Abstract
Introduction
We previously reported the efficacy and safety of low-dose (12.5 mg/day) spironolactone for chronic kidney disease (CKD) with diabetes. Few studies have examined the characteristics of patients who may have reduced urinary albumin-creatinine ratio (UACR) on mineralocorticoid receptor antagonists. In this study, we aimed to identify the clinical characteristics of patients prone to benefit from UACR reduction with low-dose spironolactone.
Methods
This was a post hoc analysis of a previous trial and included 55 patients assigned to the spironolactone group. Univariate regression analysis was performed to determine the association between the change in UACR after 24 weeks of low-dose spironolactone administration and baseline exploratory parameters. Multiple regression analysis was conducted on the associated parameters, and regression models were created for analysis. A similar analysis was performed for changes in serum potassium levels and estimated glomerular filtration rate (eGFR) after 24 weeks of spironolactone administration.
Results
In the univariate analysis, baseline UACR, triglyceride levels, and eGFR were associated with changes in UACR. The regression coefficient estimates were significant for baseline UACR, triglyceride levels, and eGFR (p = 0.002, 0.017, and 0.003, respectively). The reduction in UACR was greater with higher baseline UACR and triglyceride levels, and lower baseline eGFRs. The increase in serum potassium levels due to low-dose spironolactone administration showed a negative correlation with baseline serum potassium levels and no correlation with baseline eGFR, suggesting its safety.
Conclusions
It may not be too late to start treatment with low-dose spironolactone, even in patients with relatively advanced CKD with diabetes.
Keywords: Type 2 diabetes mellitus, Chronic kidney disease with diabetes, Urinary albumin-creatinine ratio, Mineralocorticoid receptor antagonist, Spironolactone, Post hoc analysis
Introduction
Diabetes-induced chronic kidney disease (CKD) is a major cause of end-stage kidney disease. Recently, nearly 40% of patients undergoing dialysis have been reported to have diabetes [1, 2]. The general clinical course of CKD with diabetes is the appearance of urinary albumin in the early stages of the disease, followed by progression to overt proteinuria and a progressive decline in renal function leading to end-stage renal failure. Therefore, urinary albumin excretion is widely used as an indicator of diabetes-induced CKD progression. Clinical trials have shown that a decrease in urinary albumin excretion after drug therapy significantly correlates with the efficacy of renal failure treatment [3]. Although controversial, urinary albumin excretion is a useful surrogate marker of end-stage kidney disease, and it is important to reduce urinary albumin excretion to prevent the deterioration of renal function.
Mineralocorticoid receptor antagonists (MRAs) reportedly reduce urinary albumin excretion by inhibiting the hyperactivation of mineralocorticoid receptors that cause inflammation and fibrosis; hence, providing renal protection. Finerenone is a selective nonsteroidal MRA that has been reported to improve composite renal outcomes in clinical trials of patients with diabetes-induced CKD (FIDELIO-DKD/FIDELITY) [4, 5]; therefore, its use is now recommended by the guidelines of many countries. In these studies, finerenone reduced the urinary albumin-to-creatinine ratio (UACR) by 32.0–39.9% from baseline [4, 5]. Notably, in a post hoc analysis of the FIDELITY study, the reduction in UACR with finerenone was reported to mediate 84% of the treatment effect on the composite renal outcome [6], suggesting that much of the renal protective effect of MRA is mediated by improvements in the UACR.
Spironolactone is the oldest MRA available. It has long been reported that the addition of spironolactone (25–100 mg/day) to a renin-angiotensin-aldosterone system (RAS) inhibitor further reduces urinary albumin excretion in patients with diabetes [7–11]. However, hyperkalemia is frequently observed at spironolactone doses of 25 mg/day or more [8]; hence, its active use is avoided in clinical practice. We conducted a multicenter randomized controlled trial and found that spironolactone at 12.5 mg/day, half the lowest dose ever used, significantly reduced the UACR at 24 weeks compared with controls [12]. Furthermore, we found that low-dose spironolactone is safe to use, with no cases of hyperkalemia, resulting in a potassium level of more than 5.5 mEq/L. Spironolactone is a very inexpensive drug compared with other MRAs; therefore, in terms of cost-effectiveness, low-dose spironolactone has the potential for widespread clinical use.
Consequently, reducing the UACR with MRAs improves composite renal outcomes. However, few studies have examined the characteristics of patients who may have reduced UACR on MRAs (responders). Identifying responders to low-dose spironolactone, and consequently to MRAs, could allow clinicians to use MRAs more effectively. Therefore, we conducted a post hoc analysis of the low-dose spironolactone trial in order to identify the clinical characteristics of patients who may benefit from UACR reduction with low-dose spironolactone.
Methods
Study Design and Inclusion Criteria for the Low-Dose Spironolactone Trial
The low-dose spironolactone trial is a multicenter, randomized, open-label, placebo-controlled, parallel-group trial. The participants were adult patients with type 2 diabetes and CKD treated with RAS inhibitors. The inclusion criteria included patients with: diabetes aged 20 years or older, a UACR of 30 mg/gCr or more, an estimated glomerular filtration rate (eGFR) of 30 mL/min/1.73 m2 or more calculated from serum creatinine levels, and RAS inhibitor use for 180 days or more. The exclusion criteria included patients with: serum potassium levels of 5.0 mEq/L or more, prior use of MRA, history of adverse events with MRA, and pregnancy or lactation. The participants were randomized into two groups: one group received spironolactone (12.5 mg/day) in addition to their RAS inhibitor, and the other group received 24 weeks of spironolactone-free medical therapy. Follow-up visits were conducted at 12 and 24 weeks to confirm changes from baseline in serum potassium levels, eGFR, and other parameters. Further details are available in the initial report of this trial [12].
Primary Outcomes and Endpoints
The primary outcome of this study was to identify the participant characteristics that influenced the effect of the low-dose spironolactone trial on UACR reduction. The secondary outcome was to evaluate safety (change in serum potassium levels and eGFR by administration of low-dose spironolactone). Similar to the previous trial [12], the primary endpoint in this analysis was the change in UACR from baseline to the 24-week intervention period. The secondary endpoint was the change in serum potassium levels and eGFR from baseline to the 24-week intervention period.
Study Population
The previous trial [12] included 66 patients randomized into a low-dose spironolactone administration group. However, of these, six had a UACR of less than 30 mg/gCr at baseline or unmeasured UACR, resulting in 60 patients in the full analysis set. Furthermore, 5 patients were noncompliant or failed to complete the 24-week intervention period, resulting in a per-protocol set of 55 patients. Therefore, 55 patients were included in the final analysis.
Statistical Analysis
Baseline participant characteristics are summarized using descriptive statistics, with continuous variables expressed as medians and interquartile range. Binary variables are expressed as frequencies (number of participants [N] and percentages [%]).
The primary endpoint was the change in UACR, defined as the value obtained by subtracting the baseline UACR from the UACR at 24 weeks after spironolactone administration. When the baseline UACR and change in UACR data were not normalized using the Shapiro-Wilk test, the sinh-arcsinh normal distribution transformation (SHASH) was applied to normalize data distribution, as in the previous trial [12]. In the primary outcome analysis, we performed a univariate analysis with change in UACR (SHASH-transformed variables) as the objective variable and the participants’ baseline demographic information (continuous variables such as age, body mass index, blood pressure, and serum creatinine, eGFR, cystatin C, serum potassium, uric acid, glycated hemoglobin, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, aldosterone, and baseline UACR levels; or binary variables such as sex, and sodium-glucose cotransporter inhibitor use) as explanatory variables. A regression model was constructed using parameters that showed a statistically significant association in the univariate analysis as explanatory variables. Additionally, to consider baseline UACR values, we analyzed the regression model that added baseline UACR (SHASH-transformed variables) as a covariate to the explanatory variables. The estimated regression coefficients, standard errors, p values, standard beta, and variance inflation factors (VIFs) were calculated for each model. The threshold for the VIFs was set to 10, and if it was higher than this, it was judged that there was a problem of multicollinearity in the model. A supplementary analysis was also performed for F-UACR (UACR/eGFR × 100), which is the value dividing UACR by eGFR.
Next, we performed a supplementary analysis in which change in blood pressure was added as an explanatory variable to evaluate the effect of change in blood pressure on change in UACR. Further, to advance clinical understanding and application, a supplementary analysis using the responder classification was also performed. Cases in which UACR decreased by 30% or more from baseline were defined as the responder group, and cases in which UACR decreased by less than 30% from baseline were defined as the nonresponder group. The two groups were compared in baseline clinical parameters. Continuous variables were analyzed using the Mann-Whitney U test, and binary variables were analyzed using Fisher’s exact test.
Regarding the secondary outcome analysis for the safety evaluation, we created a regression model with the change in serum potassium levels as the objective variable and baseline eGFR and covariate baseline potassium levels as the explanatory variables. Additionally, we created a regression model with changes in eGFR as the objective variable, baseline potassium levels, baseline aldosterone, and covariate baseline eGFR as explanatory variables.
A two-sided p value of less than 0.05 was considered statistically significant. All analyses were conducted using JMP Pro 17 software (SAS Institute, Inc., Cary, NC, USA).
Results
Because the statistical normality of the baseline UACR and the change in UACR were excluded, SHASH transformation was performed according to the method. The transformed data showed a normal distribution (shown in Table 1). Therefore, the change in UACR and baseline UACR transformed by SHASH (SHASH [change in UACR] and SHASH [baseline UACR], respectively) were used as objective variables and covariates in this study. Table 2 shows the background information of the 55 patients in the low-dose spironolactone group. The median participant age was 71 years, eGFR was 66.7 mL/min/1.73 m2, and UACR was 155.0 mg/gCr; hence, the participants primarily consisted of elderly patients with preserved renal function but high UACRs.
Table 1.
Before and after data conversion of UACRs
| | Before data conversion | After data conversion | ||
|---|---|---|---|---|
| baseline UACR | change in UACR | SHASH [baseline UACR] | SHASH [change in UACR] | |
| Mean | 385.111 | −108.024 | 0.067 | −0.015 |
| SD | 617.270 | 339.937 | 1.048 | 1.007 |
| SE | 83.233 | 45.837 | 0.141 | 0.136 |
| Skewness | 2.777 | −3.674 | 0.231 | −0.014 |
| Kurtosis | 7.593 | 19.109 | 0.359 | 0.055 |
| Shapiro-Wilk | 0.572 | 0.649 | 0.976 | 0.986 |
| p value | <0.001* | <0.001* | 0.340 | 0.759 |
Change in UACR, the value obtained by subtracting the baseline UACR from the UACR at 24 weeks after spironolactone administration; UACR, urine albumin-creatinine ratio; SHASH, sinh-arcsinh normal distribution transformation; SD, standard deviation; SE, standard error of the mean.
*p < 0.05.
Table 2.
Baseline participant characteristics
| | | N = 55 |
|---|---|---|
| Age, years | 71 | (66–78) |
| Male, N | 38 | (69.0) |
| BMI, kg/m2 | 26.6 | (23.5–28.8) |
| Systolic blood pressure, mm Hg | 132.0 | (122.0–141.5) |
| Diastolic blood pressure, mm Hg | 70.0 | (63.0–79.5) |
| UACR, mg/gCre | 155.0 | (73.4–348.2) |
| Serum creatinine, mg/dL | 0.85 | (0.70–0.95) |
| eGFR, mL/min/1.73 m2 | 66.7 | (57.5–73.5) |
| Cystatin C, mg/L | 0.99 | (0.86–1.11) |
| Serum potassium, mmol/L | 4.20 | (4.0–4.4) |
| UA, μmol/L | 339.0 | (285.5–377.7) |
| HbA1c, % | 7.2 | (6.8–7.7) |
| Triglyceride, mmol/L | 1.65 | (1.10–2.26) |
| HDL cholesterol, mmol/L | 1.27 | (1.07–1.49) |
| LDL cholesterol, mmol/L | 2.55 | (2.13–2.85) |
| Aldosterone, pg/mL | 104.0 | (73.3–127.1) |
| Duration of diabetes, N | ||
| ≥10 years | 18 | (32.7) |
| <10 years | 37 | (67.3) |
| eGFR, N | ||
| ≥60 mL/min/1.73 m2 | 40 | (72.7) |
| <60 mL/min/1.73 m2 | 15 | (27.3) |
| UACR, N | ||
| 30 to <300 mg/gCre | 38 | (69.1) |
| ≥300 mg/gCre | 17 | (30.9) |
BMI, body mass index; UACR, urine albumin-to-creatinine ratio; eGFR, estimated glomerular filtration rate; UA, uric acid; HDL, high-density lipoprotein; LDL, low-density lipoprotein; HbA1c, glycated hemoglobin.
The univariate analysis was performed prior to the analysis of the main outcomes, and it was found that SHASH (change in UACR) was significantly associated with SHASH (baseline UACR) (R2 = 0.223, p < 0.001), baseline triglyceride (R2 = 0.176, p = 0.002), and baseline eGFR (R2 = 0.113, p = 0.012) (shown in Table 3). The correlations among these variables are presented in online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000549942). A regression model was constructed using SHASH (change in UACR) as the objective variable, and SHASH (baseline UACR), baseline triglyceride, and baseline eGFR as the explanatory variables. Furthermore, it was constructed using the interaction between SHASH (baseline UACR) and baseline eGFR and between SHASH (baseline UACR) and baseline triglyceride as covariates (shown in Table 4). Consequently, the VIF in this model was between 1.137 and 2.787, and it was judged that there was no multicollinearity. Additionally, the regression coefficient estimates (p values) of this model showed significant values of −0.387 (p = 0.002), −0.002 (p = 0.017), and 0.021 (p = 0.003) for SHASH (baseline UACR), baseline triglyceride, and baseline eGFR, respectively. In contrast, the interaction terms between baseline eGFR and SHASH (baseline UACR) and between baseline triglyceride and SHASH (baseline UACR) were not significant, indicating that the association between baseline eGFR and baseline triglyceride with the objective variable SHASH (change in UACR) was not affected by SHASH (baseline UACR). These results suggest a reduction in UACR in cases with high baseline UACR and triglyceride levels, as well as in cases with low baseline eGFRs (shown in Fig. 1). Although triglycerides had the highest standardized partial regression coefficients (β: −0.426), the power was 0.683, which was lower than baseline UACR and baseline eGFR. Moreover, neither baseline aldosterone nor sodium-glucose cotransporter inhibitor use significantly correlated with the SHASH (change in UACR). Next, considering the possibility that the reduction in UACR may be affected by renal function, we performed a supplementary analysis similar to that for UACR on F-UACR (UACR/eGFR × 100), which is the value dividing UACR by eGFR. A regression model was constructed using SHASH (change in F-UACR) as the objective variable and SHASH (baseline F-UACR) and baseline triglyceride as the explanatory variables. The regression coefficient estimates (p values) of this model showed significant values of −0.501 (p < 0.001) and −0.001 (p = 0.027) for SHASH (baseline F-UACR) and baseline triglyceride, respectively (shown in online suppl. Table 2). As a further supplementary analysis, we investigated the effect of change in blood pressure on change in UACR and found no statistically significant correlation between the change in systolic blood pressure and SHASH-transformed change in UACR (p = 0.126). Whereas, baseline UACR, eGFR, and triglyceride were all confirmed as significant factors, as in the main analysis (shown in online suppl. Table 3). Furthermore, the participants were divided into a responder and nonresponder group based on the amount of UACR reduction, and baseline clinical parameters were compared between the two groups; however, no significant differences were observed between the two groups in any item (shown in online suppl. Table 4).
Table 3.
Bivariate relationships with SHASH (change in UACR)
| | | Variable (baseline data) | R-squared | Adjusted R-squared | RMSE | F value | Intercept estimate | Slope estimate | Slope p value |
|---|---|---|---|---|---|---|---|---|---|
| Continuous data | Covariate factor | SHASH (baseline UACR) | 0.223 | 0.208 | 0.896 | 15.204 | −0.026 | −0.480 | <0.001* |
| Age | 0.005 | −0.014 | 1.014 | 0.246 | 0.490 | −0.007 | 0.622 | ||
| BMI | 0.002 | −0.017 | 1.000 | 0.130 | 0.231 | −0.010 | 0.720 | ||
| Systolic BP | 0.065 | 0.047 | 0.983 | 3.662 | 2.170 | −0.017 | 0.061 | ||
| Diastolic BP | 0.051 | 0.033 | 0.990 | 2.838 | 1.279 | −0.018 | 0.098 | ||
| Creatinine | 0.041 | 0.022 | 0.995 | 2.241 | 0.812 | −0.956 | 0.140 | ||
| eGFR | 0.113 | 0.096 | 0.957 | 6.752 | −1.324 | 0.020 | 0.012* | ||
| Cystatin C | 0.027 | 0.008 | 0.995 | 1.400 | 0.683 | −0.692 | 0.242 | ||
| K | 0.002 | −0.018 | 1.020 | 0.090 | −0.498 | 0.117 | 0.765 | ||
| UA | 0.036 | 0.018 | 0.998 | 1.976 | 0.845 | −0.153 | 0.166 | ||
| Triglyceride | 0.176 | 0.160 | 0.930 | 11.084 | 0.348 | −0.002 | 0.002* | ||
| HDL cholesterol | 0.064 | 0.046 | 0.992 | 3.528 | −1.000 | 0.019 | 0.066 | ||
| LDL cholesterol | 0.005 | −0.015 | 1.022 | 0.241 | −0.285 | 0.003 | 0.626 | ||
| HbA1c | 0.001 | −0.018 | 1.016 | 0.065 | −0.320 | 0.042 | 0.800 | ||
| Aldosterone | 0.003 | −0.018 | 1.035 | 0.126 | 0.164 | −0.001 | 0.724 | ||
| Binary data | Sex | 0.010 | −0.008 | 1.011 | 0.552 | N/A | N/A | 0.461 | |
| Use of SGLT-2 inhibitor | 0.001 | −0.018 | 1.016 | 0.035 | N/A | N/A | 0.853 |
RMSE, root mean square error; SHASH, sinh-arcsinh normal distribution transformation; UACR, urine albumin-creatinine ratio; BMI, body mass index; BP, blood pressure; eGFR, estimated glomerular filtration rate; K, serum potassium; UA, uric acid; HDL, high-density lipoprotein; LDL, low-density lipoprotein; HbA1c, glycated hemoglobin; SGLT-2 inhibitor, sodium-glucose cotransporter 2 inhibitor.
*p < 0.05.
Table 4.
Multiple regression analysis of SHASH (change in UACR)
| Variable | Estimate (β) | Standard error | t value | p value | Standardized β | VIF | Power |
|---|---|---|---|---|---|---|---|
| Intercept | −1.015 | 0.433 | −2.340 | 0.023 | 0.000 | | |
| Baseline eGFR | 0.021 | 0.006 | 3.180 | 0.003* | 0.350 | 1.140 | 0.876 |
| Baseline TG | −0.002 | 0.001 | −2.480 | 0.017* | −0.426 | 2.768 | 0.683 |
| SHASH [baseline UACR] | −0.387 | 0.118 | −3.270 | 0.002* | −0.361 | 1.147 | 0.894 |
| (eGFR_0w – 66.0269) × (SHASH [baseline UACR] − 0.01977) | 0.010 | 0.007 | 1.450 | 0.153 | 0.160 | 1.137 | |
| (TG_0w – 204.352) × (SHASH [baseline UACR] − 0.01977) | 0.000 | 0.001 | 0.470 | 0.642 | 0.081 | 2.787 | |
Change in UACR, the value obtained by subtracting the baseline UACR from the UACR at 24 weeks after spironolactone administration; VIF, variance inflation factor; SHASH, sinh-arcsinh normal distribution transformation; UACR, urine albumin-creatinine ratio; eGFR, estimated glomerular filtration rate; TG, triglyceride.
*p < 0.05.
Fig. 1.
Relationship between baseline UACR, eGFR, and TG levels and change in UACR.
In the secondary outcome analysis for safety, the change in serum potassium levels was negatively correlated with baseline serum potassium levels, and the results of the univariate analysis were significant (r = −0.45, R2 = 0.201, p < 0.001). In the regression model, in which the change in serum potassium levels was the objective variable and baseline eGFR and covariate baseline potassium levels were added as interaction terms, the only significant factor was the baseline potassium level. The regression coefficient estimate was −0.401 (p = 0.001), which means that for every 1 mEq/L increase in baseline serum potassium levels, potassium levels at 24 weeks decreased by 0.401 mEq/L (shown in Table 5) (shown in Fig. 2).
Table 5.
Multiple regression analysis of change in blood potassium level following spironolactone administration
| Variable | Estimate (β) | Standard error | t value | p value | Standardized β | VIF | Power |
|---|---|---|---|---|---|---|---|
| Intercept | 2.078 | 0.514 | 4.040 | 0.000 | 0.000 | | |
| Baseline eGFR | −0.003 | 0.002 | −1.350 | 0.182 | −0.177 | 1.063 | |
| Baseline K | −0.401 | 0.109 | −3.670 | 0.001* | −0.488 | 1.105 | 0.949 |
| (baseline eGFR – 66.430) × (baseline K – 4.192) | 0.003 | 0.007 | 0.380 | 0.708 | 0.050 | 1.101 | |
VIF, variance inflation factor; eGFR, estimated glomerular filtration rate; K, serum potassium level.
*p < 0.05.
Fig. 2.
Relationship between baseline potassium levels and eGFR and change in serum potassium levels.
Because baseline eGFR and baseline serum potassium levels have been reported to be factors that affect the decrease in eGFR due to MRA administration [13, 14], and from a clinical perspective, we performed a multivariate analysis with change in eGFR as the objective variable and baseline eGFR, baseline potassium, and baseline aldosterone as the explanatory variables. The only significant factor was baseline eGFR. The estimated regression coefficient was −0.118 (p = 0.032), indicating that for every 1 mL/min/1.73 m2 increase in baseline eGFR, the eGFR at 24 weeks decreases by 0.118 mL/min/1.73 m2 (shown in Table 6) (shown in Fig. 3). As normality was not denied for the change in serum potassium levels, change in eGFR, baseline serum potassium levels, baseline eGFR, baseline aldosterone data, no statistical transformation was performed.
Table 6.
Multiple regression analysis of change in eGFR following spironolactone administration
| Variable | Estimate (β) | Standard error | t value | p value | Standardized β | VIF | Power |
|---|---|---|---|---|---|---|---|
| Intercept | 13.78 | 13.21 | 1.04 | 0.302 | 0 | 0 | |
| Baseline eGFR | −0.118 | 0.053 | −2.21 | 0.032* | −0.318 | 1.057 | 0.468 |
| Baseline K | −2.422 | 2.910 | −0.83 | 0.410 | −0.122 | 1.097 | |
| Baseline aldosterone | −0.005 | 0.025 | −0.19 | 0.852 | −0.027 | 1.045 | |
VIF, variance inflation factor; eGFR, estimated glomerular filtration rate; K, serum potassium level.
*p < 0.05.
Fig. 3.
Relationship between baseline potassium levels, eGFR, and aldosterone and change in eGFR.
Discussion
This post hoc analysis focused on the reduction in UACR and safety in a low-dose spironolactone trial for patients with diabetes-induced CKD. We found that the reduction in UACR by low-dose spironolactone administration was greater in patients with higher baseline UACRs, higher baseline triglyceride levels, and lower baseline eGFRs. However, in this study, the change in UACR was shown as a quantity rather than a rate; therefore, it cannot be simply interpreted that the effect of UACR reduction was greater at a higher baseline UACR. This analysis showed that low-dose spironolactone reduced UACR regardless of the baseline UACR, and its effect was sufficient even in patients with impaired renal function (low eGFR or high UACR). This suggests that it is not too late to start treatment with low-dose spironolactone, even in patients with relatively advanced CKD and type 2 diabetes.
UACR
Few studies have investigated the clinical characteristics of patients who have benefited from MRAs. Post hoc analyses of large clinical trials of finerenone have shown that patients with severely elevated albuminuria have better renal composite outcomes and UACR improvement than those with moderately elevated albuminuria. This suggests that patients with severely elevated urine albumin levels are more likely to benefit from finerenone [15, 16]. Additionally, spironolactone at its usual dose (25 mg/day) significantly reduces proteinuria in patients with CKD who have a UACR greater than 1,000 mg/gCr, an eGFR of 60 mL/min/1.73 m2 or higher, and low baseline aldosterone levels (<80 pg/mL) [17]. Baseline aldosterone levels have also been reported to predict the degree of proteinuria reduction with 25 mg/day spironolactone [18]. Our analysis also showed that the greater the UACR at baseline, the greater the decrease in the UACR. As mentioned previously, the decrease in the UACR was not shown as a percentage; therefore, it may be inaccurate to state that patients with a higher baseline UACR are more likely to experience a decrease in their UACR. However, our study suggests that low-dose spironolactone may reduce the UACR regardless of the baseline UACR. In this analysis, baseline aldosterone levels were not associated with a reduction in the UACR. Although various reports have been published regarding the effects of “aldosterone escape” on UACR reduction, no consensus has been reached. In this study, blood sampling conditions, such as rest or food restriction, were not specified for aldosterone measurements, which may have contributed to the inaccuracy of the aldosterone levels. The eGFR results differed from those of previous reports; however, this may be due to differences in spironolactone doses and patient backgrounds. The substantial method of assessing urinary protein, filtration-adjusted proteinuria, which divides absolute proteinuria by renal function (Uprot/eGFR × 100), was proposed by Provenzano et al. [19], and is a useful index that may evaluate renal prognosis more precisely than Uprot. In this study, to examine the effect of renal function on UACR, we also evaluated F-UACR in the same way as UACR. The results also showed that the greater the F-UACR at baseline, the greater the decrease in F-UACR. We suggest that the consistent findings for both UACR and F-UACR further support the robustness of our study’s main conclusion.
In online supplementary Table 4, no significant difference was observed between the responders (≥30% decrease in UACR from baseline) and nonresponders (<30% decrease) in all baseline clinical parameters. This result may be due to the fact that converting continuous variables to binary values reduces the amount of information in the data and leads to a drop in statistical power. Royston et al. [20] have made the important point that “dichotomizing continuous predictors in regression analysis should be avoided” [20], and Altman and Royston [21] have similarly discussed the costs and inefficiencies of dichotomization in detail.
The baseline UACR, triglyceride levels, and eGFR influenced the UACR reduction in our analysis. They may be due to several mechanisms. MR hyperactivation induces vascular injury, tubulointerstitial inflammation, fibrosis, and glomerular injury [22]. MR expression is increased and hyperactivated in patients with diabetes [23, 24] and elevated urinary protein levels [25]. Moreover, the response to MRAs may be enhanced in individuals with a greater degree of renal interstitial fibrosis [26]. These reports are consistent with the results of our study, suggesting that low-dose spironolactone is sufficient for improving the UACR in patients with a high baseline UACR and low eGFR. These findings provide a rationale for actively considering the introduction of MRAs in patients with advanced diabetes-induced CKD without considering it “too late.” Triglycerides have been reported to be associated with renal function decline through foam cell formation by the incorporation of very low-density lipoprotein into mesangial cells [27] and increased coagulation and intrarenal microthrombi by the induction of plasminogen activator inhibitor-1 by very low-density lipoprotein [28]. Triglycerides are known to vary significantly with diet; however, postprandial blood samples were included in this study. Additionally, the power to detect triglycerides in the multiple regression analysis was low (68.3%) compared with other factors. Therefore, it is difficult to make a clear suggestion regarding the influence of triglycerides from the results of this study and further investigation is needed. Further, the possibility that changes in blood pressure affect UACR is an important consideration; however, our supplementary analyses did not reveal a causal relationship. This result supports the view that the main factor in the urinary albumin-lowering effect of spironolactone may be a mechanism other than blood pressure reduction, as mentioned above.
Safety
Regarding spironolactone administration, previous studies have raised concerns regarding the risk of hyperkalemia and decreased eGFR during the initial phase of treatment. No discontinuations due to hyperkalemia (potassium ≥5.5 mEq/L) occurred with the use of low-dose spironolactone among the participants of this study [12]. In this post hoc analysis, the increase in serum potassium levels due to low-dose spironolactone administration was negatively correlated with baseline serum potassium levels. Similar negative correlations were found in the placebo group, which may have been influenced by a feedback mechanism due to elevated potassium levels. In other words, a feedback mechanism similar to that of the placebo may still be present with low-dose spironolactone administration, suggesting that low-dose spironolactone is safe for use. Moreover, there was no apparent correlation between the change in serum potassium levels and baseline eGFR after low-dose spironolactone administration, indicating that a lower eGFR does not necessarily increase serum potassium levels when using low-dose spironolactone. The eGFR temporarily decreased immediately after MRA administration [16, 29]. In this study, a significant decrease in eGFR was also observed in the low-dose spironolactone group [12]. This analysis showed that the decline in eGFR with low-dose spironolactone administration is minor in patients with a low baseline eGFR. Although this result should be interpreted with caution due to low statistical power, it suggests that low-dose spironolactone may be used safely even in patients with a low baseline eGFRs. These results suggest that low-dose spironolactone may be a safe treatment option, even for patients with relatively low eGFR or high serum potassium levels.
Study Limitations
This study has a few limitations. First, this was a post hoc analysis and was mostly exploratory and hypothesis-generating. Second, this study was limited to Japanese patients; hence, our results may not be generalizable to other racial groups. Third, it is unclear whether the present results apply to regular-dose spironolactone or to other MRAs. Fourth, it remains controversial whether UACR can be a surrogate marker for renal endpoints; therefore, this study does not precisely demonstrate that low-dose spironolactone improves long-term renal endpoints. Fifth, the number of participants in this study was relatively small (N = 55); therefore, our statistical analysis may have been underpowered. In particular, the sample size may have been insufficient to detect small effect sizes. Hence, future large-scale multicenter prospective studies may be required to validate our findings.
Conclusion
This analysis suggests that low-dose spironolactone is effective in improving the UACR, even in patients with high baseline UACR and low eGFR; moreover, the effect is more pronounced among patients with a relatively low eGFR. This study also showed a low risk of side effects such as hyperkalemia and decreased renal function with low-dose spironolactone administration. This suggests that low-dose spironolactone may be a safe and effective treatment option for patients with advanced diabetes-induced CKD. Furthermore, as the least expensive MRAs, low-dose spironolactone has great potential for widespread application because it is less affected by CKD progression and is affordable.
Acknowledgment
We would like to thank Editage (www.editage.com) for the English language editing.
Statement of Ethics
The clinical study protocol was conducted in accordance with the Declaration of Helsinki and was reviewed and approved by the Ethics Committee of the Shinshu University Graduate School of Medicine (approval No. 3378). This study was registered with the Japanese University Hospital Medical Information Network Clinical Trials Registry (UMIN-CTR: UMIN 000022999). The participants provided written informed consent to participate in this study.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
This study was not supported by any sponsor or funder.
Author Contributions
Writing – original draft preparation: S.T. and A.O.; writing – review and editing: M. Komatsu; statistical analyses: M. Koinuma and S.T.; collecting data: T.T., T.M., and D.H. All authors have read the manuscript and approved the manuscript for publication.
Funding Statement
This study was not supported by any sponsor or funder.
Data Availability Statement
The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon reasonable request.
Supplementary Material.
References
- 1. Coresh J, Astor BC, Greene T, Eknoyan G, Levey AS. Prevalence of chronic kidney disease and decreased kidney function in the adult US population: third National Health and Nutrition Examination Survey. Am J Kidney Dis. 2003;41(1):1–12. [DOI] [PubMed] [Google Scholar]
- 2. The Japanese Society for Dialysis Therapy . Current status of chronic maintenance dialysis therapy in Japan. 2022. Available from: https://www.jstage.jst.go.jp/article/jsdt/56/12/56_473/_article/-char/en (accessed December 28, 2023).
- 3. Heerspink HJ, Kröpelin TF, Hoekman J, de Zeeuw D; Reducing Albuminuria as Surrogate Endpoint REASSURE Consortium . Drug-induced reduction in albuminuria is associated with subsequent renoprotection: a meta-analysis. J Am Soc Nephrol. 2015;26(8):2055–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Bakris GL, Agarwal R, Anker SD, Pitt B, Ruilope LM, Rossing P, et al. Effect of finerenone on chronic kidney disease outcomes in type 2 diabetes. N Engl J Med. 2020;383(23):2219–29. [DOI] [PubMed] [Google Scholar]
- 5. Agarwal R, Filippatos G, Pitt B, Anker SD, Rossing P, Joseph A, et al. Cardiovascular and kidney outcomes with finerenone in patients with type 2 diabetes and chronic kidney disease: the FIDELITY pooled analysis. Eur Heart J. 2022;43(6):474–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Agarwal R, Tu W, Farjat AE, Farag YM, Toto R, Kaul S, et al. Impact of finerenone-induced albuminuria reduction on chronic kidney disease outcomes in type 2 diabetes: a mediation analysis. Ann Intern Med. 2023;176(12):1606–16. [DOI] [PubMed] [Google Scholar]
- 7. Sun LJ, Sun YN, Shan JP, Jiang GR. Effects of mineralocorticoid receptor antagonists on the progression of diabetic nephropathy. J Diabetes Investig. 2017;8(4):609–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Hou J, Xiong W, Cao L, Wen X, Li A. Spironolactone add-on for preventing or slowing the progression of diabetic nephropathy: a meta-analysis. Clin Ther. 2015;37(9):2086–103.e10. [DOI] [PubMed] [Google Scholar]
- 9. Oxlund CS, Henriksen JE, Tarnow L, Schousboe K, Gram J, Jacobsen IA. Low dose spironolactone reduces blood pressure in patients with resistant hypertension and type 2 diabetes mellitus: a double blind randomized clinical trial. J Hypertens. 2013;31(10):2094–102. [DOI] [PubMed] [Google Scholar]
- 10. Chrysostomou A, Becker G. Spironolactone in addition to ACE inhibition to reduce proteinuria in patients with chronic renal disease. N Engl J Med. 2001;345(12):925–6. [DOI] [PubMed] [Google Scholar]
- 11. Sato A, Hayashi K, Naruse M, Saruta T. Effectiveness of aldosterone blockade in patients with diabetic nephropathy. Hypertension. 2003;41(1):64–8. [DOI] [PubMed] [Google Scholar]
- 12. Oiwa A, Hiwatashi D, Takeda T, Miyamoto T, Kawata I, Koinuma M, et al. Efficacy and safety of low-dose spironolactone for chronic kidney disease in type 2 diabetes. J Clin Endocrinol Metab. 2023;108(9):2203–10. [DOI] [PubMed] [Google Scholar]
- 13. Rossignol P, Dobre D, McMurray JJ, Swedberg K, Krum H, van Veldhuisen DJ, et al. Incidence, determinants, and prognostic significance of hyperkalemia and worsening renal function in patients with heart failure receiving the mineralocorticoid receptor antagonist eplerenone or placebo in addition to optimal medical therapy: results from the Eplerenone in mild patients Hospitalization and Survival Study in Heart Failure (EMPHASIS-HF). Circ Heart Fail. 2014;7(1):51–8. [DOI] [PubMed] [Google Scholar]
- 14. Beldhuis IE, Myhre PL, Bristow M, Claggett B, Damman K, Fang JC, et al. Spironolactone in patients with heart failure, preserved ejection fraction, and worsening renal function. J Am Coll Cardiol. 2021;77(9):1211–21. [DOI] [PubMed] [Google Scholar]
- 15. Ruilope LM, Pitt B, Anker SD, Rossing P, Kovesdy CP, Pecoits-Filho R, et al. Kidney outcomes with finerenone: an analysis from the FIGARO-DKD study. Nephrol Dial Transplant. 2023;38(2):372–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Bakris GL, Ruilope LM, Anker SD, Filippatos G, Pitt B, Rossing P, et al. A prespecified exploratory analysis from FIDELITY examined finerenone use and kidney outcomes in patients with chronic kidney disease and type 2 diabetes. Kidney Int. 2023;103(1):196–206. [DOI] [PubMed] [Google Scholar]
- 17. Kim HY, Bae EH, Ma SK, Kim SW. Effects of spironolactone in combination with angiotensin-converting enzyme inhibitors or angiotensin receptor blockers in patients with proteinuria. Kidney Blood Press Res. 2014;39(6):573–80. [DOI] [PubMed] [Google Scholar]
- 18. Bianchi S, Bigazzi R, Campese VM. Long-term effects of spironolactone on proteinuria and kidney function in patients with chronic kidney disease. Kidney Int. 2006;70(12):2116–23. [DOI] [PubMed] [Google Scholar]
- 19. Provenzano M, Chiodini P, Minutolo R, Zoccali C, Bellizzi V, Conte G, et al. Reclassification of chronic kidney disease patients for end-stage renal disease risk by proteinuria indexed to estimated glomerular filtration rate: multicentre prospective study in nephrology clinics. Nephrol Dial Transplant. 2020;35(1):138–47. [DOI] [PubMed] [Google Scholar]
- 20. Royston P, Altman DG, Sauerbrei W. Dichotomizing continuous predictors in multiple regression: a bad idea. Stat Med. 2006;25(1):127–41. [DOI] [PubMed] [Google Scholar]
- 21. Altman DG, Royston P. The cost of dichotomising continuous variables. BMJ. 2006;332(7549):1080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Brown NJ. Aldosterone and vascular inflammation. Hypertension. 2008;51(2):161–7. [DOI] [PubMed] [Google Scholar]
- 23. Taira M, Toba H, Murakami M, Iga I, Serizawa R, Murata S, et al. Spironolactone exhibits direct renoprotective effects and inhibits renal renin-angiotensin-aldosterone system in diabetic rats. Eur J Pharmacol. 2008;589(1–3):264–71. [DOI] [PubMed] [Google Scholar]
- 24. Guo C, Martinez-Vasquez D, Mendez GP, Toniolo MF, Yao TM, Oestreicher EM, et al. Mineralocorticoid receptor antagonist reduces renal injury in rodent models of types 1 and 2 diabetes mellitus. Endocrinology. 2006;147(11):5363–73. [DOI] [PubMed] [Google Scholar]
- 25. Quinkler M, Zehnder D, Eardley KS, Lepenies J, Howie AJ, Hughes SV, et al. Increased expression of mineralocorticoid effector mechanisms in kidney biopsies of patients with heavy proteinuria. Circulation. 2005;112(10):1435–43. [DOI] [PubMed] [Google Scholar]
- 26. Lindhardt M, Persson F, Oxlund C, Jacobsen IA, Zürbig P, Mischak H, et al. Predicting albuminuria response to spironolactone treatment with urinary proteomics in patients with type 2 diabetes and hypertension. Nephrol Dial Transplant. 2018;33(2):296–303. [DOI] [PubMed] [Google Scholar]
- 27. Anami Y, Kobori S, Sakai M, Kasho M, Nishikawa T, Yano T, et al. Human beta-migrating very low density lipoprotein induces foam cell formation in human mesangial cells. Atherosclerosis. 1997;135(2):225–34. [DOI] [PubMed] [Google Scholar]
- 28. Olufadi R, Byrne CD. Effects of VLDL and remnant particles on platelets. Pathophysiol Haemost Thromb. 2006;35(3–4):281–91. [DOI] [PubMed] [Google Scholar]
- 29. Esteghamati A, Noshad S, Jarrah S, Mousavizadeh M, Khoee SH, Nakhjavani M. Long-term effects of addition of mineralocorticoid receptor antagonist to angiotensin II receptor blocker in patients with diabetic nephropathy: a randomized clinical trial. Nephrol Dial Transplant. 2013;28(11):2823–33. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon reasonable request.
