Hemodynamic Effects of Sacubitril/Valsartan in Patients with Reduced Left Ventricular Ejection Fraction Over 24 Months: A Retrospective Study

Ejection fraction (EF) is defined as the volume of blood ejected during left ventricular contraction and is measured using the Simpson biplane method. TAPSE indicates right ventricular longitudinal function and is measured using M-mode echocardiography between the end-diastole and peak systole, with the cursor along the tricuspid lateral annulus in the apical four-chamber view. PAPsys indicates the peak pressure of blood in the right ventricle during systole and is calculated from the tricuspid regurgitation jet maximal pressure gradient and right atrial pressure. Tricuspid regurgitation peak systolic velocity is a quantitative measurement of the peak velocity of blood flow across the tricuspid valve. Left atrial and right atrial size are qualitative descriptions of the size of the atria and are measured in a four-chamber apical view at end-systole. E wave/A wave ratio is determined as the ratio of peak early to peak late transmitral flow velocities reported without units. It is measured to evaluate the left ventricular diastolic function. The VCI measurement is assessed in the subcostal view 1–2 cm from the junction with the right atrium in the long-axis view. VCI values are used to determine systolic pulmonary artery pressure. An echocardiographic response was defined as improved LVEF with an increase of ≥ 5%, and a functional response was defined as an improvement of one class category in NYHA classification [9, 10].

Congestion was defined as the presence of one or more symptoms of fluid overload. The following symptoms are considered: pulmonary rales, third heart sound, jugular venous stasis, hepatomegaly, peripheral edema, high level of N-terminal pro-B-type natriuretic peptide (NT-proBNP), acute depression of heart function diagnosed by echocardiography, and chest radiograph signs of congestion. Hospitalization was defined as any readmission to a hospital.

We presented continuous variables with a non-normal distribution as median (minimum; maximum), continuous variables with a normal distribution as mean ± standard deviation, and categorical variables as frequency (%). The Kolmogorov–Smirnov test was used to assess normal distribution. Student’s t test and the Mann–Whitney U test were used to compare continuous variables with normal and non-normal distributions, respectively. Box plot diagrams were presented as medians with 25th and 75th percentiles (boxes) and 5th and 95th percentiles (whiskers). Dots indicated extreme values. The Kruskal–Wallis test with Dunn’s multiple comparisons test was used for multiple comparisons. Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated in the survival analysis using Cox regression. P values < 0.05 were recognized as statistically significant. Univariate analyses were performed to study the relationship between mortality and other variables, for example, patient characteristics, medical history, heart valve insufficiency, arrhythmias, electronic cardiac disease, and drugs on admission. Predictors of mortality were identified using univariate analysis. Predictors with p < 0.05 were analyzed using Cox multivariate regression. Multivariable Cox regression to investigate predictors of mortality included the following variables: medical history such as type 2 diabetes mellitus, congestion at admission, coronary heart disease, MitraClip, non-sustained ventricular tachycardia, aldosterone antagonist, and insulin. Statistical analysis was performed using SPSS 25.0.

The median (minimum; maximum) age of patients was 66.9 years (32; 89); 55.4% of patients were aged > 65 years. Males predominated (79.6%). Ischemic cardiomyopathy (ICMP) was diagnosed in 51.7% of patients, whereas 51.3% of patients had dilated cardiomyopathy (DCMP). ACEIs were used in 56.5% of patients, whereas 24.3% of patients received ARBs before initiating sacubitril/valsartan. Among cardiovascular risk factors and comorbidities, 36.3% of patients had diabetes mellitus and 72.9% had arterial hypertension. Baseline characteristics are presented in Table 1.

Continuous improvement in LVEF was observed, from a median (minimum; maximum) of 28% (3; 65) at initiation of sacubitril/valsartan to 34% (13; 64) at 24-month follow-up (p < 0.001). The TAPSE improved from an initial median (minimum; maximum) of 17 mm (3; 31) to 18 mm (2.5; 31) at 24-month follow-up (p = 0.47). However, the largest increase of TAPSE was revealed at 12-month follow-up: 20 mm (9; 30); p=0.007. In addition, PAPsys, as an indicator of possible PH, decreased from a median (minimum; maximum) 30 mmHg (13; 115) to 25 mmHg (20; 80) at 24-month follow-up (p = 0.005) (Table 2). Data about the hemodynamic effects at follow-up are available in the appendix in the electronic supplementary material (ESM). Transthoracic echocardiography before initiation of sacubitril/valsartan revealed a higher incidence of severe and moderate valvular insufficiency, whereas numerically more valvular insufficiency was classified as mild after sacubitril/valsartan (Fig. 1). The incidence of mitral, tricuspid, and aortic insufficiency before and after sacubitril/valsartan is presented in Fig. 1.

The glomerular filtration rate decreased from a median (minimum; maximum) 55 ml/min (10; 128.8) before treatment to 49 ml/min (8; 109) after sacubitril/valsartan at 24-month follow-up (p = 0.98). The largest decrease in glomerular filtration rate was observed at 18-month follow-up: median (minimum; maximum) 48 ml/min (15; 125); p < 0.001. However, NT-proBNP values reduced significantly from a median (minimum; maximum) of 1445 pg/ml (48; 74,676) before sacubitril/valsartan to 569 pg/ml (13; 4571) at 24-month follow-up (p < 0.001) (Table 2). Data on NT-proBNP levels at follow-up are available in the ESM.

Echocardiographic and functional nonresponse occurred more in patients with ICMP than in those with DCMP (echocardiographic nonresponse 63.8 vs. 41.4%; functional nonresponse 57.7 vs. 43.3%). NT-ProBNP was significantly reduced in echocardiographic responders compared with nonresponders at 24-month follow-up (median [minimum; maximum] 452 pg/ml [13; 4203] vs. 1097 pg/ml [159; 4571]; p = 0.003). The mortality rate in echocardiographic and functional nonresponders was higher than in responders (12.1 vs. 5.2%; p = 0.1 and 11.3 vs. 3.1%; p = 0.01, respectively). The rate of congestion was significantly higher in echocardiographic nonresponders than in responders (20 vs. 4.3%; p = 0.05). Table 3 presents the differences between echocardiographic and functional responders compared with nonresponders.

Cox multivariate analysis for mortality determined congestion at admission as an independent predictor for mortality (HR 5.57; 95% CI 1.45–21.48; p = 0.01), Table 4.

The hospitalization rate was 58.3% at 12 months before treatment and 39.7% at 3 months, 41.1% at 6 months, 51.3% at 12 months, 45% at 18 months, and 51.1% at 24 months after treatment (p < 0.001). Congestion also decreased from 34.2% at treatment initiation to 10% at 24-month follow-up (p = 0.04) (Fig. 2).

Consistent with our data, Martens et al. [5] reported that improvements in LVEF after sacubitril/valsartan were probably due to reverse remodeling. TAPSE has also been observed to increase from 18.5 to 19.7 mm at 6-month follow-up after initiation of sacubitril/valsartan [11]. This result is comparable to our result, where TAPSE increased from 17 to 18 mm at 6-month follow-up. Furthermore, PAPsys has been observed to reduce from 31 to 25 mmHg at 6-month follow-up [11]. Our data showed a decrease of PAPsys from 30 to 25 mmHg at 24-month follow-up. In our cohort, PAPsys at 6-month follow-up was equivalent to that at baseline before initiation of sacubitril/valsartan. Regarding heart valve insufficiency, our data illustrated an improvement in grade of mitral, tricuspid, and aortic valve insufficiency after sacubitril/valsartan at follow-up. Kang et al. [12] reported a reduction of effective mitral regurgitant orifice area in patients treated with sacubitril/valsartan as compared with the valsartan group. In this regard, we revealed a decrease of severe and moderate mitral insufficiency at follow-up (23 and 31.1% before vs. 13.2 and 23.7% after sacubitril/valsartan). Treatment with combined sacubitril/valsartan may improve the reverse remodeling that is associated with a reduction of left ventricular volume and sphericity and may improve left ventricular function, which decreases the severity of mitral insufficiency [13]. Our study also found a reduction in the percentage of severe and moderate tricuspid and aortic insufficiency at 24-month follow-up (tricuspid 12.2 and 24.4% before vs. 7.6 and 13.6% after treatment; aortic 1.8 and 31.6% before vs. 0 and 10% after treatment). In 205 patients with HFrEF, tricuspid velocity reduced from 2.8 ± 0.55 to 2.64 ± 0.59 m/s at 6-month follow-up after sacubitril/valsartan [8]. Data about the systematic evaluation of tricuspid and aortic insufficiency remain limited.

Our study presents a reduction in NT-proBNP from a median (minimum; maximum) of 1445 pg/ml (48; 74,676) before sacubitril/valsartan to 569 pg/ml (13; 4571) at 24-month follow-up. The PIONEER-HF trial also revealed a decrease in NT-proBNP in patients with de novo and chronic HF after sacubitril/valsartan [14].

In our study, nonresponse to treatment occurred more in patients with ICMP than in those with DCMP. This showed the impact of HF etiology in the treatment of patients with HFrEF. In addition, mortality and congestion rates after initiation of sacubitril/valsartan were higher in echocardiographic and functional nonresponders than in responders. Data in this regard are limited.

We observed that all-cause hospitalization and congestion rates reduced significantly at follow-up after initiation of sacubitril/valsartan. All-cause hospitalization was 58.3% before sacubitril/valsartan at 12 months and 51.1% at 24-month follow-up after treatment. ACEIs or ARBs were used in 80.8% of all patients before sacubitril/valsartan. However, the largest decrease in hospitalization rate was 39.7% at 3-month follow-up, maybe because of superior medication adherence at the beginning of the treatment. Medication adherence was reported to be superior at the beginning of treatment compared with later. In this context, patient education about self-management minimized the risk of medication nonadherence, which was associated with a reduction in all-cause hospital readmissions and improved quality of life [15‐17]. The PARADIGM-HF trial indicated that sacubitril/valsartan was superior to enalapril in terms of HF-related hospitalization rates (12.8 vs. 15.6%) [1]. In a systematic review, all-cause hospitalizations decreased after sacubitril/valsartan compared with standard of care [18]. In our study, the incidence of congestion was 34.2% at 12 months before starting the treatment and 10% at 24-month follow-up after initiation of sacubitril/valsartan. The rate of congestion was lower in patients with HFrEF treated with sacubitril/valsartan than in those receiving enalapril [18]. Consistent with our data, congestion reduced from the beginning compared with at 19-month follow-up in the sacubitril/valsartan group [18].

This study has several limitations. It was a monocenter study, had a limited follow-up time, and had a limited number of patients. Therefore, caution is required when interpreting data. The NYHA class was evaluated without using a qualitative evaluation questionnaire. Diet and exercise habits were not systematically analyzed. Cardiac magnet resonance tomography was not used to evaluate the left ventricular and right ventricular volumes and LVEF. Some echocardiographic parameters useful for the evaluation of reverse remodeling, such as left ventricular mass index, left atrial volume index, and strain analysis, were not presented. Interobserver variability regarding echocardiographic parameters cannot be excluded.