numbers and percent (%). Paired data analysis of clinical (NYHA and KCCQ score) and functional (6MWT distance, BNP, NT-proBNP) markers was performed at 3-, 6- and 12-month time follow-ups vs the previous timeline step by using the Pearson model with an alpha level of 0.01 for statistical consistency. Moreover, analysis of variance by ANOVA with Levene’s and Tukey’s correction was performed to establish the between-group homogeneity. Interobserver variability for echocardiographic measurements was calculated in each laboratory as by 0.052 (95% CI 0.034–0.063). The null hypothesis was rejected at two-tailed p < 0.01.

Results

A total of 90 HFrEF patients, mean age 64.5 ± 10.9 years, 74 males (82%), predominantly of an ischaemic aetiology (72%), were enrolled from an initial study population of 96 patients. Of these, 6 were excluded due to ARNI intolerance in 3, allergic reaction in one and non-cardiac hospitalization (COVID-19) in another 2 cases. Demographic and clinical characteristics are summarized in Table 1. Forty-six patients (51%) had received an ICD, and 17 (19%) CRT-D.

Table 1 Demographic and clinical characteristics of the study population
Full size table

The study duration was 12 months (range 11–13) in length, and nobody was lost to follow-up. According to the median cumulative dose of the ARNI (75 mg BID), 52 patients (58%) entered group A (≤ 75 mg) and 38 (42%) group B (> 75 mg).

Most patients gradually required a less amount of loop diuretics on follow-up (online Table 2). Systolic BP was mildly higher in patients from group B, on both admission and follow-up, and this allowed keeping a greater dose of medication.

A clinical improvement was observed in most patients from both groups, according to their QoL. No relevant side effects or adverse events were observed. Renal function (eGFR) showed just a trivial decline in both medication groups.

Regarding exercise tolerance, mildly greater WD and walk speed were observed in group B, especially after the first 6-month treatment (Fig. 1). This difference, however, become trivial as normalized for sex and weight. The proportion of patients walking > 350 m increased from 21 to 58% in group A (p < 0.001), and from 29 to 82% in group B (p < 0.001).

Fig. 1
figure 1

Functional achievements at 6-min walk test in both study groups. NT-proBNP serum levels are also displayed on the bottom right panel. In-group differences and detailed measurements are reported in Table 3. NT-proBNP, N-terminal pro B–type natriuretic peptide; *p < 0.05

Full size image

Clinical condition improved according to a lessening in NT-proBNP levels, quicker in group B. Left ventricular volumes decreased in all patients, with a mild improvement in the stroke volume and LVEF (3 points % in group A, 5 points % in group B, p = NS). A positive effect was also seen in the E/E′ ratio and GLS (online Table 3; Fig. 2).

Fig. 2
figure 2

Overtime changes in left ventricular (LV) volumes, ejection fraction (LVEF) and global longitudinal strain (GLS) at transthoracic echocardiography. LVEDVi, left ventricular end-diastolic volume index; LVESVi, left ventricular end-systolic volume index; NT-proBNP, N-terminal pro B–type natriuretic peptide. LVEF value is percent (%). *p < 0.001

Full size image

Discussion

Twelve-month ARNI therapy was clinically advantageous in our HFrEF outpatient population. We analysed the impact of different cumulative ARNI dosage on clinical status, physical performance, QoL and echocardiographic parameters. After switching to ARNI, both groups moved towards better clinical conditions. Also, reverse LV remodelling and slight systo-diastolic functional amelioration were observed according to a gradual decrease in circulating NT-proBNP levels. Overall, patients from group B attained greater and quicker clinical achievements, which were however satisfactory even in group A at the end of follow-up. Such improvements may be consistent with a positive impact on the progression and prognosis of HFrEF in the real-life patient population.

Despite the proportion of high-dose ARNI recipients in large pivotal trials, the adherence to multidrug therapy remains a challenging item in clinical practice, even if adults with cardiovascular diseases are prone to polypharmacy due to comorbidities and complexity of medication regimens [15, 16]. Though we have already observed drug-related benefit in patients at a high risk of arrhythmic disorders, high-dose ARNI therapy has some practical limitations, especially in patients with low BP symptoms [17, 18]. In fact, despite suitable clinical characteristics, Martens et al. demonstrated that high doses were administered only to 32% of HFrEF patients [19].

Unfortunately, the PARADIGM-HF study showed that patients scheduled to a dose reduction were at a higher risk of major cardiovascular events [5, 9]. Nevertheless, Vardeny et al. reported a reduced risk of death and HF hospitalization even by taking lower ARNI doses, compared to ACE inhibitors [20]. A recent meta-analysis confirmed the dose of 200 mg BID to be possible in 35% of European patients, with a potential of discontinuation in 12.8% of cases [21]. These findings likely support the fact that real-life patients are different from pivotal studies, and encourage clinicians to search for OMT without extremes.

In our study, exercise performance improved in both groups, and this is an important therapeutic target. The proportion of patients walking > 350 m at the end of study was greater in group B (82%) than in group A (58%), but we also demonstrated that confounders like weight, body mass, sex and physical inactivity can affect their performance.

The interplay between clinical and functional achievements and NT-proBNP levels in this study confirmed previously published data by Pandey et al. who found that even lower than standard doses of ARNI were able to reduce the circulating NT-proBNP and loop diuretic requirement, without any relevant change in potassium or serum creatinine [22]. Though the dose of 50 mg BID has been considered the lowest ARNI dose to be given to HFrEF patients, clinical advantages have also been shown with very low-doses (12/13 mg BID) [23].

Regarding LV function by echocardiography, drug-related reverse remodelling likely remains a controversial issue. In previous studies, ARNI was effective when initiated early after diagnosis, and at least for 3 months [24,25,26]. Though most patients in our series have gotten a decrease in cardiac chamber volumes, chiefly related to a reduced end-systolic volume, case-by-case and interobserver variability represent important shortcomings for a correct interpretation of these findings. The weak improvement in LVEF (4 points %) in our patients, a bit more relevant in group B, confirmed the results of Almufleh et al. who reported + 5% LVEF in high-dose vs + 4% in low-dose recipients [26]. Conversely, we did not see the complete functional recovery found in 17% of non-ischaemic HF patients by Chang et al. [27].

Tissue Doppler and strain imaging could be the way to overcome the limitation of LVEF variability. Improved GLS and LV diastolic function was recognized in our subset of 77 patients, though their absolute values remained quite low. Strain echocardiography likely represents an interesting imaging modality to explore myocardial impairment in HF patients as consequence of tissue oedema, inflammation and fibrosis [28], which can be counterbalanced by ARNI treatment [29, 30].

Limitations

The small sample size is the main limitation of the present study. Our patients, however, were strongly asked to follow the whole scheduled observational period, and this allowed the study completion at 1-year follow-up. Findings from the study population were not compared to controls, and this may have affected the true assessment of clinical outcomes. However, the aim of the study was not to establish the ARNI effectiveness vs conventional therapy, but dose-related discrepancies, if any. Though no serious outcomes occurred in our study population, informative findings on subclinical events, like intercurrent atrial fibrillation or ventricular arrhythmias, are missing except for the ICD-monitored patients. Echocardiography examination was only partially blinded, and the examiner could retain therapeutic information by the patient. Prospective longitudinal studies with a central echo data reading should be encouraged to avoid any interference in data analysis. Further functional information might have been achieved by cardiopulmonary exercise testing, which remains the gold standard for functional assessment of HF patients.

Though the prevalence of type 2 diabetes was not so high in our study population, metabolic parameters interact with the cardiomyocyte function and lead to increased inflammation, apoptosis, reactive oxygen species and altered calcium signalling [31]. Diabetes cardiomyopathy likely impairs prognosis and deserves more awareness in HF populations, also in view of the most recent studies on SGLT2 inhibitors in both diabetic and non-diabetic patients [32].

Current literature also suggests early initiation of the ARNI (50 mg BID) as an effective, risk-free, therapeutic approach that increases quality-adjusted life expectancy and cost savings compared with no initiation or initiation late after hospitalization [33, 34].

Conclusion

One-year sacubitril/valsartan treatment on top of OMT was advantageous in our real-like HFrEF patient population. Beneficial effects on clinical condition, QoL, exercise performance, LV remodelling and function were observed in both study groups, though slightly greater and with a shorter time to benefit in patients from group B.