Cardiac and kidney outcomes after sacubitril-valsartan therapy: recovery of cardiac function relative to kidney function decline

Study patients

Data were retrieved from the Severance Clinical Research Analysis Portal (SCRAP). Clinical data at Severance Hospital are de-identified and loaded on a clinical data warehouse (CDW). The CDW data were accessed and customized through the institution’s proprietary software, known as SCRAP. A total of 8,604 patients with HFrEF who first took sacubitril-valsartan or valsartan were retrospectively enrolled from May 2015 to June 2022 using diagnostic codes recorded in the International Classification of Diseases, 10th revision (ICD-10) clinical modification (I50.00 or I50.04). Patients who did not meet the medication usage criteria were excluded from the study cohort (n = 4,075). In addition, patients with end-stage kidney disease (n = 144), no available baseline estimated glomerular filtration rate (eGFR) (n = 1,453), no eGFR follow-up data within 7 days, 8–89 days, and 90 days to 1 year of drug initiation (n = 1,957), and those diagnosed with sepsis, pneumonia, or urinary tract infection within 30 days of drug initiation (n = 71) were further excluded. A total of 512 patients were included in the cohort 1. From this cohort, patients were divided into two groups, AKI and non-AKI groups, based on whether they experienced AKI. The definition of AKI, which is an increase in the serum creatinine level of ≥0.3 mg/dL within 48 hours or >1.5 times than the baseline value within 7 days, followed the KDIGO (Kidney Disease: Improving Global Outcomes) guidelines [15]. Furthermore, we excluded 41 patients without transthoracic echocardiography (TTE) follow-up and one patient who underwent dialysis from the AKI group, creating cohort 2 (n = 42). Each cohort was then categorized into two groups: one group receiving sacubitril-valsartan and the other group receiving valsartan (Fig. 1).

Figure 1.

Flow diagram of the study cohort.

AKI, acute kidney injury; ARB, angiotensin receptor blocker; eGFR, estimated glomerular filtration rate; ESKD, end-stage kidney disease; HFrEF, heart failure with reduced ejection fraction; TTE, transthoracic echocardiography.

Data collection and measurements

Baseline demographic data and medical history, including age, history of hypertension, diabetes mellitus, dyslipidemia, coronary artery disease, and chronic kidney disease (CKD), were collected at drug initiation. Medical conditions regarding comorbidities were collected based on ICD-10 codes. Hypertension and diabetes mellitus histories were further collected if they had received any medication. Anthropometric data, including height, weight, and BP measurements, were also collected. Data on blood samples, including hemoglobin, albumin, creatinine, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and electrolytes, were obtained prior to drug initiation. Serum creatinine levels were measured using an isotope-dilution mass spectrometry-traceable method, and the eGFR was calculated using the CKD Epidemiology Collaboration equation. The data for eGFR were collected at baseline and within 7 days of drug initiation for AKI evaluation, 8 to 89 days for AKI recovery evaluation, and 90 days to 1 year for long-term kidney outcome evaluation. Additionally, data on TTE measurements were obtained for assessing cardiac function, and ejection fraction (EF) was collected at baseline and during follow-up. Follow-up TTE results were collected between 1 month and 1 year after the initiation of either sacubitril-valsartan or valsartan treatment.

Exposure and study outcome assessment

The exposure was a start of sacubitril-valsartan or valsartan for the first time. A 3-month washout period was established to exclude prior use of other ARBs before starting the study medication. During this washout period, participants were excluded if they had taken any ARBs other than valsartan in the sacubitril-valsartan group. For the valsartan group, participants were excluded if they had taken any other form of ARBs during the washout period. The index date was defined as the date of first drug initiation (Fig. 2). We further excluded cases where both drugs were prescribed simultaneously, and where they were prescribed in doses not within the standard range: 50–400 mg for sacubitril-valsartan and 80–320 mg for valsartan. Moreover, to ensure medication adherence, patients were included in the study if they had medication adherence of 80% or more during the 90-day period starting from the index date.

Figure 2.

Study flow and outcomes over time.

(A) Cohort 1, for kidney outcome. (B) Cohort 2, for cardiac and composite outcomes.

AKD, acute kidney disease; AKI, acute kidney injury; TTE, transthoracic echocardiography.

According to each cohort, various outcomes were analyzed. For cohort 1, short- and long-term kidney changes were analyzed compared to baseline. For short-term kidney outcomes, we assessed the occurrence of AKI within 7 days of drug initiation and further evaluated its recovery within 8 to 89 days after drug initiation in each drug group. AKI recovery was defined as a return of serum creatinine to ≤150% of baseline without dialysis, and non-recovery was defined as the need for dialysis or no return of serum creatinine to ≤150% of baseline. Additionally, long-term kidney outcomes, specifically declines in eGFR to 30%, 40%, or 50% beyond 90 days to 1 year from the index date, were analyzed within each drug group. Lastly, for cohort 2, we analyzed the changes in cardiac function and the composite outcome. Composite outcome was defined as changes in cardiac function in relation to changes in kidney function using the following calculation.

Statistical analysis

Baseline characteristics and clinical outcomes were presented according to sacubitril-valsartan and valsartan use. Continuous variables are presented as mean ± standard deviation and categorical variables are presented as number and percentage. Two-sample t test was used to evaluate the difference in eGFR and EF between the two groups. Additionally, a paired t test was employed to analyze the changes in eGFR and EF between baseline and follow-up. The ΔEF/ΔeGFR ratio between the two groups was compared using both the two-sample t test and the Mann-Whitney U test for parametric and nonparametric methods, respectively. Moreover, multivariable linear regression was conducted to examine the association between two drugs and ΔEF/ΔeGFR. Adjustments were performed with covariates having a significance level of 0.1 in the univariate analysis, including systolic BP, history of coronary artery disease, hemoglobin, serum calcium, sodium, and the use of spironolactone. All statistical analyses were performed using Stata version 18 (Stata Corp.), with statistical significance determined at p-value of <0.05.

Baseline characteristics

The baseline characteristics of the 512 patients are presented in Table 1. A total of 233 and 279 patients were included in the sacubitril-valsartan and valsartan groups, respectively. The mean age of the patients was 68.3 ± 15.1 years, and 294 of them (57.4%) were male . Patients in the valsartan group were older and had a higher BP. Patients in the sacubitril-valsartan group had a higher prevalence of diabetes mellitus but showed no significant difference in other comorbidities. Baseline laboratory data showed that the mean eGFR of total patients was 76.5 ± 28.7 mL/min/1.73 m², and no significant difference between drug groups. Levels of N-terminal pro–B-type natriuretic peptide (NT-proBNP), which is an indicator of the degree of HF, were significantly higher in the sacubitril-valsartan group (7,038.0 ± 9,098.8 vs. 5,067.4 ± 7,954.7, p = 0.03). In addition, anti-hypertensive drugs such as beta-blockers, furosemide, and spironolactone were more frequently used in the sacubitril-valsartan group along with high use of sodium-glucose cotransporter-2 inhibitors.

Table 1.

Baseline characteristics of cohort 1 according to sacubitril-valsartan and valsartan groups

Cohort 1: short-term kidney outcome after drug initiation

From cohort 1, AKI occurred in 72 patients, with one patient requiring dialysis due to AKI. Patients who experienced AKI were older and had a higher prevalence of CKD compared to the non-AKI group. In laboratory parameters, the AKI group had lower albumin and eGFR levels as well as higher levels of NT-proBNP than the non-AKI group. Additionally, there was no significant difference in medication use between the two groups (Supplementary Table 1, available online). Among AKI patients, 37 (15.9%) were in the sacubitril-valsartan group, and 35 (12.5%) were in the valsartan group, with no significant difference between these two groups (p = 0.28). Furthermore, 29 (78.4%) in the sacubitril-valsartan group, and 25 (17.4%) in the valsartan group recovered from AKI but showed no significant difference between the two groups (Table 2).

Table 2.

Short and long-term kidney outcomes after drug initiation in cohort 1

Cohort 1: long-term kidney outcome after drug initiation

Long-term changes in kidney function after drug initiation are presented in Table 2. For long-term kidney outcomes, a decline in eGFR from baseline between 90 days and 1 year after the index date across both groups was compared. A decrease in eGFR by 30%, 40%, and 50% occurred in 173 (33.8%), 110 (21.5%), and 63 patients (12.3%), respectively, with a higher rate in the sacubitril-valsartan group compared to the valsartan group. However, no statistically significant difference was observed between the two drug groups. When comparing long-term kidney outcomes separately in the AKI and non-AKI groups, the decline in eGFR was higher in the sacubitril-valsartan group compared to the valsartan group within the AKI cohort; however, these differences did not reach statistical significance (Supplementary Table 2, available online). Similarly, there was no significant difference between the two drugs in the non-AKI cohort.

Kidney function, the eGFR values, over time and the difference compared to baseline are presented in Supplementary Table 3 (available online). In sacubitril-valsartan and valsartan groups, the eGFR decreased by 16.4% and 15.4% from baseline, respectively, 1 month after starting the medication. Compared to the change in eGFR 1 month after medication use, the declines in eGFR at 3, 6, and 12 months were minimal. Throughout all follow-up periods, including baseline, there were no statistically significant differences in eGFR values between the two medication groups. Although a decline in eGFR was observed after medication use, the dosage of the medication for the patients did not decrease; instead, the dose increased in both drug groups. While a decrease in eGFR was observed following the use of the two drugs, the drug concentration tended to increase. Additionally, potassium levels showed a slight increase in both drug groups after drug initiation, but there was no significant difference between the groups (Supplementary Table 4, available online).

Cohort 2: changes in kidney and cardiac function in patients with acute kidney injury

The baseline characteristics of patients in cohort 2 are presented in Table 3. The mean age of the patients was 71.9 ± 14.3 years, and 25 of them (59.5%) were male . Patients in the sacubitril-valsartan group had lower systolic BP and had higher use of spironolactone. The differences in cardiac and kidney function between drug initiation and follow-up in each group are shown in Table 4. There was a significant EF improvement in the sacubitril-valsartan group after follow-up (31.3 ± 10.2 vs. 43.2 ± 12.4, p for paired t test < 0.001), while no improvement in the valsartan group (46.6 ± 17.2 vs. 48.6 ± 15.6, p for paired t test = 0.32). Moreover, the difference in EF and ratio between before and after drug initiation showed a significant difference between the two drug groups, showing greater improvement in the sacubitril-valsartan group compared to the valsartan group. This trend was similarly found in both CKD and non-CKD groups. In addition, a decline in eGFR was prominent in both drug groups, with a similar pattern in both the CKD and non-CKD subgroups.

Table 3.

Characteristics of cohort 2 in the sacubitril-valsartan and valsartan groups

Table 4.

Effects of sacubitril-valsartan and valsartan in patients with acute kidney injury in cohort 2

To further analyze the degree of improvement in cardiac function compared to the degree of decrease in kidney function, the ΔEF/ΔeGFR ratio was used. The parametric method showed that the mean ΔEF/ΔeGFR ratio was –1.76 ± 2.58 in the sacubitril-valsartan group and –0.20 ± 0.58 in the valsartan group, indicating that sacubitril-valsartan had a positive effect on improving cardiac function compared to kidney function decline after drug initiation (p = 0.03). The nonparametric method also showed an advantage in long-term cardiac function protection against temporary kidney failure decline, with a ΔEF/ΔeGFR ratio of –0.76 (–1.61 to –0.09) in the sacubitril-valsartan group (p = 0.006) (Table 5). Similar results were found in CKD subgroups; however, there was no statistical significance in non-CKD subgroups. Due to several factors that can influence kidney and cardiac function, we conducted a linear regression analysis. In the multiple regression analysis, after adjusting for all significant variables, the ΔEF/ΔeGFR ratio was significantly 1.27 lower with sacubitril-valsartan compared to valsartan (Supplementary Table 5, available online). This suggests that sacubitril-valsartan leads to a greater improvement in cardiac function relative to kidney function compared to valsartan.

Table 5.

Analysis of △EF compared to △eGFR between each group stratified by parametric and nonparametric methods