Interplay of parathyroid hormone and aldosterone antagonist in prevention of heart failure hospitalizations in chronic kidney disease

Abstract

Background:

Aldosterone antagonists may mediate their effects on heart failure through parathyroid hormone (PTH) in chronic kidney disease (CKD) patients.

Methods:

Patients with CKD on spironolactone were selected and matched for age, gender, race, use of a vitamin D analogue, the number of antihypertensive medications, and CKD stage. PTH levels before and after the first prescription of spironolactone were measured. A thorough chart review was conducted to assess for heart failure hospitalizations. An adjusted Cox proportional model was used to calculate the hazard ratio (HR) for heart failure hospitalizations among cases versus controls.

Results:

There were a total of 950 (mean age 67±13 years, 40% men) patients with CKD. Of these, there were 48 hospitalizations for heart failure among the cases and 82 among the controls (HR 0.37; 95% confidence interval (CI) 0.19–0.74, p=0.005). We noted a more significant decrease in PTH levels among the cases when compared to the controls (p<0.0001). The adjusted hazard for heart failure hospitalization increased with higher PTH levels (p=0.002) and mediation analysis revealed change in PTH level as a significant mediator of heart failure hospitalization (p=0.04).

Conclusion:

Aldosterone antagonists may be helpful in preventing hospitalizations for heart failure exacerbation in CKD patients through a PTH-mediated effect.

Keywords

Aldosterone antagonist chronic kidney disease heart failure hospitalization

Introduction

Heart failure hospitalizations continue to be a major cause of morbidity and mortality among invididual patients whilst presenting a serious economic and public health burden in the larger framework.¹ Over the next two decades, the total direct medical costs of cardiovascular disease (CVD) are projected to triple to $818 billion.² Six year mortality rates for heart failure patients are 84% in men and 77% in females.^3,4 Given these overwhelming odds, it is imperative that preventative strategies continue to be investigated.

Elevated parathyroid hormone (PTH) levels have been associated with greater degrees of severity in heart failure.^5–8 Indeed PTH levels have been suggested as potential biomarkers of risk for increased heart failure hospitalization rates.⁹ Perhaps underlying this relationship is that fact that a direct link between the renin-angiotensin-aldosterone system and PTH has been demonstrated in humans. The implications of this are significant, and may present potential therapeutic avenues for the treatment of a number of PTH-related conditions.¹⁰

PTH levels are often measured in patients with chronic kidney disease (CKD) in order to evaluate and reduce the risk of renal osteodystrophy.^11–14 Such data may yield useful information towards predicting patients at risk for further heart failure admissions. Whether this association between aldosterone and PTH levels persists among patient with CKD remains unknown. Whether this relationship can be manipulated by means of an aldosterone antagonist in order to prevent heart failure hospitalizations is also unknown.

We theorized that use of aldosterone antagonists in patients with CKD results in reduced serum PTH levels and that both these factors are associated with reduced hospitalization in patients with heart failure.

Methods

This is a retrospective study of adult (≥18 years) ambulatory patients who were diagnosed with chronic kidney disease stage 2 or worse as defined by the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation.¹⁵ Patients with prior history of heart failure, on hemodialysis, with contraindications for spironolactone use, diagnosed malignancy, and acute renal failure were excluded from the study.

Cases

A case was defined as an individual with CKD on spironolactone for a minimum duration of 12 months. We reviewed spironolactone prescriptions of patients with CKD between 2000–2005 and identified patients who had filled prescriptions for at least 12 months after commencement of the medication. Individuals were selected randomly using a random number generator. There were two prerequisites for selection of cases:

A serum parathyroid hormone level was measured at least once in the year preceding the first prescription of spironolactone and that at least one parathyroid hormone level was obtained after 12 months of spironolactone use. We also stipulated a minimal interval of 18 months between measurement of the two assays.

All individuals had received a minimum of five years of follow-up care after the diagnosis of CKD had been made.

The control group (comparison group) was matched for age (for each year), gender, race, use of a vitamin D analogue, number of antihypertensive medications, and CKD stage from the Henry Ford Hospital database with at least two values of PTH 18 months apart and a follow up period of at least five years while off spironolactone. The study protocol was approved by the institutional review board.

Patient demographics, medications and other laboratory variables within three months of measurement of the second PTH level were collected. These were obtained by detailed chart review.

The primary outcome was heart failure exacerbation diagnosed by the presence of at least one symptom (shortness of breath, lower extremity swelling, orthopnea, or paroxysmal nocturnal dyspnea) and at least one sign (wet crackles, Jugular venous pressure (JVP)>5 cm, dependent edema) of heart failure along with chest radiograph findings or Brain Natriuretic Peptide (BNP)>100 pg/ml, that required hospital admission. These findings were checked for concordance with the diagnosis and treatment employed by the treating physician as evidenced by chart documentation. The secondary outcome was all-cause mortality which we identified through chart review and the social security death index.

Sample size calculation

In order to detect a mean difference of 20 pg/ml in the serum PTH level with a standard deviation (SD) of 120, a type I error of 0.05, and power of 80%, we calculated that the total required sample size would need to be 570 cases and controls.

Statistical analysis

Baseline characteristics were compared between the cases and the controls. Continuous variables were expressed in mean±SD. In the case of variables that did not follow a normal distribution, geometric means with 95% confidence intervals (CIs) were reported. Categorical variables were reported as percentages. The Student’s t-test was used to compare continuous variables and chi- square or Fischer’s exact test were used to compare categorical variables.

The hazard for heart failure hospitalizations amongst the cases was assessed by Cox conditional regression model adjusted for demographic, laboratory and historical covariates (age, gender, race, vitamin D levels, serum alkaline phosphatase, hypertension, diabetes, estimated glomerular filtration rate, and urine protein creatinine ratio)

Once spironolactone use was confirmed, we established an association of change in serum PTH levels with heart failure admissions in another Cox regression model adjusted for the same covariates given above.

After establishing this association, we employed regression models to evaluate the association between spironolactone use and changes in PTH levels and then to evaluate the association between PTH levels and heart failure hospitalizations. The regression coefficients of both of these models were used to calculate Sobel’s test statistic. If significant, this statistic would indicate that change in PTH levels is a mediator of association between spironolactone use and heart failure hospitalization. All p-values were two-sided and p<0.05 was considered significant. Analyses were performed by using PASW v 18.0 (SPSS Inc., Chicago, Illinois, USA).

Results

There was a total of 425 cases and 425 controls; patients were matched for age, gender, ethnicity, number of blood pressure medications, vitamin D analogues, and CKD stage. Significant differences between the cases and the controls were noted in the serum levels of alkaline phosphatase as well as levels of PTH (Table 1).

Table 1.

Baseline characteristics of cases and controls.

Characteristics	Spironolactone group (cases) n=425	Controls n=425	p-value
Age, years	67±13 (16%)	67±13 (16%)	>0.99
Men	169 (40%)	169 (40%)	>0.99
White	160 (38%)	160 (38%)	>0.99
African American	223 (52%)	223 (52%)
Others	42 (10%)	42 (10%)
25-hydroxy vitamin D level, ng/ml	24±12 (6%)	26±12 (6%)	0.36
Alkaline phosphatase, IU/l	110±84 (26%)	110±88 (26%)	0.99
<130 IU/l	328 (77%)	336 (79%)
≥130 IU/l	95 (22%)	88 (20%)
1 year post spironolactone use alkaline phosphatase level, IU/l	87±54 (20%)	104±81 (24%)	0.001
<130 IU/l	378 (77%)	349 (82%)
≥130 IU/l	47 (11%)	76 (18%)
Prior parathyroid hormone level, pg/ml	141±133 (33%)	139±125 (33%)	0.82
<65 pg/ml	136 (32%)	133 (31%)
65–100 pg/ml	80 (19%)	85 (20%)
>100 pg/ml	207 (49%)	207 (49%)
Post parathyroid hormone level, pg/ml	101±87 (24%)	139±126 (33%)	<0.0001
<65 pg/ml	171 (40%)	141 (33%)
65–100 pg/ml	100 (23%)	63 (15%)
>100 pg/ml	149 (35%)	221 (50%)
Hemoglobin A1c (%)	6.2±1.5	5.7±1.5	<0.0001
Urine protein/creatinine ratio (g/g)	1.037±0.212	1.021±0.208	0.16
<0.5 (g/g)	21 (5%)	19 (4%)
0.5–3.4 (g/g)	39 (9%)	41 (10%)
≥3.5 (g/g)	40 (9%)	40 (9%)
eGFR (ml/min/1.73m2)	41±24 (10%)	41±23 (10%)	0.27
<15 (non-dialysis)	32 (7%)	32 (7%)
15–29	89 (20%)	89 (20%)
30–59	124 (29%)	124 (29%)
60–90	180 (42%)	180 (42%)
Medications
Spironolactone dose, mg/day	64±44	–
Vitamin D analogues	178 (42%)	178 (42%)	>0.99
Phosphate scavengers	102 (24%)	112 (26%)	0.48
Calcium salts	356 (84%)	362 (85%)	0.64
Cinacalcet	34 (8%)	48 (11%)	0.13
Cardiovascular medications
ACE inhibitors	183 (43%)	181(42%)	0.94
ARBs	140 (33%)	143 (34%)	0.94
Beta blockers	284 (67%)	298 (70%)	0.34
Aspirin	207 (49%)	203 (48%)	0.84
Clopidogrel	38 (9%)	32 (7%)	0.53

ACE: angiotensin-converting enzyme; ARB: angiotensin receptor blocker; eGFR: estimated glomerular filtration rate; IU: international units.

Values of n (%) or mean±standard deviation provided.

Spironolactone and its effect on PTH

The mean PTH level for cases prior to enrollment was 141±126 pg/ml and was not found to be significantly different from that of the controls 140±125 pg/ml (p=0.59). After one year of spironolactone use, the cases were noted to have significantly lower levels of PTH when compared to the controls (101±87 vs 139±126, p<0.0001). The decrease in PTH levels was consistent across all stages of CKD amongst the cases, but not in the controls (Figure 1).

Figure 1.

Serum parathyroid hormone level change across various groups based on estimated glomerular filtration rates for cases and controls.

Spironolactone lowers rates of heart failure hospitalization

There were a total of 130 hospitalizations for heart failure with 48 instances among the cases and 82 in the controls. The hazard for heart failure hospitalization was lower in the cases than in the controls (HR 0.37; 95% CI 0.19–0.74, p=0.005). A Kaplan-Meier survival plot demonstrates this as a function of cases versus controls (log rank p-value<0.0001) (Figure 2).

Figure 2.

Time to heart failure hospitalization plot for cases and controls.

There were 369 deaths over a median follow up duration of 4.6 years (interquartile range (IQR) 3.3–5.8 years). Spironolactone use was not associated with reduction in all-cause mortality (HR 0.87; 95% CI 0.66–1.13, p=0.29). The Kaplan Meier plot demonstrates the survival probability of cases versus controls (log rank p-value=0.29) (Figure 3).

Figure 3.

Kaplan-Meier survival plot for all-cause mortality of cases and controls.

PTH mediates the association of spironolactone and decreased heart failure hospitalizations

In order to evaluate the role of PTH in the causal pathway of spironolactone and heart failure, we first evaluated the association of spironolactone use in a model adjusted for age, gender, race, vitamin D levels, serum alkaline phosphatase, hypertension, diabetes, estimated glomerular filtration rate, and urine protein creatinine ratio. After that, we added change in PTH levels in the model which led to significant change in effect size of association of spironolactone use and reduced risk of heart failure hospitalization (HR=0.47; 95% CI 0.23–0.96, p=0.04). There was a significant interaction of change in PTH and spironolactone use (p=0.002). To test significance of mediation, Sobel’s test was used and found to be significant (test statistic 1.97, p=0.04). In a separate model adjusted for demographic and clinical covariates (age, gender, race, vitamin D levels, serum alkaline phosphatase, hypertension, diabetes, estimated glomerular filtration rate, and urine protein creatinine ratio), every 50 pg/ml increment in the serum PTH level was associated with increased hazard for heart failure hospitalization, (HR 1.25; 95% CI 1.09 – 1.44, p = 0.002).

Discussion

In this study, we assessed the association between use of aldosterone antagonists, serum PTH levels and heart failure hospitalizations. We found that spironolactone use was associated with risk of heart failure hospitalization. Similarly, we found increased PTH levels were associated with risk of heart failure hospitalization. The association of spironolactone use with heart failure hospitalization was influenced by changes in PTH levels. Improved PTH levels in the spironolactone group (cases) additionally reduced the risk of heart failure hospitalization over and above that which was explained by spironolactone use only.

Two studies had previously evaluated the effect of mineralocorticoid receptor antagonism in lowering PTH in humans and attempted to elucidate the effects of aldosterone antagonist on PTH. Brown et al. used data from four human interventional studies and detected significant changes in PTH.¹⁰ However, it is not a common practice to use aldosterone antagonists in CKD patients especially due to the great risk of hyperkalemia and the fact that a level of serum potassium ≥5.5 meq/l is considered a relative contraindication.¹⁶ It is worth noting however that spironolactone is well tolerated in patients with mild to moderate CKD but requires close monitoring in the first month of its initiation.^17,18 In our study, we had a one year incidence of mild hyperkalemia (5.5–6 meq/l) in 5% of cases and severe hyperkalemia (>6 meq/l) in 3% of cases compared to 4% mild hyperkalemia and 2% severe hyperkalemia in CKD patients not on spironolactone. The results of this study should be interpreted cautiously as hyperkalemia can increase the risk of fatal arrhythmias, and further risk benefit analysis of these findings needs to be validated.

Previous studies had not only associated elevated PTH levels with greater severity in heart failure patients but to also be independently associated with cardiovascular mortality in these patients.¹⁹ The elevation in PTH seems to be driven by aldosterone related calcium wasting in the kidneys.²⁰ This pathway assumes a primary role for modulating PTH levels in patients with CKD due to inappropriate activation of the renin-angiotensin-aldosterone axis as a result of autoregulatory mechanisms in place to preserve glomerular filtration rate.^21–23

Brown et al. had speculated that aldosterone can directly act on mineralocorticoid receptors (MRs) at the parathyroid tissue to influence PTH synthesis as well as indirectly through increased urinary calcium excretion. This may explain why curative therapy for primary aldosteronism usually results in decreased PTH levels.¹⁰ In another study, primary aldosteronism was associated with reduced serum calcium levels and increased baseline urinary calcium excretion. The authors of the study concluded that primary aldosteronism with secondary hyperparathyroidism can be treated with MR antagonists.²⁴ Sodium is another mediator which has been shown to impose adverse effects on the renal tubules through aldosterone.²⁵ Calcium homeostasis can influence renal sodium filtration and reabsorption which in turn could lead to the observed association in our study.²⁵ Elevated levels of aldosterone are often associated with heart failure, especially heart failure with preserved ejection fraction.²⁶ This study elucidates the pathophysiological basis of this observation.

The findings of this study are consistent with several other studies illustrating the association between PTH and heart failure^5,19,27,28 while also being inconsistent with some others.^29,30 Though prior studies have shown similar associations, this is the largest study to date that has evaluated this relationship in patients with CKD.^31,32 Moreover, this study attempts to explain the clinical association between the use of aldosterone antagonists and observed changes in PTH levels.²⁰ These findings are consistent with basic scientific literature supporting the claim that aldosterone-mediated effects on tissues rely on MR sensitivity and signaling.³³

Another observation of our study was that decreased levels of PTH associated with concurrent use of an aldosterone antagonist did not improve mortality rates when compared to the controls. One may surmise that individuals placed on aldosterone antagonists were more likely to have had severe resistant hypertension requiring more aggressive therapy and therefore our inability to demonstrate a mortality benefit when compared to the matched controls. CKD is a strong risk factor for heart failure with preserved ejection fraction.^34,35 Aldosterone antagonists have not been shown to improve outcomes in individuals with heart failure with preserved ejection fraction.^36,37 This may also be true for individuals with CKD.

Strengths of this study include the completeness of our data and the relatively prolonged follow-up duration for the cases and the controls given their continued care at the same institution. There remains the possibility, however, that some of these individuals may have been admitted to outside hospitals, rendering us unable to collect their information. All of our patients were insured and in instances where they were admitted outside our primary institution, they were eventually transferred back or subsequently followed up at our hospital system, further assuring us of the completeness of our follow up. The study is limited by its retrospective design and the fact that the clinical diagnosis was based upon documentation of the treating physicians which may have led to information bias. The study findings help to further our understanding of the mechanistic properties of aldosterone inhibition in individuals with CKD. In the future, a controlled study design may allow us to better interpret these findings.

Conclusions

This study attempts to describe the use of aldosterone antagonism and its effects on PTH in patients with CKD. It also explores the pathophysiological aspects of this pathway and their potential utility in reducing and preventing heart failure hospitalization in a subgroup of patients who, though not widely studied in heart failure trials, nevertheless account for a significant proportion of heart failure related hospitalizations.

Footnotes

Conflict of interest

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Biermann

Neumann

Angermann

. Economic burden of patients with various etiologies of chronic systolic heart failure analyzed by resource use and costs. Int J Cardiol 2012; 156: 323–325.

Heidenreich

Trogdon

Khavjou

. Forecasting the future of cardiovascular disease in the United States: A policy statement from the American Heart Association. Circulation 2011; 123: 933–944.

Jaarsma

Van Der Wal

Hogenhuis

. Design and methodology of the coach study: A multicenter randomised coordinating study evaluating outcomes of advising and counselling in heart failure. Eur J Heart Fail 2004; 6: 227–233.

Khalid

Qureshi

. Impact of restarting warfarin therapy in renal disease anticoagulated patients with gastrointestinal hemorrhage. Ren Fail 2013; 35: 1228–1235.

Altay

Colkesen

Parathyroid hormone and heart failure: Novel biomarker strategy. Endocr Metab Immune Disord Drug Targets 2013; 13: 100–104.

Sugimoto

Dohi

Onishi

. Interrelationship between haemodynamic state and serum intact parathyroid hormone levels in patients with chronic heart failure. Heart 2013; 99: 111–115.

Altay

Zorlu

Binici

. Relation of serum parathyroid hormone level to severity of heart failure. Am J Cardiol 2012; 109: 252–256.

Loncar

Bozic

Dimkovic

. Association of increased parathyroid hormone with neuroendocrine activation and endothelial dysfunction in elderly men with heart failure. J Endocrinol Invest 2011; 34: e78–e85.

Sugimoto

Tanigawa

Onishi

. Serum intact parathyroid hormone levels predict hospitalisation for heart failure. Heart 2009; 95: 395–398.

10.

Brown

Williams

Luther

. Human interventions to characterize novel relationships between the renin-angiotensin-aldosterone system and parathyroid hormone. Hypertension 2014; 63: 273–280.

11.

Hamdy

Kanis

Beneton

. Effect of alfacalcidol on natural course of renal bone disease in mild to moderate renal failure. Brit Med J 1995; 310: 358–363.

12.

Arenas

Alvarez-Ude

Torregrosa

. Consequences of the implementation of K/DOQI clinical practice guidelines for bone metabolism and disease in chronic kidney disease in a population of patients on chronic hemodialysis. J Nephrol 2007; 20: 453–461.

13.

Bailie

Massry

SG.

Clinical practice guidelines for bone metabolism and disease in chronic kidney disease: An overview. Pharmacotherapy 2005; 25: 1687–1707.

14.

National Kidney Foundation K/DOQI clinical practice guidelines for bone metabolism and disease in chronic kidney disease. Am J Kidney Dis 2003; 42: S1–S201.

15.

Levey

Stevens

Schmid

. A new equation to estimate glomerular filtration rate. Ann Int Med 2009; 150: 604–612.

16.

Lima

Ochiai

Cardoso

. Hyperkalemia during spironolactone use in patients with decompensated heart failure. Arq Bras Cardiol 2008; 91: 177–182, 194–179.

17.

Edwards

Steeds

Chue

. The safety and tolerability of spironolactone in patients with mild to moderate chronic kidney disease. Br J Clin Pharmacol 2012; 73: 447–454.

18.

Heshka

Ruzicka

Hiremath

. Spironolactone for difficult to control hypertension in chronic kidney disease: An analysis of safety and efficacy. J Am Soc Hypertens 2010; 4: 295–301.

19.

Schierbeck

Jensen

Bang

. Parathyroid hormone and vitamin D-markers for cardiovascular and all cause mortality in heart failure. Eur J Heart Fail 2011; 13: 626–632.

20.

Chhokar

Sun

Bhattacharya

. Hyperparathyroidism and the calcium paradox of aldosteronism. Circulation 2005; 111: 871–878.

21.

Sim

Shi

Calara

. Association of plasma renin activity and aldosterone-renin ratio with prevalence of chronic kidney disease: The Kaiser Permanente Southern California cohort. J Hypertens 2011; 29: 2226–2235.

22.

Ponda

Hostetter

TH.

Aldosterone antagonism in chronic kidney disease. Clin J Am Soc Nephrol 2006; 1: 668–677.

23.

Hostetter

Ibrahim

HN.

Aldosterone in chronic kidney and cardiac disease. J Am Soc Nephrol 2003; 14: 2395–2401.

24.

Pilz

Kienreich

Drechsler

. Hyperparathyroidism in patients with primary aldosteronism: Cross-sectional and interventional data from the GECOH study. J Clin Endocrinol Metab 2012; 97: E75–E79.

25.

Tomaschitz

Pilz

. Interplay between sodium and calcium regulatory hormones: A clinically relevant research field. Hypertension 2014; 63: 212–214.

26.

Ohtani

Ohta

Yamamoto

. Elevated cardiac tissue level of aldosterone and mineralocorticoid receptor in diastolic heart failure: Beneficial effects of mineralocorticoid receptor blocker. Am J Physiol Regul Integr Comp Physiol 2007; 292: R946–R954.

27.

Gruson

Lepoutre

Ahn

. Comparison between intact and bioactive parathyroid hormone assays in patients with severe heart failure. Clin Biochem 2013; 46: 391–394.

28.

Hagstrom

Ingelsson

Sundstrom

. Plasma parathyroid hormone and risk of congestive heart failure in the community. Eur J Heart Fail 2010; 12: 1186–1192.

29.

Diez-Perez

Nogues

Mellibovsky

. Bone mass, vitamin D deficiency and hyperparathyroidism in congestive heart failure. Am J Med 1998; 105: 358–359.

30.

Zittermann

Schleithoff

Gotting

. Poor outcome in end-stage heart failure patients with low circulating calcitriol levels. Eur J Heart Fail 2008; 10: 321–327.

31.

Laguardia

Dockery

Bhattacharya

. Secondary hyperparathyroidism and hypovitaminosis D in African-Americans with decompensated heart failure. Am J Med Sci 2006; 332: 112–118.

32.

Khouzam

Dishmon

Farah

. Secondary hyperparathyroidism in patients with untreated and treated congestive heart failure. Am J Med Sci 2006; 331: 30–34.

33.

Tomaschitz

Ritz

Pieske

. Aldosterone and parathyroid hormone interactions as mediators of metabolic and cardiovascular disease. Metabolism 2014; 63: 20–31.

34.

Smith

Thorp

Gurwitz

. Chronic kidney disease and outcomes in heart failure with preserved versus reduced ejection fraction: The Cardiovascular Research Network Preserve study. Circ Cardiovasc Qual Outcomes 2013; 6: 333–342.

35.

McAlister

Ezekowitz

Tarantini

. Renal dysfunction in patients with heart failure with preserved versus reduced ejection fraction: Impact of the new Chronic Kidney Disease-Epidemiology Collaboration group formula. Circ Heart Fail 2012; 5: 309–314.

36.

Neil

Singh

Frenneaux

. Aldosterone inhibition in patients with heart failure with preserved ejection fraction. JAMA 2013; 310: 205–207.

37.

Patel

Fonarow

Kitzman

. Aldosterone antagonists and outcomes in real-world older patients with heart failure and preserved ejection fraction. JACC Heart Fail 2013; 1: 40–47.