Abstract
Rationale & Objective
Angiotensin-converting enzyme inhibitors (ACEI) are beneficial in heart failure with reduced ejection fraction (HFrEF). We sought to describe longitudinal trends in estimated glomerular filtration rate (eGFR) in HFrEF and how ACEI therapy influences these changes.
Study Design
Post-hoc analysis of trial data
Settings & Participants
Symptomatic (Treatment Trial, n=2,423) and asymptomatic (Prevention Trial, n=4,094) patients from the Studies Of Left Ventricular Dysfunction (SOLVD) trial.
Exposures
Enalapril vs placebo
Outcomes
Acute and chronic eGFR slope, and 4 kidney endpoints: 1) creatinine increase by ≥0.3 mg/dl; 2) >30% eGFR decline; 3) >40% eGFR decline; 4) incident eGFR of <30 ml/min/1.73m2
Analytic Approach
Shared parameter models, multivariable Cox regression models
Results
Baseline mean (SD) eGFR was lower in the Treatment Trial 69.5 (19.8) ml/min/1.73m2 vs 76.2 (18.6) ml/min/1.73m2 in the Prevention Trial. Following acute eGFR decline after randomization to enalapril, chronic slopes were not statistically different by randomization arm in either Treatment Trial (−0.84 in enalapril vs −1.36 ml/min/1.73m2/year in placebo; p=0.08) or Prevention Trial (−1.27 in enalapril vs −1.36 ml/min/1.73m2/year in placebo; p=0.7) over median 3-year follow-up. Randomization to enalapril increased the risk for all 4 outcomes in the Treatment Trial in the first 6-week period (HR [95% CI] were 1.48, [1.10, 1.99] for creatinine increase by ≥0.3 mg/dl; 1.38 [0.98, 1.94] for eGFR decline >30%; 2.60 [1.30, 5.21] for eGFR decline >40%; 4.71 [1.78, 12.50] for eGFR<30 ml/min/1.73m2) but after the first year was not significantly associated with increased risk. Similar, albeit less pronounced, pattern was observed in the Prevention Trial with risks only present in the early period.
Limitations
Creatinine results were not blinded, making it possible that ACEI/placebo dosing was influenced by creatinine.
Conclusion
Kidney function decline is slow in HFrEF. Although randomization to enalapril results in statistically increased risk of kidney surrogates, the risk is limited to the acute phase and chronic slopes are not different by randomized group.
Keywords: ACE inhibitor, cardiorenal, chronic kidney disease (CKD) Stage 4, heart failure with reduced ejection fraction (HFrEF), kidney function
Plain-word Summary
Certain medications called angiotensin-converting enzyme inhibitors (ACEI) are known to improve survival in patients with heart failure with reduced ejection fraction (HFrEF). They can be associated with reductions in one measurement of kidney function called estimated glomerular filtration rate (eGFR) in the short-term, but the longer-term effects on eGFR in patients with HFrEF are unknown. Using two large trials of HFrEF patients, we observed that being randomized to ACEI therapy did lead to an initial reduction in eGFR but that this effect did not persist over the longer term. These findings may encourage clinicians that ACEIs will not only promote survival but also have no detrimental, albeit also no beneficial, longer-term effect on kidney function in patients with HFrEF.
Introduction
Blockade of the renin-angiotensin-aldosterone system (RAAS) with either angiotensin-converting enzyme inhibitors (ACEI) or angiotensin-receptor blockers (ARB) has been shown to be beneficial in reducing the risk of cardiovascular events in patients with heart failure with reduced ejection fraction (HFrEF) (1, 2), and is recognized as an integral component of the management of patients with HFrEF (3). In the kidney, ACEI/ARB therapy leads to dilation of the efferent arteriole, and thereby can be associated with acute declines in estimated glomerular filtration rate (eGFR). Prior trials in patients with diabetes have shown that acute declines in eGFR following initiation of ACEI/ARB therapy are associated with slower long-term declines (4), but whether this trend also applies to HFrEF is unclear.
It is well-recognized that reduced level of kidney function is a strong predictor of poor outcomes in the HFrEF population, with the risk of mortality increased in those with eGFR <60 ml/min/1.73 m2; this risk increases progressively as eGFR declines (5–8). However, few studies have assessed longitudinal trends in eGFR in patients with HFrEF. In the general non-HFrEF population, eGFR declines of 30–40% or greater and progression to chronic kidney disease (CKD) Stage 4 (as defined by an eGFR of 15–29 ml/min/1.73 m2) are both associated with increased risk for death and hospitalizations (9–12); yet, incidence of these kidney function endpoints among patients with HFrEF is not well described. Whether initiation of ACEI/ARB therapy, given its acute effects on decreasing eGFR, would hasten progression to these surrogate endpoints in patients with HFrEF, or alternatively slow progression because of improved cardiac remodeling and decreasing glomerular pressure, is also unclear.
Using the Studies Of Left Ventricular Dysfunction (SOLVD) trials, we sought to describe the longitudinal changes in kidney function over the trial’s multiyear follow-up period based on randomization to enalapril or placebo. We also sought to assess whether enalapril use influenced the risk of reaching various surrogate kidney function endpoints including increases in creatinine of ≥0.3 mg/dl, eGFR declines of >30–40%, and incident CKD stage 4 or 5 as defined by a new eGFR <30 ml/min/1.73 m2.
Methods
Study Population and Design
The SOLVD studies were two National Heart, Lung and Blood Institute sponsored multicenter, double-blind, randomized controlled trials that evaluated the effects of the ACEI enalapril versus placebo on mortality in patients with HFrEF (ejection fraction <35%) (13, 14). Patients who were symptomatic (manifested by dyspnea, evidence of fluid retention and diuretic requirement) were enrolled into the Treatment Trial, and those who were asymptomatic and not requiring diuretics or vasodilators were enrolled into the Prevention Trial. Pertinent exclusion criteria included serum creatinine of >2.5 mg/dl, age >80 years, uncontrollable hypertension, or suspected renal artery stenosis. Patients were randomized to either enalapril or placebo. Enrollment into the trial was performed on a rolling basis from July 1986–1989, with trial closure on January 31, 1991 for the Treatment trial and August 31, 1991 for the Prevention trial (15). Patients without baseline and/or follow-up creatinine measurements (n=263) were excluded from the present analysis. Participants in SOLVD provided informed consent at the time of enrollment; the present study was deemed exempt from review by the institutional ethics review board.
Exposure
Initial dosing of enalapril was 5 mg twice daily, which was uptitrated to 10 mg twice daily at the 2-week follow-up visit as tolerated by patient symptoms. Uptitration of enalapril was encouraged unless participants reported dizziness or fainting.
Outcomes
The kidney function outcomes of interest included the annual rate of eGFR decline as well as the occurrence of four kidney function endpoints: (1) increase in serum creatinine of ≥0.3 mg/dl as this has been used in prior studies of patients with HFrEF to define acute kidney injury (16–18) and in studies in the general population is associated with adverse outcomes; (2) decline of eGFR of >30% from baseline and (3) decline of eGFR of >40% since these were both considered surrogate end-points in trials of CKD (9, 12, 19); and (4) incident eGFR of <30 ml/min/1.73 m2 given the high risk for adverse outcomes and propensity for development of complications from the loss of renal clearance (10, 11). The time at risk for each of these outcomes began at the time of randomization into the respective trials. Serum creatinine was measured at the time of randomization, at 2- and 6-week visits, and annually thereafter. Measurements that occurred at other time-points, besides those pre-specified by the trial protocol were excluded, since the indication for checking kidney function was unclear. Measured serum creatinine values were lowered by 5%, as is usual practice for serum creatinine values measured prior to assay standardization (20). GFR was estimated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) calculation (21). While the CKD-EPI equation has not yet been validated in the heart failure population, prior published findings have shown its accuracy as compared to other estimating equations in the general population, as well as in patients with CKD and HF (21–23).
Covariates
While our analyses were based on randomized groups, several covariates were selected a priori and included in our regression models given their clinical relevance. These included demographic characteristics (age, sex, race), comorbid conditions (hypertension, diabetes), cardiovascular characteristics (New York Heart Association functional class, ischemic etiology of left ventricular dysfunction, current smoking), diuretic use, and baseline eGFR.
Statistical Analysis
We used descriptive statistics to compare the characteristics of patients randomized to enalapril versus placebo separated by trial (Treatment or Prevention) given differences in the severity of illness and thus potentially differing associations between enalapril and kidney function declines. Normally distributed continuous variables were compared using the t-test, and non-normally distributed variables were evaluated using the Wilcoxon Mann-Whitney test. Categorical variables were compared using the χ2 test and Fisher’s exact test as appropriate.
Trend of mean eGFR data among randomized arms within each trial was visualized and two-slope linear mixed effects models were used to estimate the annual rate of decline in eGFR with a knot at 6-weeks to account for the observed acute ACEI-related effect on slope. A shared parameter model was also included to account for informative dropout due to death. Kaplan-Meier estimates of the percentage of patients who were outcome-free were performed seperately by trial for each of the four kidney function outcomes. The association between enalapril and the risks of each endpoint were assessed using multivariable Cox proportional hazards regression models in the early (<6 weeks), middle (6 weeks to 1 year), and late (>1 year) follow-up periods given a violation in the proportional hazards assumption in the Treatment Trial. Patients were censored at the date of their last kidney function measurement. As a sensitivity analysis, discrete time proportional hazards models were also used to estimate risk of each outcome given that outcome events could only be ascertained at certain visit time points. Death was treated as a censoring event in the primary analysis, then alternatively as a competing event in sensitivity analyses. Time to event analyses were repeated using the 6-week timepoint as baseline in sensitivity analyses. Models were adjusted for the covariates described above. All analyses were performed using SAS Enterprise Guide (Version 7.12, Cary, NC) and R language (version 3.3.1, R Foundation for Statistical Computing, Vienna Austria).
Results
A total of 2,423 patients were included from the Treatment Trial and 4,094 from the Prevention Trial. The mean (SD) age was 60 (10) years in the Treatment Trial and 59 (10) years in the Prevention Trial, with slightly higher percentage of women in Treatment than in Prevention at 19% and 11% respectively (Table 1). While baseline characteristics including prevalence of hypertension and diabetes and baseline laboratory values were similar across randomized groups within each trial, there were notable differences between the Treatment Trial and Prevention Trials (Table 1). There was a greater prevalence of diabetes, hypertension and diuretic use among those in the Treatment Trial. Baseline kidney function was also worse among those in the Treatment Trial, with an eGFR of 69.5±19.8 ml/min/1.73 m2 as compared to 76.2±18.6 ml/min/1.73 m2 in the Prevention Trial.
Table 1.
Baseline characteristics according to randomized groups and by trial.
| Characteristic | TREATMENT TRIAL | PREVENTION TRIAL | ||
|---|---|---|---|---|
| Enalapril | Placebo | Enalapril | Placebo | |
| (n = 1207) | (n = 1216) | (n = 2047) | (n = 2047) | |
| Age – years | 60 ± 10 | 61 ± 10 | 59 ± 10 | 59 ± 11 |
| Female | 226 (19) | 241 (20) | 229 (11) | 227 (11) |
| Black | 190 (16) | 172 (14) | 183 (9) | 197 (10) |
| eGFR – ml/min/1.73 m2 | 70.2 ± 19.9 | 68.7 ± 19.6 | 76.1 ± 18.5 | 76.3 ± 18.7 |
| eGFR < 60 ml/min/1.73 m2 | 403 (33) | 445 (37) | 420 (21) | 425 (21) |
| Diabetes | 293 (24) | 324 (27) | 313 (15) | 312 (15) |
| Hypertension | 508 (42) | 498 (41) | 748 (37) | 769 (38) |
| Ischemia | 849 (70) | 876 (72) | 1714 (84) | 1708 (83) |
| NYHA Functional Class | ||||
| NYHA Class I | 135 (11) | 133 (11) | 1359 (66) | 1372 (67) |
| NYHA Class II | 684 (57) | 684 (56) | 687 (34) | 672 (33) |
| NYHA Class III/IV | 388 (32) | 399 (33) | 1 (0) | 3 (0) |
| Hematocrit–% | 42.5 ± 4.8 | 42.0 ± 4.8 | 42.8 ± 4.4 | 43.1 ± 4.3 |
| Previous MI | 805 (67) | 792 (65) | 1658 (81) | 1640 (80) |
| Smoking, current | 276 (23) | 259 (21) | 463 (23) | 486 (24) |
| Diuretic Use | 1032 (86) | 1031 (85) | 329 (16) | 349 (17) |
Values are represented by mean ± standard deviation or n (%). GFR estimated using the Chronic Kidney Disease Epidemiology Collaboration equation. NYHA, New York Heart Association; eGFR, estimated glomerular filtration; MI, myocardial infarction
Rate of Decline in eGFR
Median length of follow-up was 34 months, with a range of 0.5 to 62.3 months, reflective of rolling enrollment into the trials. Of the patients who were alive and reached each follow-up visit, there was no substantial difference in missing eGFR data by randomization arm in either the Treatment or Prevention Trials (Table S1). Overall, there were acute declines in eGFR that were more pronounced in the enalapril arms of both trials followed by parallel progressive decline in eGFR over the remaining course of follow-up (Figure 1). In the first 6 weeks following randomization, the rate of decline was significantly more rapid among the enalapril arms as compared to the placebo arms, although there was an observed acute decline among all arms (Table 2). In the follow-up period after the 6-week visit, the slope in the enalapril arm of the Treatment Trial was −0.84 (95% CI −1.24, −0.43) ml/min/1.73 m2/year, not significantly different from −1.36 (95% CI −1.77, −0.94) ml/min/1.73 m2/year in the placebo arm (difference of 0.52 [95% CI −0.06, 1.09], p=0.08). In the Prevention Trial, the chronic slopes within the enalapril and placebo arms were also not significantly different at −1.27 (95% CI −1.56, −0.99) and −1.36 (95% CI −1.65, −1.07) respectively (difference of 0.09 [−0.31, 0.49] ml/min/1.73 m2/year; p=0. 7).
Figure 1.
Mean estimated glomerular filtration rates (eGFR) over course of follow-up, stratified by trial and by randomization arm. Dotted lines represent 95% confidence intervals. Slopes were not significantly different between the Treatment and Prevention Trials (p=0.5).
Table 2.
Estimated intercept and rates of decline in estimated glomerular filtration rate (eGFR) per year stratified by trial and randomized groups, in early and late follow-periods
| Trial | Follow-up Period | eGFR intercept, ml/min/1.73 m2 and eGFR slope, ml/min/1.73 m2 per year (95% CI) | |||
|---|---|---|---|---|---|
| Enalapril | Placebo | Enalapril–Placebo | p-value | ||
| Treatment Trial | Intercept | 69.8 (68.7, 70.9) | 68.6 (67.5, 69.7) | 1.16 (−0.38, 2.69) | 0.1 |
| Initial 6 weeks (acute slope) | −27.33 (−33.57, −21.09) | −7.78 (−14.07, −1.48) | −19.55 (−28.41, −10.69) | <0.001 | |
| After 6 weeks (chronic slope) | −0.84 (−1.24, −0.43) | −1.36 (−1.77, −0.94) | 0.52 (−0.06, 1.09) | 0.08 | |
| Prevention Trial | Intercept | 75.9 (75.1, 76.7) | 76.1 (75.3, 76.9) | −0.18 (−1.29, 0.93) | 0.7 |
| Initial 6 weeks (acute slope) | −12.86 (−17.28, −8.43) | −5.85 (−10.28, −1.42) | −7.01 (−13.27, −0.75) | 0.03 | |
| After 6 weeks (chronic slope) | −1.27 (−1.56, −0.99) | −1.37 (−1.65, −1.07) | 0.09 (−0.31, 0.49) | 0.7 | |
A two-slope linear mixed effects model using longitudinal eGFR with a knot at 6 weeks was used to determine slopes. Model included random effects for the intercept and the acute (before 6-weeks) and chronic (after 6-weeks) slope, and fixed effects of treatment and treatment by time interaction. A shared parameter model was used to account for drop-out caused by death
eGFR, estimated glomerular filtration rate
Increase in serum creatinine by ≥0.3 mg/dl
There were 607 (25.1%) patients enrolled in the Treatment Trial and 607 (14.8%) patients enrolled in the Prevention Trial who experienced an increase in serum creatinine by ≥0.3 mg/dl (Table 3). Differences in reaching this endpoint by randomization group appeared greatest early in follow-up in the Treatment Trial and remained parallel between the randomization groups thereafter (Figure 2), with hazard ratio (HR) not meeting significance in the follow-up period beyond 1-year (Table 4). For the Prevention Trial, enalapril was associated with a 22% higher risk of meeting this outcome (HR=1.22 [95% CI 1.04, 1.43]), with no significant interaction with time (Table 5). For both trials, when risk was assessed starting at the 6-week visit, randomization to enalapril was no longer significantly associated with increased risk (Table S2).
Table 3.
Unadjusted and adjusted event rates per 100-person years (95% CI) stratified by study and randomization arm
| Enalapril | Placebo | |||||||
|---|---|---|---|---|---|---|---|---|
| Study | Outcome | Model | N | Events | Event rate | N | Events | Event rate |
| Treatment Trial | Creatinine Increase by ≥0.3 mg/dl | Unadjusted | 1207 | 349 | 18.3 (16.5, 20.3) | 1216 | 258 | 13.2 (11.7, 14.9) |
| Adjusted | 17.7 (15.9, 19.8) | 12.8 (11.3, 14.5) | ||||||
| Decline in eGFR by >30% | Unadjusted | 1207 | 298 | 14.9 (13.3, 16.7) | 1216 | 208 | 10.2 (8.9, 11.7) | |
| Adjusted | 13.6 (12.1, 15.3) | 9.4 (8.2, 10.8) | ||||||
| Decline in eGFR by >30% | Unadjusted | 1207 | 133 | 6.0 (5.0, 7.1) | 1216 | 96 | 4.5 (3.7, 5.4) | |
| Adjusted | 5.4 (4.5, 6.5) | 4.1 (3.4, 5.1) | ||||||
| Incident eGFR<30 ml/min/1.73 m2 | Unadjusted | 1195 | 61 | 2.6 (2.1, 3.4) | 1207 | 46 | 2.1 (1.6, 2.8) | |
| Adjusted | 0.9 (0.6, 1.3) | 0.6 (0.4, 0.9) | ||||||
| Prevention Trial | Creatinine Increase by ≥0.3 mg/dl | Unadjusted | 2047 | 334 | 8.9 (8.0, 9.9) | 2047 | 273 | 7.3 (6.5, 8.2) |
| Adjusted | 8.3 (7.4, 9.3) | 6.8 (6.0, 7.7) | ||||||
| Decline in eGFR by >30% | Unadjusted | 2047 | 288 | 7.5 (6.7, 8.4) | 2047 | 236 | 6.2 (5.5, 7.1) | |
| Adjusted | 6.5 (5.7, 7.3) | 5.4 (4.7, 6.2) | ||||||
| Decline in eGFR by >40% | Unadjusted | 2047 | 111 | 2.7 (2.3, 3.3) | 2047 | 85 | 2.1 (1.7, 2.7) | |
| Adjusted | 2.2 (1.8, 2.7) | 1.7 (1.4, 2.2) | ||||||
| Incident eGFR<30 ml/min/1.73 m2 | Unadjusted | 2042 | 31 | 0.7 (0.5, 1.1) | 2040 | 27 | 0.7 (0.5, 1.0) | |
| Adjusted | 0.3 (0.2, 0.5) | 0.2 (0.1, 0.4) | ||||||
Analyses performed using Poisson regression with the logarithm of the follow-up time as an offset parameter
Adjusted for age, sex, race, NYHA functional class, current smoking, ischemic etiology of left ventricular dysfunction, diuretic use and baseline eGFR. eGFR=estimated glomerular filtration rate
Figure 2.
Kaplan-Meier plots of outcome-free survival, separated into the Treatment Trial and the Prevention Trial. Plots of incidence of creatinine increase by ≥0.3 mg/dl, decline in estimated glomerular filtration rate (eGFR) by >30%, decline in eGFR by >40% and incident eGFR to <30 ml/min/1.73 m2. The number at risk can be found below each time separate plot.
Table 4.
Hazard ratios for kidney outcomes in association with randomization to enalapril for the Treatment Trial
| Enalapril | Placebo | Interaction with time p-value | Period of Follow-up | |||
|---|---|---|---|---|---|---|
| < 6 weeks | 6 weeks to 1 year | > 1 year | ||||
| Creatinine Increase By ≥0.3 mg/dl | ||||||
| Unadjusted HR (95% CI) | 1.37 (1.17, 1.61) | Ref | 0.013 | 1.49 (1.11, 2.00) | 1.51 (1.12, 2.04) | 1.21 (0.94, 1.56) |
| *Adjusted HR (95% CI) | 1.37 (1.17, 1.61) | Ref | 0.014 | 1.48 (1.10, 1.99) | 1.50 (1.11, 2.03) | 1.21 (0.95, 1.56) |
| Decline in eGFR By >30% | ||||||
| Unadjusted HR (95% CI) | 1.44 (1.21, 1.72) | Ref | 0.043 | 1.41 (1.01, 1.98) | 1.65 (1.16, 2.34) | 1.36 (1.05, 1.76) |
| *Adjusted HR (95% CI) | 1.43 (1.20, 1.70) | Ref | 0.056 | 1.38 (0.98, 1.94) | 1.62 (1.14, 2.29) | 1.36 (1.05, 1.76) |
| Decline in eGFR By >40% | ||||||
| Unadjusted HR (95% CI) | 1.34 (1.04, 1.75) | Ref | 0.005 | 2.64 (1.32, 5.29) | 1.21 (0.74, 1.99) | 1.16 (0.82, 1.65) |
| *Adjusted HR (95% CI) | 1.32 (1.02, 1.72) | Ref | 0.003 | 2.60 (1.30, 5.21) | 1.19 (0.73, 1.96) | 1.14 (0.80, 1.62) |
| Incident eGFR<30 ml/min/1.73 m2 | ||||||
| Unadjusted HR (95% CI) | 1.29 (0.88, 1.89) | Ref | 0.014 | 4.24 (1.60, 11.25) | 0.81 (0.41, 1.61) | 1.03 (0.59, 1.82) |
| *Adjusted HR (95% CI) | 1.48 (1.01, 2.17) | Ref | 0.012 | 4.71 (1.78, 12.50) | 0.92 (0.47, 1.84) | 1.21 (0.69, 2.14) |
Analyses performed using Cox proportional hazards ratios. Given significant interaction with time in the Treatment trial, follow-up period was divided into 3 separate intervals of <6-weeks, 6-weeks to 1 year, and >1 year.
Adjusted for age, sex, race, NYHA functional class, current smoking, ischemic etiology of left ventricular dysfunction, diuretic use and baseline eGFR.
eGFR = estimated glomerular filtration rate; HR = hazard ratio; CI = confidence interval; ref = reference
Table 5.
Hazard ratio for kidney outcomes in association with randomization to enalapril for the Prevention Trial
| Enalapril | Placebo | Interaction with time p-value | Period of Follow-up | |||
|---|---|---|---|---|---|---|
| < 6 weeks | 6 weeks to 1 year | > 1 year | ||||
| Creatinine Increase by ≥0.3 mg/dl | ||||||
| Unadjusted HR (95% CI) | 1.22 (1.04, 1.43) | Ref | 0.706 | 1.01 (0.75, 1.36) | 1.62 (1.16, 2.27) | 1.19 (0.95, 1.50) |
| *Adjusted HR (95% CI) | 1.22 (1.04, 1.43) | Ref | 0.638 | 1.01 (0.75, 1.37) | 1.64 (1.17, 2.29) | 1.19 (0.94, 1.49) |
| Decline in eGFR By >30% | ||||||
| Unadjusted HR (95% CI) | 1.21 (1.02, 1.44) | Ref | 0.846 | 0.95 (0.66, 1.36) | 1.50 (1.05, 2.14) | 1.22 (0.97, 1.55) |
| *Adjusted HR (95% CI) | 1.21 (1.02, 1.43) | Ref | 0.990 | 0.95 (0.66, 1.36) | 1.50 (1.05, 2.15) | 1.21 (0.96, 1.53) |
| Decline in eGFR By >40% | ||||||
| Unadjusted HR (95% CI) | 1.28 (0.96, 1.69) | Ref | 0.683 | 1.05 (0.57, 1.90) | 1.40 (0.72, 2.71) | 1.34 (0.93, 1.93) |
| *Adjusted HR (95% CI) | 1.29 (0.97, 1.71) | Ref | 0.871 | 1.05 (0.58, 1.91) | 1.42 (0.73, 2.75) | 1.35 (0.94, 1.96) |
| Incident eGFR<30 ml/min/1.73 m2 | ||||||
| Unadjusted HR (95% CI) | 1.12 (0.67, 1.87) | Ref | 0.352 | 1.19 (0.36, 3.91) | 1.13 (0.41, 3.13) | 1.08 (0.54, 2.17) |
| *Adjusted HR (95% CI) | 1.17 (0.69, 1.98) | Ref | 0.279 | 1.15 (0.35, 3.77) | 1.16 (0.42, 3.20) | 1.18 (0.59, 2.36) |
Analyses performed using Cox proportional hazards ratios. Given significant interaction with time in the Treatment trial, follow-up period was divided into 3 separate intervals of <6-weeks, 6-weeks to 1 year, and >1 year. Even though this interaction with time did not reach significance in the Prevention trial, follow-up period was divided in similar fashion for consistency.
Adjusted for age, sex, race, NYHA functional class, current smoking, ischemic etiology of left ventricular dysfunction, diuretic use and baseline eGFR.
eGFR = estimated glomerular filtration rate; HR = hazard ratio; CI = confidence interval; ref=reference
Decline in eGFR by >30%
Among those enrolled in the Treatment Trial, 506 (20.9%) experienced this outcome, while 524 (12.8%) had this outcome in the Prevention Trial (Table 3). Differences in reaching this endpoint by randomization group appeared early in follow-up in the Treatment Trial and remained parallel between the randomization groups thereafter (Figure 2). The differences by randomized group were smaller in the Prevention Trial (Figure 2). Randomization to enalapril carried a significantly higher hazard of reaching this endpoint in both the Treatment and Prevention Trial with HR=1.43 (95% CI 1.20, 1.70) and HR=1.21 (95% CI 1.02, 1.43) respectively, with significant interaction with time over the follow-up period noted in the Treatment Trial but not in the Prevention Trial (Tables 4 and 5 respectively).
Decline in eGFR by >40%
Fewer patients reached this endpoint with 229 (9.5%) in the Treatment Trial and 196 (4.8%) in the Prevention Trial (Table 3). Differences in reaching this endpoint by randomized group appeared in the Treatment Trial but not in the Prevention Trial, where Kaplan-Meier curves were mostly overlapping (Figure 2). In the Treatment Trial, randomization to enalapril carried a significantly higher hazard of this endpoint in the first 6 weeks of follow-up with a HR=2.60 (95% CI 1.30, 5.21) that then diminished and no longer reached significance in the subsequent intervals of follow-up (Table 4). In the Prevention Trial, randomization to enalapril was not significantly associated with an increased risk of an eGFR decline by >40% nor was there an interaction with time over the follow-up period (Table 5).
Incident eGFR <30 ml/min/1.73 m2
One hundred and seven (4.5%) patients in the Treatment Trial and 58 (1.4%) patients in the Prevention Trial reached this endpoint (Table 3). Kaplan-Meier curves for incidence of this endpoint among the enalapril and placebo arms were largely overlapping in both the Treatment and the Prevention Trials (Figure 2). Randomization to enalapril in in the Treatment Trial was significantly associated with a HR=4.71 (95% 1.78, 12.50) in the early follow-up period, but not in the subsequent intervals of follow-up (Table 4). In the Prevention Trial, randomization to enalapril was not significantly associated with an increased hazard of reaching this endpoint in either unadjusted or adjusted analysis (HR=1.12, 95% CI 0.67, 1.87 in unadjusted analysis; HR=1.17 [95% CI 0.69, 1.98] in adjusted analysis).
Given the increased rate of events in the first 6-week period, analyses were repeated using the kidney function from the 6-week point as baseline and showed that randomization to enalapril was not significantly associated with outcomes in either trial (Table S2). Analyses incorporating death as a competing risk for all the above endpoints did not significantly change the results (Table S2). Estimates of risks of each outcome at each follow-up time were not different between Cox proportional hazards regression estimates and logistic regressions (Table S3).
Discussion
This study demonstrated overall slow progression of eGFR decline among patients with HFrEF, with only a small percentage of patients reaching incident CKD Stage 4–5 over a median 3-year follow-up. Despite starting with lower baseline kidney function, patients with symptomatic HFrEF (Treatment Trial) also had slow overall rates of eGFR decline despite an initial pronounced acute decline after randomization to enalapril. Among both symptomatic and asymptomatic patients, randomization to enalapril carried a higher risk of experiencing an increase in serum creatinine of ≥0.3 mg/dl, an eGFR decline of >30%−40%, or reaching incident CKD Stage 4–5 although these differences were small and the risks were no longer significant when the initial acute follow-up period was excluded.
The current paradigm of cardiorenal syndrome would suggest that patients with more advanced and symptomatic HFrEF, such as those in the Treatment Trial, would tend to have more rapidly progressive kidney function decline from various mechanisms such as reduced renal perfusion, increased renal congestion, and/or neurohormonal activation. However, few studies have systematically examined the natural course of kidney function among patients with HFrEF. While we did find that baseline eGFR among those in the Treatment Trial started about 6.7 ml/min/1.73 m2 lower than the Prevention Trial, the slopes of decline were very similar. That is, our observed chronic rates of decline, in the follow-up period after the initial 6-weeks among the symptomatic and asymptomatic patients were −0.84 and −1.36 ml/min/1.73m2 in the enalapril and placebo arms of the Treatment Trial and −1.27 and −1.36 ml/min/1.73m2 in the enalapril and placebo arms of the Prevention Trial, respectively. These rates are also not very different from those observed in healthier community cohorts such as −0.90 ml/min/1.73 m2/year (cohort of healthy men) (24) and −1.52 ml/min/1.73 m2/year (cohort of patients with hypertension) (25).
While use of ACEI/ARB treatment is known to lead to an acute decline in eGFR, there has been limited data describing long-term kidney function after initiation of these medications among HFrEF patients. We observed that the chronic slope among those randomized to enalapril was not significantly different from the slope among those randomized to placebo, both within the Treatment Trial as well as the Prevention Trial. Following an acute initial decline, the eGFR curves remain parallel throughout the course of follow-up, suggesting that there is no long-term detriment from a kidney perspective with use of enalapril but also no evidence of benefit to preservation of kidney function either. These results add to the small number of prior analyses demonstrating that even though there can be an acute decline in eGFR shortly after starting ACEI/ARB treatment, there is generally no rapid long-term progression. A post-hoc analysis of the Heart failure Endpoint evaluation of Angiotensin II Antagonist Losartan (HEAAL) trial, which comprised of 3834 patients with HFrEF randomized to either high dose losartan 150 mg daily or low dose losartan of 50 mg daily, found slightly faster rates of eGFR decline among the high dose group that was largely driven by early changes (26). The eGFR curves then remained largely parallel after the first 4 months. The conclusion from that study was similar to our study in that there is an early drop in eGFR, but no evidence of accelerated progression of kidney function decline with higher doses of ARB therapy, and no benefit on slowing the decline of eGFR.
In contrast, ACEI/ARB therapy in patients with diabetic nephropathy has been shown to slow eGFR decline when compared to placebo or alternative antihypertensives (4, 27). There are two possible explanations for the difference. First, a 3-year follow-up may not be sufficient enough to demonstrate a plateau of the eGFR slope in the enalapril arm versus a progressive decline in the placebo arm, especially considering the degree of the initial acute drop in eGFR. Second and more likely, is the fact that the disease process is different in diabetic nephropathy, which initially manifests with proteinuria. Treatment with ACEI/ARB decreases proteinuria and thereby slows decline of renal function. Significant proteinuria is often not seen in patients with HFrEF. The current dataset did not have detailed data on proteinuria; however, the Prospective Comparison of ARNI with ACEI to Determine Impact on Global Mortality and Morbidity in Heart Failure (PARADIGM-HF) trial demonstrated that despite enalapril leading to a decrease in proteinuria, the rate of decline was not attenuated over their median 27-months of follow-up. This is consistent with the paradigm that kidney function decline in HFrEF patients is not primarily associated with proteinuria and therefore ACEI/ARB therapy may not necessarily slow eGFR decline as in diabetic nephropathy.
With regards to our surrogate kidney function endpoints, the current analysis found that randomization to enalapril among the Treatment Trial was consistently associated with a significantly increased hazard of meeting all endpoints in the early 6-week follow-up period when compared to those randomized to placebo. For the Prevention Trial, randomization to enalapril was only associated with increased risk for an increase in serum creatinine by ≥0.3 mg/dl and having a >30% eGFR decline. Importantly, it should be emphasized that when analyses were repeated excluding the initial 6-week follow-up period, there was no longer any significant risk of meeting any of these endpoints associated with randomization with enalapril in either trial. While acute changes in kidney function after ACEI/ARB initiation have been previously demonstrated in HFrEF, (28–30), there are few prior data regarding reaching these endpoints over long-term follow-up. While some of the patients in SOLVD may have reached the kidney function endpoints as a result of acute decline from initiation of enalapril (as shown by the early separation of the enalapril and placebo curves on Kaplan-Meier plots), occurrence of the outcomes continued throughout the length of follow-up and were parallel between the groups. For patients with HFrEF, it remains unclear whether reaching the kidney function endpoints are predictors for poor outcomes as they are in the general CKD population, or can be considered relatively benign changes if they are occurring in the setting of ACEI/ARB therapy.
This study has several strengths. The analysis was performed using data from 2 landmark, rigorously conducted randomized controlled trials with long follow-up periods. Kidney function was measured frequently and based on protocol-driven intervals, and measurements of kidney function that were obtained outside of pre-specified intervals were excluded. Consideration was taken with death considered both as a competing and a non-competing event and eGFR slopes were calculated using mixed effects modeling, allowing for incorporation of all values as well as with shared parameter modeling to account for drop-out due to death. With regards to study limitations, patients with a serum creatinine of >2.5 mg/dl were excluded from the original SOLVD trials, and thus applying these findings to patients with more advanced CKD may not be appropriate. Dosing of enalapril was encouraged to be maximized in all patients, but it is possible that some clinicians were modifying doses based on changes of serum creatinine while others were not. Urinalyses were not collected as part of the trial, preventing incorporation of the proteinuric response to enalapril. It is also important to acknowledge that the SOLVD trials were conducted from 1986 to 1991; therapies for patients with HFrEF have evolved significantly in the time since. However, despite extraordinary achievements in therapies for advanced heart failure, the clinical question regarding the hemodynamic eGFR response to ACEI/ARB initiation in HFrEF remains highly relevant.
Decline in eGFR on average was relatively slow in both symptomatic and asymptomatic patients with HFrEF and only a small percentage reached incident CKD Stage 4–5 over a median follow-up of 3-years. Randomization to enalapril only minimally hastened the progression of decline in kidney function. Despite a slightly increased risk of reaching several kidney function surrogate endpoints, the overall incidence of these endpoints was low and occurred early. RAAS inhibitors have been shown to have a clear mortality benefit for patients with HFrEF as demonstrated in multiple landmark trials including SOLVD, and it is encouraging that their use appears to have minimal added risk of detriment to kidney function although admittedly also no kidney benefit.
Supplementary Material
Acknowledgments
Sources of Funding
NIH Training Grant T32 DK007777. Funders had no role in study design, data collection, analysis, reporting, or the decision to submit for publication.
Abbreviations
- SOLVD
Studies Of Left Ventricular Dysfunction
- HFrEF
Heart Failure with reduced Ejection Fraction
- RAAS
renin-angiotensin-aldosterone system
- eGFR
estimated glomerular filtration rate
- CKD
chronic kidney disease
- ACEI
angiotensin converting enzyme inhibitor
- ARB
angiotensin-receptor blockers
- CI
confidence interval
- HR
hazard ratio
- SD
standard deviation
Footnotes
None of the authors have any conflicts of interest to disclose.
References
- 1.Effects of enalapril on mortality in severe congestive heart failure. Results of the Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS). The New England journal of medicine. 1987;316(23):1429–35. Epub 1987/06/04. doi: 10.1056/nejm198706043162301. [DOI] [PubMed] [Google Scholar]
- 2.Carson P, Tognoni G, Cohn JN. Effect of Valsartan on hospitalization: results from Val-HeFT. Journal of cardiac failure. 2003;9(3):164–71. Epub 2003/06/20. doi: 10.1054/jcaf.2003.22. [DOI] [PubMed] [Google Scholar]
- 3.Yancy CW, Jessup M, Bozkurt B, Butler J, Casey DE Jr., Drazner MH, et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation. 2013;128(16):e240–327. Epub 2013/06/07. doi: 10.1161/CIR.0b013e31829e8776. [DOI] [PubMed] [Google Scholar]
- 4.Holtkamp FA, de Zeeuw D, Thomas MC, Cooper ME, de Graeff PA, Hillege HJ, et al. An acute fall in estimated glomerular filtration rate during treatment with losartan predicts a slower decrease in long-term renal function. Kidney international. 2011;80(3):282–7. Epub 2011/04/01. doi: 10.1038/ki.2011.79. [DOI] [PubMed] [Google Scholar]
- 5.Dries DL, Exner DV, Domanski MJ, Greenberg B, Stevenson LW. The prognostic implications of renal insufficiency in asymptomatic and symptomatic patients with left ventricular systolic dysfunction. Journal of the American College of Cardiology. 2000;35(3):681–9. Epub 2000/03/15. [DOI] [PubMed] [Google Scholar]
- 6.McAlister FA, Ezekowitz J, Tonelli M, Armstrong PW. Renal insufficiency and heart failure: prognostic and therapeutic implications from a prospective cohort study. Circulation. 2004;109(8):1004–9. Epub 2004/02/11. doi: 10.1161/01.cir.0000116764.53225.a9. [DOI] [PubMed] [Google Scholar]
- 7.Hillege HL, Girbes AR, de Kam PJ, Boomsma F, de Zeeuw D, Charlesworth A, et al. Renal function, neurohormonal activation, and survival in patients with chronic heart failure. Circulation. 2000;102(2):203–10. Epub 2000/07/13. [DOI] [PubMed] [Google Scholar]
- 8.Khan NA, Ma I, Thompson CR, Humphries K, Salem DN, Sarnak MJ, et al. Kidney function and mortality among patients with left ventricular systolic dysfunction. Journal of the American Society of Nephrology : JASN. 2006;17(1):244–53. Epub 2005/11/18. doi: 10.1681/asn.2005030270. [DOI] [PubMed] [Google Scholar]
- 9.Coresh J, Turin TC, Matsushita K, Sang Y, Ballew SH, Appel LJ, et al. Decline in estimated glomerular filtration rate and subsequent risk of end-stage renal disease and mortality. Jama. 2014;311(24):2518–31. Epub 2014/06/04. doi: 10.1001/jama.2014.6634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sud M, Tangri N, Pintilie M, Levey AS, Naimark DM. Progression to Stage 4 chronic kidney disease and death, acute kidney injury and hospitalization risk: a retrospective cohort study. Nephrology, dialysis, transplantation : official publication of the European Dialysis and Transplant Association - European Renal Association. 2016;31(7):1122–30. Epub 2015/11/22. doi: 10.1093/ndt/gfv389. [DOI] [PubMed] [Google Scholar]
- 11.Abboud H, Henrich WL. Clinical practice. Stage IV chronic kidney disease. The New England journal of medicine. 2010;362(1):56–65. Epub 2010/01/08. doi: 10.1056/NEJMcp0906797. [DOI] [PubMed] [Google Scholar]
- 12.Inker LA, Lambers Heerspink HJ, Mondal H, Schmid CH, Tighiouart H, Noubary F, et al. GFR decline as an alternative end point to kidney failure in clinical trials: a meta-analysis of treatment effects from 37 randomized trials. American journal of kidney diseases : the official journal of the National Kidney Foundation. 2014;64(6):848–59. Epub 2014/12/03. doi: 10.1053/j.ajkd.2014.08.017. [DOI] [PubMed] [Google Scholar]
- 13.Yusuf S, Pitt B, Davis CE, Hood WB, Cohn JN. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. The New England journal of medicine. 1991;325(5):293–302. Epub 1991/08/01. doi: 10.1056/nejm199108013250501. [DOI] [PubMed] [Google Scholar]
- 14.Yusuf S, Pitt B, Davis CE, Hood WB Jr., Cohn JN. Effect of enalapril on mortality and the development of heart failure in asymptomatic patients with reduced left ventricular ejection fractions. The New England journal of medicine. 1992;327(10):685–91. Epub 1992/09/03. doi: 10.1056/nejm199209033271003. [DOI] [PubMed] [Google Scholar]
- 15.Studies of left ventricular dysfunction (SOLVD)--rationale, design and methods: two trials that evaluate the effect of enalapril in patients with reduced ejection fraction. The American journal of cardiology. 1990;66(3):315–22. Epub 1990/08/01. [DOI] [PubMed] [Google Scholar]
- 16.Forman DE, Butler J, Wang Y, Abraham WT, O’Connor CM, Gottlieb SS, et al. Incidence, predictors at admission, and impact of worsening renal function among patients hospitalized with heart failure. Journal of the American College of Cardiology. 2004;43(1):61–7. Epub 2004/01/13. [DOI] [PubMed] [Google Scholar]
- 17.Metra M, Nodari S, Parrinello G, Bordonali T, Bugatti S, Danesi R, et al. Worsening renal function in patients hospitalised for acute heart failure: clinical implications and prognostic significance. European journal of heart failure. 2008;10(2):188–95. Epub 2008/02/19. doi: 10.1016/j.ejheart.2008.01.011. [DOI] [PubMed] [Google Scholar]
- 18.Damman K, Valente MA, Voors AA, O’Connor CM, van Veldhuisen DJ, Hillege HL. Renal impairment, worsening renal function, and outcome in patients with heart failure: an updated meta-analysis. European heart journal. 2014;35(7):455–69. Epub 2013/10/30. doi: 10.1093/eurheartj/eht386. [DOI] [PubMed] [Google Scholar]
- 19.Levey AS, Inker LA, Matsushita K, Greene T, Willis K, Lewis E, et al. GFR decline as an end point for clinical trials in CKD: a scientific workshop sponsored by the National Kidney Foundation and the US Food and Drug Administration. American journal of kidney diseases : the official journal of the National Kidney Foundation. 2014;64(6):821–35. Epub 2014/12/03. doi: 10.1053/j.ajkd.2014.07.030. [DOI] [PubMed] [Google Scholar]
- 20.Levey AS, Coresh J, Greene T, Marsh J, Stevens LA, Kusek JW, et al. Expressing the Modification of Diet in Renal Disease Study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clinical chemistry. 2007;53(4):766–72. Epub 2007/03/03. doi: 10.1373/clinchem.2006.077180. [DOI] [PubMed] [Google Scholar]
- 21.Inker LA, Schmid CH, Tighiouart H, Eckfeldt JH, Feldman HI, Greene T, et al. Estimating glomerular filtration rate from serum creatinine and cystatin C. The New England journal of medicine. 2012;367(1):20–9. Epub 2012/07/06. doi: 10.1056/NEJMoa1114248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Manzano-Fernandez S, Flores-Blanco PJ, Perez-Calvo JI, Ruiz-Ruiz FJ, Carrasco-Sanchez FJ, Morales-Rull JL, et al. Comparison of risk prediction with the CKD-EPI and MDRD equations in acute decompensated heart failure. Journal of cardiac failure. 2013;19(8):583–91. Epub 2013/08/06. doi: 10.1016/j.cardfail.2013.05.011. [DOI] [PubMed] [Google Scholar]
- 23.Oh J, Kang SM, Hong N, Youn JC, Han S, Jeon ES, et al. The CKD-EPI is more accurate in clinical outcome prediction than MDRD equation in acute heart failure: data from the Korean Heart Failure (KorHF) Registry. International journal of cardiology. 2013;167(3):1084–7. Epub 2012/11/30. doi: 10.1016/j.ijcard.2012.10.054. [DOI] [PubMed] [Google Scholar]
- 24.Lindeman RD, Tobin J, Shock NW. Longitudinal studies on the rate of decline in renal function with age. Journal of the American Geriatrics Society. 1985;33(4):278–85. Epub 1985/04/01. [DOI] [PubMed] [Google Scholar]
- 25.Hanratty R, Chonchol M, Miriam Dickinson L, Beaty BL, Estacio RO, MacKenzie TD, et al. Incident chronic kidney disease and the rate of kidney function decline in individuals with hypertension. Nephrology Dialysis Transplantation. 2010;25(3):801–7. doi: 10.1093/ndt/gfp534. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kiernan MS, Gregory D, Sarnak MJ, Rossignol P, Massaro J, Kociol R, et al. Early and late effects of high- versus low-dose angiotensin receptor blockade on renal function and outcomes in patients with chronic heart failure. JACC Heart failure. 2015;3(3):214–23. Epub 2015/03/07. doi: 10.1016/j.jchf.2014.11.004. [DOI] [PubMed] [Google Scholar]
- 27.Bjorck S, Mulec H, Johnsen SA, Norden G, Aurell M. Renal protective effect of enalapril in diabetic nephropathy. BMJ (Clinical research ed). 1992;304(6823):339–43. Epub 1992/02/08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Testani JM, Kimmel SE, Dries DL, Coca SG. Prognostic importance of early worsening renal function after initiation of angiotensin-converting enzyme inhibitor therapy in patients with cardiac dysfunction. Circulation Heart failure. 2011;4(6):685–91. Epub 2011/09/10. doi: 10.1161/circheartfailure.111.963256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Jose P, Skali H, Anavekar N, Tomson C, Krumholz HM, Rouleau JL, et al. Increase in creatinine and cardiovascular risk in patients with systolic dysfunction after myocardial infarction. Journal of the American Society of Nephrology : JASN. 2006;17(10):2886–91. Epub 2006/08/25. doi: 10.1681/asn.2006010063. [DOI] [PubMed] [Google Scholar]
- 30.Damman K, Solomon SD, Pfeffer MA, Swedberg K, Yusuf S, Young JB, et al. Worsening renal function and outcome in heart failure patients with reduced and preserved ejection fraction and the impact of angiotensin receptor blocker treatment: data from the CHARM-study programme. European journal of heart failure. 2016;18(12):1508–17. Epub 2016/07/19. doi: 10.1002/ejhf.609. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.