Predictors of Decreasing Glomerular Filtration Rate and Prevalence of Chronic Kidney Disease After Treatment of Primary Aldosteronism: Renal Outcome of 213 Cases

Context:

In primary aldosteronism (PA), glomerular hyperfiltration due to excessive aldosterone is considered to underestimate actual renal damage.

Objective:

Our objectives were to determine the prevalence of chronic kidney disease (CKD) in PA and identify the predictors of decreasing estimated glomerular filtration rate (eGFR) after treatment.

Design and Setting:

This was a 12-month prospective study of patients with PA treated at Tohoku University Hospital.

Patients and Interventions:

All patients were treated according to the results of adrenal venous sampling; 102 patients with aldosterone-producing adenoma underwent adrenalectomy, and 111 with bilateral hyperaldosteronism were treated with mineralocorticoid receptor antagonists.

Main Outcome Measures:

Electrolytes, blood pressure, and indicators of renal function were determined at 1 and 12 months after intervention.

Results:

Blood pressure, urinary albumin excretion (UAE), and eGFR, which significantly decreased at 1 month after treatment of PA, did not further decrease at 12 months. Prevalence of CKD, which was 15.7% in aldosterone-producing adenoma and 8.1% in bilateral hyperaldosteronism at the first visit, increased to 37.1% and 28.3%, respectively, at the end of study (P < .0001). Multivariate regression analysis revealed that higher UAE and lower serum potassium levels were found to be independent predictors of decreasing eGFR after intervention.

Conclusions:

This large cohort study shows that the prevalence of CKD in PA was increased after treatment and that higher UAE and lower serum potassium levels at the first visit were predictors of decreasing eGFR after treatment of PA. To prevent a large decrease of eGFR after intervention, PA patients should be diagnosed before evolution to severe albuminuria and hypokalemia.

Patients with primary aldosteronism (PA) are at a higher risk for cardiovascular morbidity than patients with essential hypertension (1, 2), but the risk can be reduced in the long term after treatment (3). Recent studies demonstrated higher morbidity of renal damage in PA patients compared with primary hypertension (4–6). In the Primary Aldosteronism Prevalence in Italy (PAPY) study, 24-hour urinary albumin excretion (UAE) was significantly greater in patients with PA than in hypertensive controls (4). Sechi et al (5) demonstrated that increased UAE and relative glomerular hyperfiltration were rapidly reversed after removal of excessive aldosterone. They also reported that, in the patients with essential hypertension, blood pressure (BP) and UAE were significantly reduced by antihypertensive agents, but creatinine clearance was not. Moreover, their UAE reduction was significantly smaller than that in PA patients (5). These results suggested that hyperaldosteronism could cause greater glomerular hyperfiltration and albuminuria as compared with essential hypertension, which was also supported by another report (6). Reincke et al (7) showed that plasma aldosterone concentration and serum potassium were independent risk factors of low glomerular filtration rate (GFR) in untreated PA patients and that specific interventions were associated with a further decrease in GFR.

Thus, a clinically critical problem is the decrement of eGFR after treatment of PA. Actually, we sometimes deal with patients with aldosterone-producing adenoma (APA) who develop chronic kidney disease (CKD) after adrenalectomy because of a large decrease of estimated GFR (eGFR). However, there have been few reports investigating the predictive factors of renal function after treatment of PA. Therefore, we conducted a large prospective cohort study including over 200 PA subjects. Our main goal was to determine the prevalence of CKD in PA patients after treatment and to identify predictors of decreasing eGFR after treatment for PA.

Patients and Methods

Study population and follow-up

The study group comprised patients diagnosed with PA who were treated at Tohoku University Hospital between April 2007 and December 2010. Written consent was obtained from the patients after giving them sufficient information about the study; the study protocol was approved by the ethics committee of Tohoku University School of Medicine and adhered to the principles of the Declaration of Helsinki and Title 45, U.S. Code of Federal Regulations, part 46, Protection of Human Subjects, revised November 13, 2001, effective December 13, 2001. PA was diagnosed using previously reported procedures (8). At the first visit, tests for plasma aldosterone concentration (PAC) (nanograms per deciliter), plasma renin activity (PRA) (nanograms per milliliter per hour), serum creatinine, serum sodium, serum potassium, urinary sodium, urinary potassium, and UAE were performed in the morning after a 10-minute rest in the sitting position. PAC and PRA were measured by RIA; the former was measured using the SPAC-S aldosterone kit (TFB Inc), and the latter was measured using the renin Riabead kit for PRA (Dainabot Co Ltd). The diagnosis of PA was confirmed when the aldosterone to renin ratio (ARR) exceeded 20 after loading 50 mg captopril. Upon diagnosis of PA, antihypertensive agents were limited to calcium channel blockers and/or α₁-blockers according to clinical practice guidelines of The Endocrine Society (8). Patients who had other endocrine diseases were excluded from this study. We also excluded the patients with primary renal diseases, diabetes with evident proteinuria, renal artery stenosis, renal cell carcinoma, or other urinary findings such as hematuria and urinary tract infection. All patients were treated according to the results of adrenal venous sampling as previously reported (9, 10). The patients with unilateral disease diagnosed by adrenal venous sampling underwent laparoscopic adrenalectomy, whereas those with bilateral hyperaldosteronism (BHA) were treated with a pharmacological regimen that included mineralocorticoid receptor (MR) antagonists (spironolactone or eplerenone). In all patients who underwent adrenalectomy, APA was confirmed by histopathological and immunohistochemical analysis including immunohistochemistry of 3β-hydroxysteroid dehydrogenase and C17 (11). In particular, the absence of 3β-hydroxysteroid dehydrogenase in the hyperplastic zona glomerulosa in the nonneoplastic adrenal with APA ruled out the possibility of bilateral or idiopathic hyperaldosteronism (11).

Clinical assessments and laboratory tests were done at 1 and 12 months after a specific treatment for PA. At each visit, BP was measured using the Omron Hem-907 (Omron Healthcare Co Ltd) to verify whether the patient had achieved the target BP value of 140/90 mm Hg. In both groups, no angiotensin converting enzyme inhibitors, angiotensin II receptor blockers, or diuretics not including MR antagonists were used during the first month after treatment of PA. During follow-up and until the end of study, all classes of antihypertensive agents were allowed to achieve the target BP recommended by the guidelines (12). The eGFR was calculated using the following equation established for the Japanese population by the Japanese Society of Nephrology: eGFR (mL/min/1.73 m²) = 194 × serum creatinine^−1.094 × age^−0.287 (× 0.739 in the case of females) (13). The UAE values are expressed as a ratio of urinary albumin concentration to urinary creatinine concentration (milligrams per gram creatinine). In this study, CKD was defined by an eGFR <60 mL/min/1.73 m². Albuminuria was defined as >30 mg/g creatinine, respectively.

Statistical analysis

The results are presented as mean ± SEM values. The UAE values were log-transformed for statistical analysis due to their skewed distribution. Univariate comparisons between APA and BHA were performed using unpaired t test, as appropriate. Comparisons between before and after treatment in each group were performed by paired t test. One-way ANOVA followed by Bonferroni's correction was used to compare normally distributed quantitative variables at each time point. Univariate correlations were evaluated as Pearson's correlation coefficients. Independent correlates of eGFR decline were investigated using multivariate linear regression analysis with the backward stepwise procedure. Significance was set at P < .05. StatFlex version 6.0 software (Artech Co Ltd) was used for all the statistical analyses.

Results

Clinical characteristics of the study population

A total of 213 patients with PA participated in this study. Their clinical features at the first visit are shown in Table 1. Among these, 102 patients had APA and 111 had BHA. There were significant differences in systolic BP (SBP) and diastolic BP (DBP) between the APA and BHA groups (SBP, 154 ± 2.3 vs 148 ± 2.0 mm Hg, P < .05; DBP, 94 ± 1.4 vs 89 ± 1.3 mm Hg, P < .01, respectively). Duration of hypertension and family history of hypertension were comparable. PAC, PRA, ARR, and serum sodium and potassium levels also significantly differed. Serum creatinine levels and eGFR were similar between the APA and BHA groups, although there were significant differences in age and number of male patients. The prevalence of CKD was 11.7% in all patients at the first visit, with no significant difference between the APA and BHA groups. The prevalence of albuminuria was 47.4% with a significant difference between groups (61.8% in APA group vs 34.2% in BHA group, P < .0001).

Table 1.

Clinical Characteristics of Patients With PA at the First Visit^a

Variables	Total (n = 213)	APA (n = 102)	BHA (n = 111)	P
Age, y	54 ± 0.8	51 ± 1.3	56 ± 1.0	<.01
Males/females, n	89/124	52/50	37/74	<.01
Duration of hypertension, y	10 ± 0.7	11 ± 1.0	8.3 ± 0.9	NS
Family history of hypertension, %	68.8	67.4	70.0	NS
Body mass index, kg/m2	25.0 ± 0.3	24.2 ± 0.4	25.6 ± 0.4	<.05
SBP, mm Hg	151 ± 1.5	154 ± 2.3	148 ± 2.0	<.05
DBP, mm Hg	91 ± 1.0	94 ± 1.4	89 ± 1.3	<.01
Serum creatinine, mg/dL	0.74 ± 0.02	0.77 ± 0.03	0.71 ± 0.02	NS
eGFR, mL/min/1.73 m2	80 ± 1	81 ± 2	79 ± 2	NS
CKD, n (%)	25 (11.7)	16 (15.7)	9 (8.1)	NS
UAE, mg/g creatinine	91 ± 19	80 ± 22	100 ± 29	NS
Albuminuria, n (%)	101 (47.4)	63 (61.8)	38 (34.2)	<.0001
Serum sodium, mmol/L	142.9 ± 0.2	143.7 ± 0.2	142.1 ± 0.2	<.0001
Serum potassium, mmol/L	3.8 ± 0.0	3.5 ± 0.1	4.1 ± 0.0	<.0001
Urinary sodium, mmol/g creatinine	158.8 ± 9.3	154.9 ± 13.5	162.5 ± 12.8	NS
Urinary potassium, mmol/g creatinine	60.8 ± 2.3	56.0 ± 3.0	65.4 ± 3.5	<.05
Plasma aldosterone, ng/dL	20.9 ± 1.1	26.2 ± 1.9	15.8 ± 0.7	<.0001
PRA, ng/mL/h	0.69 ± 0.1	0.37 ± 0.1	1.00 ± 0.2	<.001
ARR, ng/dL per ng/mL/h	97 ± 11	164 ± 21	35 ± 3	<.0001
HbA1c (NGSP), %	5.8 ± 0.1	5.7 ± 0.1	5.9 ± 0.1	NS
Diabetes mellitus, n (%)	33 (15.5)	14 (13.7)	19 (17.1)	NS

Abbreviations: HbA1c, hemoglobin A1c; NGSP, National Glycohemoglobin Standardization Program; NS, not significant.

^a

The data represent the mean ± SEM Unpaired t test was used for comparison between APA and BHA. CKD was defined as <60 ml/min/1.73 m² in eGFR.

Aldosterone and BP decreased after treatment of PA

The average duration of follow-up was 11.5 and 8.3 months in the APA and BHA groups, respectively. In the APA group, PAC significantly decreased and serum potassium was corrected to levels within the normal range by the first month after adrenalectomy (PAC, from 26.2 to 8.5 ng/dL, P < .05; serum potassium, from 3.5 to 4.5 mmol/L, P < .05) (Table 2). BP significantly decreased at 1 month after treatment, with mean values of 126/79 and 124/74 mm Hg in the APA and BHA groups, respectively. BP remained stable during the 12-month follow-up period in BHA (Table 2), but SBP was significantly decreased in APA (from 126 at 1 month to 120 at 12 months, P < .05).

Table 2.

Changes in Parameters Before and After Treatment for PA^a

	APA			BHA
	At First Visit	1 Month After Treatment	12 Months After Treatment	At First Visit	1 Month After Treatment	12 Months After Treatment
SBP, mm Hg	154 ± 2.3	126 ± 1.7b,d	120 ± 1.5b,d,e	148 ± 2.0	124 ± 1.5b,d	125 ± 1.7b,d,e
DBP, mm Hg	94 ± 1.4	79 ± 1.2b,c,f	74 ± 1.1b,c	89 ± 1.3	74 ± 1.6b,d,f	74 ± 1.5b,d
PAC, ng/dL	26.2 ± 1.9	8.5 ± 0.4b,d	9.4 ± 0.5b,d	15.8 ± 0.7
PRA, ng/mL/h	0.37 ± 0.1	1.72 ± 0.4c	3.73 ± 0.7b,c	1.00 ± 0.2
Serum potassium, mmol/L	3.5 ± 0.1	4.5 ± 0.0b,g	4.5 ± 0.1b,d,f	4.1 ± 0.0	4.2 ± 0.0b,g	4.3 ± 0.0b,d,f
eGFR, mL/min/1.73 m2	81 ± 2	68 ± 2b,e	65 ± 2b,d	79 ± 2	73 ± 1b,e	69 ± 2b,d
UAE, mg/g creatinine	80 ± 22	21 ± 4b	23 ± 5b,d	100 ± 29	21 ± 5b	25 ± 7b,d

^a

One-way ANOVA followed by Bonferroni correction was performed for each time point. Unpaired t test was used for comparison between APA and BHA groups at each time point after treatment. Comparisons at the first visit between groups are shown in Table 1.

^b

P < .05 vs first visit in each group.

^c

P < .05, 1 vs 12 months after treatment in each group.

^d

No significance, 1 vs 12 months after treatment in each group.

^e

P < .05, APA vs BHA groups at each time point after treatment.

^f

P < .005, APA vs BHA groups at each time point after treatment.

^g

P < .0005, APA vs BHA groups at each time point after treatment.

Use of antihypertensive agents

At the first visit, the number of patients who took no antihypertensive medication was 18 (17.6%) and 30 (27%) in APA and BHA, respectively. The number of patients who took 1 agent was 22 (21.6%) and 30 (27%), respectively. The number of patients on multi-agent therapy was 62 (60.8%) and 51 (46%), respectively. There was significant difference only in the number of patients on multi-agent therapy between the APA and BHA group (P < .05). At the end of the study, the number of patients with no agents was 20 (19.6%) in APA. APA patients on mono- and multi-therapy were 43 (42.2%) and 39 (38.2%), respectively. In BHA, all 111 patients in this group took MR antagonists during the follow-up. However, 15 patients discontinued their spironolactone prescription due to adverse effects and changed to eplerenone. At the end of the study, BHA patients on mono- and multi-therapy were 12 (10.8%) and 99 (89.2%), respectively. There were significant differences in the number of patients on mono- (P < .0001) and multi-agent therapy (P < .0001) between the APA and BHA group (P < .05). There was no difference in the therapeutic classes at the end of the study between APA and BHA groups except for MR antagonists (angiotensin converting enzyme inhibitor, 1% vs 5%; angiotensin II receptor blocker, 38% vs 44%; diuretics, 6% vs 10% in APA vs BHA, respectively). During the first month after specific treatment, angiotensin converting enzyme inhibitors, angiotensin II receptor blockers, and diuretics were not used in either group.

Estimated GFR and prevalence of CKD after treatment of PA

At 1 month after treatment, eGFR immediately decreased in both groups (APA, from 81 to 68 mL/min/1.73 m², P < .05; BHA, from 79 to 73 mL/min/1.73 m², P < .05), but it showed no further significant decreases at 12 months after treatment (Table 2). The initial decrement was significantly different (APA, −13 mL/min/1.73 m²; BHA, −6 mL/min/1.73 m², P < .005). The prevalence of CKD (defined by eGFR <60 mL/min/1.73 m²), which was 15.7% in the APA group and 8.1% in the BHA group at the first visit, significantly increased to 37.1% and 28.3%, respectively, at 12 months after treatment (Figure 1).

Figure 1.

Prevalence of CKD. Paired t test was used for comparison between before and after treatment in each group. CKD was defined as an eGFR of <60 mL/min/1.73 m². Compared with the first visit, the prevalence of CKD increased significantly after treatment in APA and BHA patients.

UAE and prevalence of albuminuria after treatment of PA

UAE showed a significant decrease at 1 month after intervention in both groups (APA, from 80 to 21 mg/g creatinine, P < .05; BHA, from 100 to 21 mg/g creatinine, P < .05) (Table 2). In both groups, UAE at 12 months after treatment was not significantly changed compared with that at the first month. The prevalence of albuminuria significantly decreased at the end of the study in both the groups (from 61.8% to 14.7% in APA, P < .0001; and from 34.2% to 16.2% in BHA, P < .0001).

The predictors of treatment-induced eGFR decrease

In univariate analysis, albuminuria, PAC, PRA, SBP, body mass index, and serum potassium correlated with the decrease in eGFR. Backward stepwise regression analysis showed that UAE (log₁₀ UAE), serum potassium, and PRA at the first visit were significant predictors of the decrease in eGFR at 1 month after intervention (Table 3).

Table 3.

Potential Predictors of the Decrement of eGFR at 1 Month After Treatment^a

Variables	P
UAE (log10)	.0029
Serum potassium	.0092
PRA	.0144

^a

Backward stepwise regression analysis was performed. Adjusted R² = .178, F = 12.13, and P < .0001.

Discussion

In the present study, we found similar BP and renal alterations throughout the follow-up of 12 months after treatment to those of other previous studies (5–7). Based on these findings, we provided new information about the prevalence of CKD, which increased at 12 months after treatment. In addition, we identified the predictors of decreasing eGFR after treatment of PA.

Prevalence of CKD and albuminuria

In this study, we first observed that the prevalence of CKD was lower before treating PA than after treatment. However, it significantly increased approximately 20% after treatment in both APA and BHA groups because of decreasing eGFR. This means that renal damage was initially masked in 20% of our patients and that glomerular hyperfiltration can lead to underestimation of CKD at baseline in PA patients. In 1977, Danforth et al (14) demonstrated that the histological renal changes in APA patients were more severe than expected from the preoperative hypertensive and renal evaluation. This phenomenon might be explained as glomerular hyperfiltration even in patients with a damaged kidney.

Second, in the present study, both UAE and eGFR were significantly decreased after treatment of PA. Sechi et al (5) demonstrated the same alteration in PA patients with a more accurate method in which urine is collected over a 24-hour period. Thus, albuminuria might also indicate hyperfiltration. Moreover, it is known to be a significant risk factor not only for cardiovascular events but also for end-stage renal disease (15). In this respect, from the results of the present study, we emphasize that renal outcome was not deteriorated but rather normalized or ameliorated instead of increasing in CKD patients after treatment of PA.

Third, in the present study, there were 11.7% of PA patients who already had CKD before treatment of PA. In the German Conn registry of a multicenter study, Born-Frontsberg et al (2) reported the complications associated with chronic renal failure in untreated PA patients. Experimental studies previously demonstrated that high levels of aldosterone lead to renal damage (16, 17). Furthermore, in our study, albuminuria persisted in some patients even after the treatment. This suggests that in PA, renal impairment starts with reversible glomerular hyperfiltration associated with increasing losses of urinary albumin, followed by coexistence of irreversible structural and functional renal damage due to longstanding hypertension and persistent albuminuria. It might also arise due to other reasons such as metabolic syndrome, including obesity or diabetes without nephropathy in our patients, especially in BHA patients. Finally, we should repeatedly emphasize that the patients with urinary disorders due to other diseases or conditions such as diabetes nephropathy, urinary tract infection, and hematuria were strictly excluded in this study. Therefore, we could fairly analyze and consider our results in PA patients without such these potential causes of albuminuria.

Predictors of decreasing eGFR after treatment

Our data demonstrated and provided new information on the predictors of decreasing eGFR after treatment of PA. The strongest factor was the level of urinary albumin excretion at the first visit. This result indicated that higher urinary albumin might reflect glomerular hyperfiltration. It did not contradict previous reports that hyperaldosteronism could cause greater glomerular hyperfiltration and albuminuria in comparison with essential hypertension (5, 6). Serum potassium was also a predictor in our multivariate analysis; that is, the initial hypokalemia independently led to a lower decrease of eGFR after treatment. Ribstein et al (6) showed that baseline serum potassium correlated with the change in GFR by univariate analysis. Rossi et al (4) provided evidence in the PAPY study with 64 PA patients that low serum potassium at baseline predicts albuminuria. Reincke et al (7) provided evidence in the German Conn registry with more than 400 patients that hypokalemic PA patients tended to complicate renal impairment. From these results, we suggest that hypokalemia should be considered as not only a hallmark of hyperaldosteronism but also an indication of causality with hyperfiltration or renal damage. PRA was also one of the factors; that is, the state of suppressed renin activity due to hyperaldosteronism and hypervolemia can predict the decrement of eGFR after treatment.

Study limitations

There are some method- and participant-related limitations in this study. First, renal function might have been underestimated because of the method used to evaluate GFR. The equation we used in the present study was developed based on data from the Japanese population of patients with CKD (13). This formula is considered more suitable for the Japanese population than other formulas (eg, the Cockcroft-Gault formula). However, eGFR tends to be underestimated in patients with normal renal function, more so than in PA patients with glomerular hyperfiltration. Hence, eGFR of patients with glomerular hyperfiltration or without CKD might be lower than their actual GFR. Patients with diabetes generally tend to show hyporeninemic hypoaldosteronism, because of hypervolemia, especially due to the retention of sodium (18). In addition, it is reported that hyperglycemia itself can cause glomerular hyperfiltration as well as PA (19). This point might also be assumed to influence the renal outcomes. Moreover, treatment with antihypertensive agents might influence the renal outcomes at the end of the study. However, none of the patients took angiotensin converting enzyme inhibitors or angiotensin II receptor blockers during the first month after adrenalectomy in APA patients and after treatment of MR blockers in BHA. Therefore, the predictive parameters identified for decreasing eGFR at this time were considered as mainly depending on the effect of adrenalectomy or aldosterone antagonism.

Conclusions

In summary, treatment of PA resulted in increased prevalence of CKD due to decreasing eGFR. Higher UAE and lower serum potassium levels at the first visit were strong predictors of a decrement in eGFR. From our results, to prevent a large decrease of eGFR after treatment of PA, patients should be diagnosed as early as possible before presentation with evident albuminuria and hypokalemia.

Acknowledgments

We thank Ms Akane Sugawara and Ms Yasuko Tsukada for secretarial assistance.

Disclosure Summary: The authors have nothing to disclose.

Abbreviations

 
• APA

  aldosterone-producing adenoma

 
• ARR

  aldosterone to renin ratio

 
• BHA

  bilateral hyperaldosteronism

 
• BP

  blood pressure

 
• CKD

  chronic kidney disease

 
• DBP

  diastolic BP

 
• eGFR

  estimated GFR

 
• GFR

  glomerular filtration rate

 
• MR

  mineralocorticoid receptor

 
• PA

  primary aldosteronism

 
• PAC

  plasma aldosterone concentration

 
• PRA

  plasma renin activity

 
• SBP

  systolic BP

 
• UAE

  urinary albumin excretion.