Sacubitril-Valsartan Lowers Blood Pressure in Patients on Dialysis: A Randomized Controlled Multicenter Study

Chronic kidney disease (CKD) patients worldwide often see the need for dialysis treatment due to irreversible decline in renal function, which in turn leads to increased cardiovascular morbidity and mortality [1, 2]. The mortality rate of cardiovascular events in dialysis patients is about 20 times higher than that of those without renal disease, and three times higher than other high-risk groups such as those with diabetes [3]. On average, 10%–20% of dialysis patients die each year, of which 45% are attributed to cardiovascular events [4].

The prevalence of hypertension is substantially higher among dialysis patients compared to the general population [5, 6]. Inadequate blood pressure (BP) control is a significant concern among patients undergoing hemodialysis (HD) and peritoneal dialysis (PD). It was reported that approximately 86% of HD and 88% of PD patients were on suboptimal BP management [7, 8]. Uncontrolled office systolic BP (SBP) is a significant predictor of all-cause mortality in patients undergoing dialysis [9]. Hence, there is an urgent need to improve BP control in this population.

The angiotensin receptor-neprilysin inhibitor (ARNI) has emerged as a new agent that can significantly improve outcomes in patients with heart or renal failure [10, 11] and reduce BP in patients with hypertension [12‒15]. In China, sacubitril-valsartan received the indication for hypertension in 2021. However, the efficacy and safety of its use in dialysis patients is unclear. A recent small-sized, single-arm study showed that treatment with sacubitril-valsartan in patients with end-stage kidney disease (ESKD) and heart failure could reduce office BP by −20.7/-8.3 mm Hg [16]. To further investigate the potential benefits and risks of using ARNI in dialysis patients, we designed a prospective, randomized, controlled, multicenter trial, which might provide important insights into the antihypertensive treatment for stage 5 CKD patients on dialysis.

Original data are available upon reasonable request from corresponding authors.

The study protocol was reviewed and approved by the Ethics Committee (KY202-137) of the Ruijin Hospital Affiliated to Shanghai Jiaotong University School of Medicine, Shanghai, China and registered at ClinicalTrials.gov (NCT05243199). Eligible patients were aged between 18 and 75 years, diagnosed with CKD stage 5 and had been undergoing regular HD or PD for more than 1 month. Additionally, patients were required to have a serum N-terminal pro-brain natriuretic peptide (NT-proBNP) level of ≥ 2,000 pg/mL, and a diagnosis of hypertension, defined as an office SBP of ≥140 mm Hg or a diastolic BP (DBP) of ≥90 mm Hg or on antihypertensive medications. Patients who were on the treatment of renin-angiotensin-system (RAS) inhibitors, including angiotensin-converting enzyme inhibitor, angiotensin receptor blocker (ARB), and ARNI, were excluded, and those with a history of symptomatic hypotension or office SBP <100 mm Hg at screening were also excluded.

The eligible patients were randomly assigned to either the ARNI sacubitril-valsartan treatment or the ARB irbesartan treatment group. Specifically, we employed simple randomization using a computer-generated randomization sequence. Each eligible patient was assigned a random number between 0 and 1 generated by SPSS software, with 0.5 as the cut-off. Patients with a random number of <0.5 were allocated to the ARB group, while patients with a random number of ≥0.5 were allocated to the ARNI group. This randomization sequence was generated and held by the investigators; there were no stratification or block randomization procedures employed.

The ARNI (sacubitril-valsartan) was administered with an initial dose of 50 mg per day. If the patient was able to tolerate the drug without any adverse effects, the dose was increased to 50 mg twice a day after 1 week and then to 100 mg twice a day after 2 weeks of the treatment. In the ARB group, irbesartan 75 mg was initially administered per day and, if tolerated, then titrated to 75 mg twice a day and the maximum tolerated dosage was 150 mg twice a day during the follow-up. Any other antihypertensive agents (e.g., diuretics, calcium channel blockers, or β-blockers) that patients were taking at baseline were maintained at the same doses throughout the study, regardless of BP changes. If intolerance occurred, the dosage of the assigned study drug (ARNI or ARB) was down-titrated to the previously tolerated dose. Intolerance was defined as the occurrence of one of the following conditions: symptomatic hypotension; office or the 24-h SBP <100 mm Hg during follow-up; or serum potassium >6.0 mmol/L.

Clinical examinations and various biochemical tests were performed at enrollment, including office and 24-h ambulatory BP measurement, electrocardiogram, echocardiography, blood routine test, measurements of liver function, serum potassium, and NT-proBNP. Throughout the study, participants were regularly followed up. Measurements of office BP, serum NT-proBNP level and electrolytes, and echocardiography were repeated at 3, 6, 9, and 12 months after randomization. This study was conducted in accordance with the CONSORT guidelines for randomized controlled trials, and the CONSORT checklist is available in the online supplementary materials (for all online suppl. material, see https://doi.org/10.1159/000545195).

BP measurements were conducted before HD for HD patients, whereas for patients on PD, BP was measured during routine PD sessions. Specifically, office BP was measured using a validated upper arm electronic sphygmomanometer available at each center, such as the Omron HEM-907 or Microlife WatchBP Office, which are commonly used in clinical settings [17]. Standard cuffs were used, but for those with a large arm circumference of ≥32 cm, large-sized cuffs were utilized. Before the measurement, participants were asked to rest quietly for at least 5 min, and the BP was measured in the sitting position with their upper arm (non-venous/non-arterial fistula-side upper arm) placed at heart level. BP was measured twice at intervals of one to 2 min, and the average value of the two readings was recorded. If the difference between the two readings of SBP or DBP was more than 5 mm Hg, a third reading was taken, and the average of the three readings was recorded to ensure accuracy.

For the 24-h ambulatory BP monitoring, validated monitors available at each participating center were used, such as the Spacelabs 90,217 or Welch Allyn ABPM 7100 [17]. For HD patients, ambulatory BP is measured on non-dialysis days, while for PD patients, the measurement is done on dialysis days. BP was measured every 15–20 min during the day and every 30 min during sleep at night. A valid ambulatory recording should have at least 20 readings during daytime and 7 readings during nighttime [18].

The primary efficacy variable of the present study was the reduction in office BP after 12 months of treatment. The secondary variables were the reduction of the 24-h, daytime and nighttime ambulatory BPs, and change in serum NT-proBNP level and cardiac structure and function parameters evaluated by echocardiography. Safety variables included the incidence of death (including death from cardiovascular or other causes), hyperkalemia, and other adverse events.

It was assumed that if the ARNI treatment would result in an average office SBP reduction of 16.5 mm Hg and ARB treatment 10 mm Hg [13, 15, 19‒21], a within-group standard deviation (SD) of 20 mm Hg [19], loss of follow-up of 10%, approximately 330 patients would be needed for the study to ensure power of 80% at an overall two-sided alpha level of 0.05.

Statistical analyses were performed using SPSS statistical software version 25 (IBM Corp). Continuous data at baseline were presented as mean ± SD and compared with Student’s t test for normally distributed variables, and presented as median (interquartile range [IQR]) and compared with a non-parametric Mann-Whitney U test for variables with a skewed distribution. Categorical variables were presented as frequencies (percentages) and compared with the chi-square test. A two-sided p value of less than 0.05 was considered statistically significant.

Within-group mean changes of continuous variables from baseline to follow-up were tested using the paired Student’s t test, and between-group differences in the mean change from baseline of office BP and NT-proBNP levels after 12 months of treatment were tested using the covariance of analysis, with adjustment for the baseline levels. For the between-group comparisons in office BPs during follow-up, ANOVA analysis for repeated measurements was used. The incidence of death and hyperkalemia between the two groups was compared using the chi-square test. For missing data, we used the last observation carried forward method [22]. For the primary efficacy variable, we performed both intention-to-treat and per-protocol analyses.

Patients were screened, and baseline information was collected between August 2020 and February 2023. Figure 1 shows the flow chart of the trial. Totally, 330 eligible participants were randomized, of whom 166 received the ARNI treatment and 164 the ARB treatment. During the follow-up, among the ARNI and ARB treatment groups, 2 deaths and 8 deaths occurred, respectively, and 68 and 37 withdrew from the study because of change of dialysis centers, adverse events, or patients’ own decision; finally 96 patients on ARNI and 119 on ARB treatment completed the 12-month treatment and follow-up (Fig. 1).

Of the enrolled patients, 141 underwent HD and 189 underwent PD. The average age of patients was 57 years (range, 20–75 years), and the median dialysis duration was 3 years (range, 3 months to 20 years). The primary ESKD causes included glomerulonephritis accounting for 34.2%, followed by diabetic nephropathy 21.2%, hypertension-induced renal injury 8.5%, other etiologies 3.3% and cases with undetermined etiology 29.4% (Table 1). The mean kt/v for PD was 1.9 ± 0.4 and for HD was 1.4 ± 0.3. Baseline mean office BP was 150/82 mm Hg, median NT-proBNP was 6,336 pg/mL, and serum potassium level averaged 4.27 ± 0.57 mmol/L. Table 1 shows the baseline characteristics of the ARNI or ARB treatment. There were no significant differences between the two groups (p ≥ 0.060).

In the intention-to-treat analysis, both the 12-month ARB and ARNI treatment resulted in a significant reduction in office SBP (Table 2). The mean change from baseline was −5.1 mm Hg for the ARB treatment group and −9.8 mm Hg for the ARNI treatment group. After adjustment for the baseline SBP level, the SBP reduction remained significantly greater (p = 0.003) in the ARNI treatment group than the ARB treatment group (−10.4 mm Hg vs. −4.6 mm Hg). For office DBP, there was no significant change at 12 months compared to baseline for either the ARB or ARNI treatment in the intention-to-treat analysis.

The BP-lowering effect of the 12-month ARNI and ARB treatment was further compared in 141 HD and 189 PD patients, respectively (Table 3). In the HD patients, the mean SBP reduction was −8.7 mm Hg in the ARB group and −13.8 mm Hg in the ARNI group. After adjusting for baseline level, the adjusted mean reduction was significantly greater in the ARNI compared to ARB treatment group (−15.9 mm Hg vs. 6.6 mm Hg, p = 0.002). For DBP in HD patients, the adjusted mean reduction was also greater in the ARNI group than in the ARB group (−2.4 mm Hg vs. 1.1 mm Hg, p = 0.044). For PD patients, there were no significant differences between the two treatment groups in either SBP or DBP reduction at the 12-month follow-up (p ≥ 0.087). Subgroup analysis was further performed according to age (<60 vs. ≥ 60 years), gender (male vs. female), the median duration of dialysis (<3 vs. ≥ 3 years) and the median level of NT-proBNP (<6,336 vs. ≥ 6,336 pg/mL). Overall, the ARNI group demonstrated a more pronounced BP reduction compared to the ARB group with no statistical difference between the subgroups (p ≥ 0.462, online supplement eFigures 1 and 2), except for the comparison between the age groups (p for interaction = 0.031). In younger but not older patients, ARNI reduced more SBP than ARB.

Per-protocol analyses of the 215 patients revealed similar results. The ARNI treatment group had significantly lower office SBP and DBP during follow-up (Fig. 2) and at the 12-month visit (online eTable 1). Similar findings were observed in the 101 HD patients (online suppl. eTable 2), but not in the 114 PD patients (online suppl. eTable 3).

The 24-h ambulatory BP monitoring was successfully performed on 226 patients at baseline and 137 at the 12-month follow-up. The per-protocol analyses showed that the reductions in the 24-h SBP (−7.12 vs. −0.68 mm Hg, p < 0.001) and the daytime SBP/DBP (−5.96/−4.09 vs. 0.91/0.57 mm Hg, p ≤ 0.025), but not the nighttime BP (−5.32/-2.96 vs. −2.43/−0.32 mm Hg, p ≥ 0.103), were significantly greater in the ARNI treatment group than the ARB treatment group (online suppl. eTable 4).

At baseline, the Serum NT-proBNP level was a little but insignificantly higher in the ARNI group than the ARB group (8,145 vs. 6,812 pg/mL, p = 0.078, online suppl. eTable 5). At the 12-month follow-up, the NT-proBNP level was similar between the two groups (6,028 pg/mL in the ARNI vs. 6,664 pg/mL in the ARB, p = 0.459). The unadjusted mean change in the serum NT-proBNP was greater in the ARNI than the ARB treatment group (2,107 vs. 155 pg/mL, p = 0.028), but became statistically insignificant after adjustment for baseline levels (p = 0.085). Subgroup analyses in the PD or HD patients did not reveal any significant findings (online suppl. eTable 5). For data on echocardiographic measurements, there were also no significant differences in the mean changes between the two treatment groups (data not shown).

During the study period, 2 (1.2%) patients in the ARNI group and 8 (4.9%) patients in the ARB group died, and 5 (3.0%) patients in each group experienced a nonfatal event (Table 3). There was no statistically significant difference in the overall incidence of fatal and nonfatal events (p = 0.174). Hyperkalemia, a known risk associated with RAS inhibitors, occurred in both treatment groups (5%–11%), particularly in HD patients (11%–18%). However, there was no significant difference in serum potassium levels (p = 0.175) as well as in the incidence of hyperkalemia (serum potassium >5.5 mmol/L) between the two groups (online suppl. eTable 6). Except at the 3-month follow-up, the incidence of hyperkalemia was lower in patients treated with ARB compared to ARNI (5% vs. 11%, p = 0.028). During the study period, only 1 patient in the ARB group experienced hypotension and required dose adjustment, while no cases of hypotension were reported in the ARNI group.

The key finding of the present study is that the antihypertensive treatment with the ARNI sacubitril-valsartan compared to that with the ARB irbesartan can effectively reduce BP in dialysis patients with hypertension, especially in HD patients. The average office SBP decrease after the treatment with ARNI for 12 months was about 10 mm Hg, greater than the treatment of ARB irbesartan by around 5 mm Hg on average. The safety of the ARNI and ARB treatment was comparable and acceptable.

ESKD patients represent a unique patient population due to renal failure, which results in a variety of pathophysiological changes, such as an increased systemic volume load, activation of the RAS, and release of various cytokines, and all these can ultimately lead to sustained hypertension. In addition, ESKD patients are at increased risk of cardiovascular complications [23‒26]. Hypertension is a significant risk factor for cardiovascular morbidity and mortality in dialysis patients [3, 27]. Although achieving BP control is crucial for ESKD patients, it can be challenging [4, 28]. Only a small proportion of patients with ESKD have their BP adequately controlled; the optimum BP target in dialysis patients is not known [29, 30]. It is thus essential to prioritize research efforts to address the unmet needs of ESKD patients.

ARNI is a co-crystal composed of sacubitril and valsartan at a molar ratio of 1:1. Sacubitril can inhibit neprilysin activity, enhance the levels of natriuretic peptide, therefore relax blood vessels, lower BP, and improve myocardial remodeling [31]. While valsartan can block the receptor of angiotensin II, inhibit vasoconstriction, and reverse myocardial remodeling [31]. ARNI has become the first-line drug for the treatment of heart failure [32, 33]. As early as 2010, the first hypertension study of ARNI was published in Lancet, and the results showed that ARNI was more effective than valsartan in reducing SBP and DBP in patients with mild to severe hypertension [34]. ARNI can rapidly, potently, and continuously reduce BP in patients with heart failure or obesity or in the elderly [13, 14, 35, 36]. Studies have shown that ARNI and irbesartan were comparable in the protection of renal function in patients with estimated glomerular filtration rate of 20–60 mL/min/1.73 m [2, 15]. For patients with CKD stage 3 and below, ARNI was more effective than angiotensin-converting enzymes inhibitor in preserving renal function [11, 37]. Pharmacokinetics and pharmacodynamics studies have suggested that the ARNI sacubitril-valsartan at a dose of 100 mg BID may appear safe in patients with HD or PD, although it is important to note that available data are limited [38, 39]. In ESKD patients, several small-sized single-arm studies showed that ARNI has potential in controlling BP and improving heart function [16, 40, 41]. However, there is limited experience with ARNI use in patients with CKD stages 4–5, particularly those with end-stage renal disease (estimated glomerular filtration rate <15 mL/min/1.73 m²). Our present study, with a multicenter, randomized controlled trial design, was the first to investigate the safety and efficacy of ARNI in ESKD dialysis patients, providing valuable insights for hypertension management in this special patient population.

Previously, there was only one study focused on the BP-lowering effect of ARNI in dialysis patients [16]. The single-arm trial enrolled only 18 dialysis patients with refractory hypertension. After 12 weeks of the ARNI treatment, office SBP and DBP decreased by 22.4 mm Hg and 8.3 mm Hg from baseline, respectively. Compared to the above-mentioned study, the magnitude of BP reduction observed in our study by the ARNI treatment was smaller. Nevertheless, our results were in line with two other studies in hypertensive patients, in which the ARNI reduced more SBP by −4.3 mm Hg after 2 months [12] and −4.1 mm Hg after 3 months of treatment compared to the ARB treatment [13].

It is well-established that achieving BP control in HD patients is more challenging than in PD patients, which might be due to the rapid fluctuations in fluid and electrolyte levels during HD. Large volumes of fluid and electrolytes are removed from the blood, leading to rapid volume loss and a BP drop during HD, and a reflexive increase in heart rate and constriction of blood vessels, and finally increased BP levels post-dialysis. Furthermore, HD can activate the RAS and further increase BP. Therefore, in the current study we chose to measure office and 24-h ambulatory BP before dialysis. In addition, both treatment groups were balanced in hemoglobin levels, daily weight gain, dry weight and dialysis adequacy in terms of Kt/V at baseline and during the follow-up (online suppl. eTable 7), ensuring the between-group comparisons in BP lowering were not confounded by fluid fluctuations during the dialysis.

The active ingredients of ARNI have a high degree of binding to plasma proteins, have a high tissue distribution rate, and are not affected by dialysis [42, 43]. In HD patients treated with the ARNI 50 or 100 mg twice a day, the mean maximum plasma concentrations (C_max) of LBQ657 (active metabolite of sacubitril) and valsartan on the days between HD sessions were 15.46 ± 6.01 and 2.57 ± 1.23 mg/L, respectively. Compared with previous report in patients with severe renal impairment and healthy volunteers, these levels both remained within the safe concentration ranges [38]. The findings of our study demonstrate that ARNI confers a significant antihypertensive effect in HD patients, better than the ARB irbesartan. The use of ARNI might help ease the difficulties in controlling BP in HD patients.

In our study, we also observed a greater reduction in serum NT-proBNP after 12 months of ARNI treatment compared to ARB treatment. However, after considering the baseline level, the between-group difference became statistically insignificant. This may be attributed to the small sample size of our study and the relatively higher baseline level of NT-proBNP in the ARNI group. ARNI has been proven to effectively reduce NT-proBNP levels in heart failure patients. However, in dialysis patients, particularly those with unique pathophysiological conditions such as fluid overload and cardiac remodeling, it may take longer to decrease NT-proBNP level or require more intensive intervention to show a significant change. In the current study, a baseline level of NT-proBNP ≥2,000 pg/mL was arbitrarily set as one of the inclusion criteria, which is because that NT-proBNP level can increase dramatically by 10–20 times in dialysis patients due to impaired kidney function, and a high level of NT-proBNP (>1,800 pg/mL) was significantly associated with increased cardiovascular and mortality risk in dialysis patients [44, 45]. In line with our findings, a previous retrospective single-arm study involving 21 PD patients also observed a significant reduction in NT-proBNP levels and an improvement of symptoms of heart failure after ARNI treatment for 12 months [41].

Concerning the safety of the use of ARNI in dialysis patients, our study found a lower number of deaths in the ARNI treatment group compared to the ARB group, which could be attributed to the better BP control and anti-heart failure treatment effect of ARNI. However, further investigation with a larger sample size and longer follow-up is required to confirm the benefit of ARNI in reducing death. Additionally, it is important to pay attention to the adverse effects of hyperkalemia when treating ESKD with ARNI or ARB. Our study showed that hyperkalemia occurred in dialysis patients receiving either treatment, especially in HD patients. Therefore, regular electrolyte checks are necessary to avoid hyperkalemia.

The present study has to be interpreted in the context of its strengths and limitations. First, our study is the first randomized controlled trial of the ARNI in dialysis patients with a relatively large sample size. However, it was an open-label study and not double-blind, which might introduce some observer bias in the results and the open-label nature of the study predisposes to potential treatment bias. Second, all participants were Chinese, and thus, further studies in patients with different ethnic and geographical backgrounds will provide a more comprehensive understanding of ARNI’s efficacy and safety. Such studies can explore how genetic variability, regional differences in environmental and socioeconomic factors, lifestyle, comorbidities, and healthcare access may modify treatment outcomes. Third, the study was affected by the COVID-19 pandemic, influencing the registration and progression of our research, and causing delays in the ethical review and approval processes. Additionally, the pandemic resulted in missing data at certain follow-up points and in ambulatory BP measurements. Fourthly, the 24-h, but not the 44-h, ABPM was used in the current study as we consider that the 24-h monitoring period is sufficient to capture BP fluctuations during the day and night, especially on non-dialysis days, and aligns with our study objectives. The 44-h ABPM may provide a more comprehensive BP assessment in HD patients, especially for capturing nighttime BP patterns. On the other hand, extending the monitoring period to 44 h may increase patient discomfort, interfere with daily activities and sleep, thus potentially reducing patient compliance and affect data quality. Finally, while the results suggest that ARNI can effectively lower BP and NT-proBNP levels in dialysis patients, further research is required to determine the long-term benefit of ARNI in dialysis patients.

Our study showed that ARNI was more effective than ARB in reducing office BP as well as ambulatory BP in dialysis patients, especially in HD patients. Our study supports the use of ARNI in dialysis patients with hypertension. Further research is needed to explore its long-term benefit in dialysis patients and in multi-ethnic populations.

We are very grateful to Dr. Biyu Zhang (Hangzhou, Zhejiang) for her contributions to the study preparation during the early stage of our research. We also thank all the participants and local clinical center staff for their great contributions to the study.

The study was approved by the appropriate Institutional and/or National research the Ethics Committee (KY202-137) of the Ruijin Hospital Affiliated to Shanghai Jiaotong University School of Medicine, and the study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Written informed consent was obtained from all patients for being included in the study. Measures have been implemented to protect the privacy and personal information of research participants, with all data undergoing anonymous processing. Throughout the research process, we have consistently respected the rights of participants and provided them with the option to withdraw from the study at any time. Regarding conflicts of interest, we solemnly declare that there are no conflicts of interest that could potentially influence the research results. All research data are handled and stored in accordance with relevant regulations and ethical guidelines.

Li Lin, Weijing Bian, Qun Luo, Liang Wang, Na Liu, Min Yang, Jun Cen, Kedan Cai, Jia Hua, Hongwei Gu, Hualin Qi, Zhihong Wang, Jianying Niu, Yu Chen, Yizheng Gu, Chun Hu, Suhua Li, Yan Li, and Nan Chen declare no conflicts of interest. Xiao Li received grant from the Novartis Hypertension Scientific Research Innovation Fund (KY20211419).

The present investigator-initiated research is financially supported by the Novartis Hypertension Scientific Research Innovation Fund (KY20211419), and National Natural Science Foundation of China (Grant No. 82300796).

Trial design: Nan Chen and Xiao Li. Randomization sequence generation: Biyu Zhang and Xiao Li. Randomization implementation: Li Lin, Weijing Bian, Qun Luo, Liang Wang, Na Liu, Min Yang, Jun Cen, Kedan Cai, Jia Hua, Hongwei Gu, Hualin Qi, Zhihong Wang, Jianying Niu, Yu Chen, Yizheng Gu, Chun Hu, Suhua Li, and Xiao Li. Statistical analysis: Li Lin, Yan Li, and Xiao Li. Drafting of the manuscript: Li Lin and Xiao Li. Critical revision of the manuscript for important intellectual content: All authors. Administrative, technical, or material support and supervision: Yan Li, Nan Chen, and Xiao Li.

Trial Registration: The trail registered at the ClinicalTrials.gov (NCT05243199).

Data are not publicly available due to other ongoing research utilizing the data acquired. Please contact the corresponding author Xiao Li who will provide the data upon reasonable request.