Primary aldosteronism (PA) is one of the common forms of curable hypertension. Recent views have suggested that PA is far from being relatively benign, as it was previously thought, but it is associated with a variety of cardiovascular and renal sequelae that reflect the capability of inappropriately elevated aldosterone to induce tissue damage over that induced by hypertension itself. The evidence supporting these views has been obtained from experiments conducted in hypertensive animal models and studies involving patients with PA. Preclinical studies have also indicated that aldosterone causes cardiovascular and renal tissue damage only in the context of inappropriate salt status. It has been suggested that untoward effects of high-salt intake are dependent on activation of mineralocorticoid receptors (MRs) that might result from increased oxidative stress and changes in the intracellular redox potential. Unilateral adrenalectomy or treatment with MR antagonists are the current options for treating an aldosterone-producing adenoma (APA) or idiopathic adrenal hyperplasia (IHA). Treatments are effective in correcting hypertension and hypokalemia, and currently available information on their capability to prevent cardiovascular events and deterioration of renal function indicates that surgery and medical treatment are equally beneficial in the long term.
Recent evidence indicates a greater prevalence of primary aldosteronism (PA) among patients with hypertension than the previously accepted estimate of ~1%.1–3 Such increased prevalence is the result of widespread use of the aldosterone:renin ratio that has led to a more efficient screening of this condition.4,5 Because initial descriptions of patients with PA reported low incidence of cardiovascular events,6 this form of hypertension was traditionally considered relatively benign. This was generally ascribed to the suppression of the renin–angiotensin axis that occurs as a consequence of the aldosterone-induced body fluid expansion.7 More recent views, however, have suggested that exposure to inappropriately elevated aldosterone might result in substantial cardiovascular8 and renal9 damage. Now we know that PA is associated with a variety of cardiovascular and renal sequelae10 that reflect the capability of aldosterone to induce tissue damage over that induced by hypertension itself.
Aldosterone-producing adenoma (APA) and bilateral idiopathic adrenal hyperplasia (IHA) are the most common causes of PA, and unilateral adrenalectomy or chronic administration of mineralocorticoid receptor (MR) antagonists effectively reduces blood pressure in these patients. Normalization of blood pressure, however, is not the only goal of treatment of PA and effective prevention of organ complications is mandatory in these patients. Both cardiovascular and renal outcomes might benefit from treatment in the long term, but the relative efficacy of adrenalectomy and MR antagonists needs further evaluation, being that the aldosterone-induced tissue damage is the main factor that could justify the cost of increasing efforts in screening of disease and differentiation of subtypes.11 Much debate has been generated upon this issue due to the opposite views of researchers who claim the opportunity for widespread screening of PA and demonstration of lateralized aldosterone secretion by adrenal venous sampling12 and those who would restrict diagnostic procedures to a much more selected population.13 The latter position has been supported recently by the evidence of significant blood pressure response to spironolactone in a large cohort of patients with PA (both APA and IHA) and resistant hypertension.14 This review outlines the evidence of cardiovascular and renal involvement in patients with PA and emphasizes the most recent findings in the evaluation of outcomes following either surgical or medical treatment.
Experimental Evidence of MR-Related Organ Damage
A growing body of evidence suggests that exposure to inappropriate aldosterone levels for salt status or activation of the MR can produce massive myocardial, vascular, and renal tissue injury with mechanisms that are independent of blood pressure (reviewed in ref. 15). Landmark experiments demonstrated that chronic aldosterone infusion causes myocardial fibrosis in rats that are maintained on a high-salt diet.16 Later on, in a series of elegant studies, Rocha and colleagues demonstrated that aldosterone-induced myocardial fibrosis is preceded by inflammatory changes of perivascular tissue,17 and that both inflammation and fibrosis can be prevented by administration of MR antagonists or adrenalectomy.18 Similar evidence was obtained in the kidney of uninephrectomized19 and stroke-prone spontaneously hypertensive rats20 in which aldosterone produced intrarenal vascular damage, glomerular injury, and tubulointerstitial fibrosis. Elevated aldosterone also caused aortic fibrosis21 and hypertrophy22 in different rat models of hypertension, and administration of eplerenone to hypertensive rats corrected vascular remodeling and fibrosis,23 suggesting a MR-mediated mechanism.
These animal studies consistently indicate that aldosterone causes tissue damage only in the context of inappropriate salt status. It was suggested that untoward effects of high-salt intake are largely dependent on activation of MRs and that this activation might reflect increased oxidative stress.24 MRs are found in epithelial and nonepithelial tissues with high affinity for aldosterone and glucocorticoid hormones, such as cortisol and corticosterone. Under physiological conditions, the majority of MRs in nonepithelial tissues are occupied by greater concentrations of cortisol,25 whereas in epithelial tissues, binding of cortisol to MR is prevented by 11β-hydroxysteroid dehydrogenase (11β-HSD2), the enzyme that converts cortisol to the receptor-inactive cortisone. In addition to the conversion of cortisol to cortisone, activity of 11β-HSD2 generates NADH from NAD and produces changes in the intracellular redox potential that might, in turn, inactivate the glucocorticoid–MR complex.26 11β-HSD2 is not present in nonepithelial tissues including the heart, but in such tissues, changes of the intracellular redox potential can result from generation of reactive oxygen species and thereby affect the activity of the MR.26In vitro experiments have demonstrated that changes in the redox potential of cardiomyocytes by exposure to oxidized glutathione turn cortisol from being a MR antagonist to an agonist.27 More recently, it has been demonstrated that aldosterone itself induces changes in the intracellular redox potential in diverse cell types28,29 through an activation of the NOX1 catalytic subunit of NAD(P)H oxidase. This aldosterone-dependent change in the redox potential is amplified by exposure to high concentrations of salt29 leading to increased production of reactive oxygen species and thereby to cellular and tissue injury.
Thus, in addition to the well-known effects of salt loading on epithelial swelling, vascular stiffening and blood pressure increase, some effects of salt loading might depend on MR activation and reflect, in different tissues, impairment of 11β-HSD2 activity and/or increased oxidative stress, both mechanisms possibly leading to changes in the intracellular redox state. The distinction of these effects of salt from those generated at the tissue level by elevated aldosterone is complex and even genetic manipulations could not help in the understanding of their respective roles. In fact, both cardiac overexpression of the MR30 and cardiac-specific induction of aldosterone production31 do not cause cardiac fibrosis, whereas fibrosis results from knockdown of the cardiac MR by the use of antisense mRNA.32
Is PA Associated with Greater Cardiovascular Risk than Primary Hypertension?
The RALES,33 EPHESUS,34 and 4-E studies35 have provided strong evidence that MR blockade has positive effects on the circulatory system. In these clinical trials, significant benefit on cardiovascular outcomes was demonstrated with the use of MR antagonists in patients with heart failure and hypertension. Similar evidence of beneficial effects of MR antagonists has been obtained in smaller clinical trials that were conducted in patients with diabetic nephropathy36 or chronic kidney disease due to various conditions.37 However, only sparse reports have evaluated the association of cardiovascular events with PA. Conn et al.,6 in their original description of 145 patients, reported only three cases of cerebrovascular accidents, but subsequent cross-sectional studies have yielded variable results,7,38,39,40 showing a prevalence of cardiovascular complications comprised from 14 to 35%. In a prevalence analysis conducted in Japan,39 stroke was found in 9 of 58 patients (15.5%) with PA, whereas coronary artery disease was demonstrated only in one. A major limitation of these cross-sectional evaluations was due to lack of appropriate controls with primary hypertension.
The best available evidence supporting the presence of greater cardiovascular risk in patients with PA comes from longitudinal, retrospective studies. Milliez et al.41 were the first to examine a large cohort of patients with APA or IHA, and to report a significantly higher rate of events such as atrial fibrillation (7.3% vs. 0.6%), myocardial infarction (4.0% vs. 0.6%), and stroke (12.9% vs. 3.4%), in comparison with patients with primary hypertension. More recently, Catena et al.42 have documented increased odds ratios of myocardial infarction or reversible ischemia (2.80; 95% CI, 1.30–6.06), stroke or transient ischemic attack (4.36; 1.49–12.80), and sustained arrhythmias (4.93; 1.89–12.91) in patients with PA than patients with primary hypertension who were accurately matched for the severity and duration of disease. These two studies consistently reported that the rate of cardiovascular complications was comparable in patients with tumoral or idiopathic disease. In the recently published data from the German Conn's Registry,43 the largest cohort ever reported in the literature, prevalence of coronary artery disease was 16.3%, and atrial fibrillation and ventricular arrhythmias occurred in 7.1 and 5.2% of patients, respectively. Although the German study lacked a reference control group, prevalence of cardiovascular complications in PA patients was far greater than that reported in the literature for patients with essential hypertension of comparable cardiovascular risk profile. More recent evidence obtained in a French cohort of 460 patients with PA44 has reported greater frequency of heart failure (7.4% vs. 3.6%), coronary artery disease (5.6% vs. 1.0%), and atrial fibrillation (2.0% vs. 0.3%) in these patients than in 1,291 matched patients with essential hypertension. Thus, current evidence convincingly and conclusively demonstrates that patients with PA are at greater risk of cardiovascular events than those with primary hypertension.
Possible contribution to increased cardiovascular risk in patients with PA could be due to abnormalities of glucose metabolism that have been reported in these patients. An increased prevalence of the metabolic syndrome has been reported in patients with PA who were compared with essential hypertensive controls,45 although this difference was due only to significantly greater presence of diabetic patients (8.2% vs. 3.4%) that were not excluded from the study. Recent evidence indicates that increased circulating levels of resistin are associated with PA and might account for the unfavorable cardiometabolic profile demonstrated in these patients.46 The only study that used the hyperinsulinemic–euglycemic clamp to assess sensitivity to insulin in a sufficiently large group of patients with PA reported that, although these patients were more insulin-resistant than matched normotensive subjects, the severity of insulin resistance was less than in patients with essential hypertension.47 In this study, long-term follow-up demonstrated that both surgical and medical treatment of PA improved insulin sensitivity significantly. Another study, has reported a difference between treatment results in PA patients, in that surgical treatment of adrenal adenoma improved glucose tolerance, whereas medical treatment blocked further progression of the metabolic complications, rather than reversing them.48 Therefore, available evidence is insufficient to conclusively support a contribution of insulin resistance to the greater cardiovascular risk of patients with PA.
Abnormalities of Cardiac Function and Vascular Structure in PA
Cardiac structural and functional changes are common consequences of hypertensive states and have been demonstrated in patients with PA, along with some additional surrogate vascular end points. Many cross-sectional echocardiographic evaluations have reported an excess increase of left ventricular (LV) mass in patients with PA as compared with other types of hypertensive disease,49,50,51,52 although this finding has not been confirmed in other studies.53 Muiesan et al. have reported that the frequency of inappropriate LV mass is increased in patients with PA, even in the absence of LV hypertrophy as defined by current criteria.54 This observation strongly suggests that elevated aldosterone contributes to the increase in LV mass beyond the amount needed to compensate the blood pressure–related hemodynamic load. In PA, excess LV hypertrophy occurs in association with an abnormal pattern of LV filling,51,52 indicating the presence of diastolic dysfunction, whereas systolic function is generally found to be comparable with that of patients with primary hypertension. Furthermore, diastolic dysfunction of PA has been associated with the evidence of abnormal videodensitometric properties of the LV wall,51 suggesting myocardial fibrosis. Surrogate vascular end points have also been documented in patients with PA. Excess vascular fibrosis was demonstrated in resistance arteries,55 and increased arterial wall stiffness was found in conduit arteries56 of patients with PA as compared with patients with primary hypertension. The intima-media thickness was also found to be increased in carotid arteries,57 and carotid artery lesions were more frequent58 in such patients in comparison with patients with primary hypertension.
All these aldosterone-related cardiac and vascular abnormalities could contribute to the increased cardiovascular risk observed in patients with PA41,42,43,44 and account for greater incidence of heart failure, arrhythmia, and cerebrovascular disease. Mechanisms causing more frequent coronary heart disease, however, are only hypothetical, including proinflammatory effects of aldosterone on the arterial wall,17 excess LV hypertrophy with increased oxygen consumption,49,50,51,52,54 endothelial dysfunction,59 and remodeling of resistance vessels55 and will need further investigation.
Cross-sectional evidence of associations between cardiac phenotypes and possible causative factors is of limited value, and findings should be confirmed in longitudinal evaluations. Most echocardiographic observations of cardiac changes after treatment of PA are confined to short-term follow-up studies, mostly after removal of an adrenal adenoma. Initial observations showed that in patients with APA treated by adrenalectomy, both LV mass50 and LV filling patterns51 were normalized 1 year after surgery, whereas patients who were treated for a year with spironolactone showed no comparable LV hypertrophy regression. A recent study has provided data on long-term echocardiographic follow-up in a cohort of patients with PA, after either surgical or medical treatment.52 This 7-year study has demonstrated that patients treated with either adrenalectomy or spironolactone have significant and comparable decrease of LV mass, although decrease is significant within the first year only after surgery. In both treatment groups, baseline LV mass was correlated with plasma aldosterone concentration, and this concentration was independent predictor of changes of LV mass after treatment. Consistent with these findings, a recent study has reported significant decrease of LV mass in 45 patients with PA who have been re-evaluated after an average follow-up of 36 months following either surgery or treatment with spironolactone.60 Even in this study, response of LV mass in adrenalectomized patients occurred earlier than in those treated with spironolactone. These studies demonstrate that the decrease of LV mass obtained with treatment of PA is only partially explained by blood pressure reduction, clearly supporting a role of aldosterone that is independent of the hemodynamic overload.
As previously noted, MRs have been demonstrated in cardiomyocytes,61 and their activation might contribute to myocardial hypertrophy of PA via mechanisms that include modulation of ionic movements and accelerated fibrosis.62 The latter mechanism might, in turn, result from interactions of aldosterone with other hormones such as angiotensins, endothelin, and bradykinin, activation of inflammatory cells, and stimulation of fibroblast proliferation and collagen synthesis.15 Interruption of these receptor-mediated mechanisms might explain why, in the long term, treatment of PA with MR receptor antagonists has comparable effects with the removal of an aldosterone-secreting adenoma in reducing LV mass, although this response occurs later than after adrenalectomy. Persistent hyperaldosteronemia with possible persistence of the so-called nongenomic effects of aldosterone8 might hypothetically explain why regression of LV hypertrophy requires longer time to occur in patients treated with MR antagonists than in those treated with surgery. These nongenomic effects of aldosterone are detectable, at the cell level, much earlier than the classical MR-mediated effects8 and might be mediated by pathways that are independent from binding of the hormone to MR. An alternative explanation for the beneficial effects of spironolactone on LV mass regression could be that this compound may also function as an “inverse agonist” by binding to the MR, but inducing changes opposite to those induced by other agonists.63
Which Role Plays the Kidney in PA?
As noted above, MR blockade prevents the development of renal tissue damage in hypertensive animal models,19,20 and MR antagonists have been beneficial in proteinuric patients with diabetic nephropathy36 or chronic kidney disease.37 Cross-sectional evaluations in PA have shown a high degree of variability in the prevalence of clinically relevant renal damage.39,64,65,66,67,68 Prevalence of overt proteinuria varied from 8%64 to 24%,39 with a disparity that could be explained by different duration and severity of disease. Danforth et al. observed moderate renal tissue damage in renal biopsies obtained from patients with PA,64 and Nishimura et al.39 reported renal insufficiency in 7% of 58 patients with PA. In another single-center study, 24-h creatinine clearance of <60ml/min/1.73m2 has been found in 7% of 56 patients with PA,68 whereas in the German Conn's Registry, increased plasma creatinine concentration was present in a substantially higher percentage of patients.69 In the PAPY (Primary Aldosteronism Prevalence in Italy) study,67 24-h microalbuminuria was significantly greater in patients with both APA and IHA in comparison with primary hypertensive controls. Important information on the role of the kidney in PA has been obtained from two prospective studies, with short-term and long-term follow-up after treatment. Ribstein et al.70 have reported a significant decrease in urinary albumin excretion after adrenalectomy in 25 patients with adrenal adenoma who were followed up for 6 months. In a 9-year follow-up study of patients with PA and well-documented microalbuminuria, we have shown that microalbuminuria is more likely to subside to normal levels after treatment than to progress to overt proteinuria.71 In this study, restoration of normal albumin excretion was more frequent in patients with PA than in those with primary hypertension and this effect was independent of blood pressure. These prospective studies have consistently indicated that PA is characterized by partially reversible renal dysfunction, suggesting that albuminuria is, at least in part, a marker of a renal hemodynamic defect. In agreement with the findings of previous renal function studies,72 some of which were conducted in more experimental settings,73 these two studies have demonstrated the presence of relative glomerular hyperfiltration in patients with PA as compared with appropriately matched patients with primary hypertension. Glomerular hyperfiltration is rapidly reversed within the first few months after treatment, an effect that could be explained by decreased blood pressure, decreased extracellular volume, recovery of renin activity, postoperative aldosterone suppression, and removal of direct vasomotor effects of aldosterone on intrarenal vessels.74 A very recent analysis of 408 patients of the German Conn's Registry69 has confirmed that glomerular filtration declines soon after treatment of PA and remains relatively stable thereafter. Moreover, evaluation of intrarenal Doppler velocimetric indexes has demonstrated significantly lower intrarenal vascular resistance in patients with PA in comparison with patients with primary hypertension, and reversal of the abnormal intrarenal hemodynamic pattern 1 year after treatment.75
Thus, findings of longitudinal studies consistently indicate that renal dysfunction in PA is characterized by reversible glomerular hyperfiltration that is associated with decreased intrarenal vascular resistance and contributes to increase urinary albumin losses. The frequency of regression of microalbuminuria in PA suggests that urinary albumin excretion is, at least in part, a marker of functional rather structural renal changes76 (Figure 1). On the other hand, long-term persistence of albuminuria in a substantial proportion of patients with PA71 is associated with detectable baseline plasma renin levels,66,68 suggesting coexistence of structural intrarenal vascular damage presumably due to long-standing hypertension prior to treatment.
Role of the kidney in primary aldosteronism. Two separate aspects should be considered. On one hand, there are functional adaptations, induced by increased tubular sodium reabsorption and leading to expansion of extracellular volume, hypertension, increased renal perfusion pressure, and suppression of renin with decreased intrarenal vascular resistance. These changes result in glomerular hyperfiltration and increased sodium excretion, with recovery of a steady state. On the other hand, there is structural damage, involving primarily the intrarenal vessels, that might result from chronic hypertensive insult and direct effects of aldosterone. This might lead to decreased glomerular perfusion and stimulation of renin production that could escape from suppression by excess plasma aldosterone. In this view, lack of renin suppression indicates a progressed stage of renal disease that predicts low possibility of blood pressure cure. ECV, extracellular fluid volume.
Cardiovascular and Renal Outcomes of Treatment in PA
The established treatment for APA is adrenalectomy, whereas MR antagonists are used to manage patients with IHA. The effects of these treatments on blood pressure, and the predictors of blood pressure cure have been overviewed in thorough reviews1 and recent guidelines.77 Although PA is considered correctable, in many cases, hypertension may persist after surgical or medical treatment, and only approximately one-third of patients achieve values of <140/90mmHg without the use of additional antihypertensive agents.52,66,78
Current views indicate that adrenalectomy is the optimal treatment for PA patients with evidence of lateralized aldosterone secretion because surgery confers a greater possibility of cure and avoids the possible side effects of MR antagonists.1,77 Side effects are a serious problem with spironolactone, but use of low doses (25–50mg/day), as currently recommended77 or, alternatively, use of the androgen-receptor inactive eplerenone, although this drug still needs testing in patients with PA, may restrict this problem to a few patients. This is important because differentiation of lateralized aldosterone secretion is labor-intensive and expensive, and only very few centers have the technical ability to perform routine adrenal venous sampling successfully.4,12 For these reasons and because of the relatively high frequency of PA, it would be unrealistic to recommend adrenal venous sampling to all patients. Cautious clinical judgment would be advisable for every single patient, gathering together clinical, biochemical, and imaging information, before deciding which treatment to offer. In this context, it should also be considered that recent evidence has indicated possible beneficial effects of unilateral adrenalectomy in patients with bilateral PA.79
Also and most important, it should be kept in mind that the mandatory goal of treatment in patients with PA is the prevention or recovery of organ damage, in order to decrease the risk of cardiovascular events and renal failure. To this matter, evidence linking, in the long term, treatment of PA with cardiovascular and renal prevention is recent and refers to a single-center experience. Cardiovascular outcomes have been compared in 108 patients with primary hypertension and in 54 patients with PA who had comparable risk factors, but greater retrospective incidence of coronary artery disease, cerebrovascular events, and sustained arrhythmias.42 Patients were prospectively followed up for a mean of 7.4 years after adrenalectomy or treatment with spironolactone, with a combined end point including myocardial infarction, stroke, any type of revascularization procedure, and sustained arrhythmias. During follow-up, blood pressure was comparable in the PA and primary hypertension group, and 10 patients in the PA group and 19 in the primary hypertension group reached the end point (P = 0.85). Cox analysis showed that age <52 years and a history of hypertension lasting <10 years were associated with a better cardiovascular outcome in PA. Actuarial analysis of patients treated with adrenalectomy vs. spironolactone did not reveal significant difference in the occurrence of the combined end point (HR, 1.26; 95% CI, 0.36–4.44; P = 0.71). The long-term outcomes of renal function after treatment of PA were investigated in the same cohort, by measuring the rates of change of glomerular filtration rate and albuminuria.71 After the initial fall in creatinine clearance, due to correction of the aldosterone-induced intrarenal hemodynamic adaptation, subsequent declines of glomerular filtration in patients with PA (−1.15ml/min/1.73m2/year) and primary hypertension (−1.06ml/min/1.73m2/year) were comparable (P = 0.49). Urinary albumin losses did not differ between patients with PA and primary hypertension during the long-term phase of follow-up (P = 0.56). Analysis of renal outcomes in patients with PA who were treated with adrenalectomy or spironolactone did not reveal significant difference (P = 0.87). These studies clearly demonstrate that, in patients with PA, the incidence of cardiovascular events and renal impairment does not differ from patients with primary hypertension when the effects of excess aldosterone are permanently removed and that, in this view, both adrenalectomy and aldosterone antagonists are of considerable therapeutic value.
The relationship between aldosterone and hypertension is unquestioned. Today, an increasing body of evidence indicates that elevated levels of aldosterone or activation of its receptors, occurring in the presence of inappropriate salt status, cause cardiac, vascular, and renal tissue damage independent of blood pressure levels. Mechanisms leading to this damage appear to depend from generation of reactive oxygen species and changes of redox potential, and the availability of aldosterone-synthase inhibitors and aldosterone-synthase deficient mice will help to better understand the relative importance of aldosterone levels as compared with salt effects and MR activation.
In PA, excess organ complications indicate that this condition is far from benign and that these complications are not simply explained by blood pressure elevation. Initial long-term studies indicate that both surgical and medical treatment of PA effectively lower blood pressure, correct subclinical organ damage, and decrease the risk of cardiovascular events and renal disease progression. In this view, appropriate timing in the identification of this endocrine disorder remains mandatory in all hypertensive patients, in order to effectively prevent those complications that are associated with persistence of excess aldosterone effects. Based upon results of long-term studies, clinicians should not feel uncomfortable when they decide to treat their PA patients with the appropriate doses of MR antagonists when the evidence of lateralized aldosterone secretion cannot be conclusively reached.
This work was supported by research grants of the Italian Ministry of the University and from the Pier Silverio Nassimbeni Foundation.
The authors declared no conflict of interest.
- conn adenoma
- oxidative stress
- renal function
- cardiovascular system
- magnetic resonance spectroscopy
- melkersson-rosenthal syndrome
- animal model
- mineralocorticoid receptor
- surgical procedures, operative
- unilateral adrenalectomy
- adrenal hyperplasia
- cardiovascular event
- tissue damage
- medical management
- renal complications