We evaluated the prevalence of primary aldosteronism in subjects newly diagnosed with hypertension who were referred to specialized hypertension centers. An aldosterone-producing adenoma was diagnosed in the subjects with lateralized aldosterone secretion, adenoma at surgery and on pathologic evaluation, and a blood pressure fall after adrenalectomy. Evidence of excess autonomous aldosterone secretion without such criteria led to a diagnosis of idiopathic hyperaldosteronism. Aldosterone-producing adenoma and idiopathic hyperaldosteronism were conclusively diagnosed in 4.8% and 6.4% of the subjects, respectively. Thus, with a prevalence of 11.2%, primary aldosteronism is quite common in patients with newly diagnosed hypertension.
The prevalence of primary aldosteronism, a curable cause of arterial hypertension, has remained uncertain because of the retrospective nature and small size of most studies, which involved selected subjects.1 Determining the prevalence of primary aldosteronism is difficult because it requires: (1) prospective investigation of a large cohort of newly diagnosed, consecutive, unselected patients with hypertension to minimize the chance of selection bias; (2) the use of predefined state-of-the-art diagnostic criteria; and (3) attaining a certain diagnosis of primary aldosteronism. Attaining this unequivocal diagnosis is crucial because this condition can only be conclusively diagnosed retrospectively in the patients cured or improved by removal of an aldosterone-producing adenoma or unilateral adrenal hyperplasia.2 The diagnosis of idiopathic hyperaldosteronism, the other major primary aldosteronism subtype, can only be presumed because it cannot be cured surgically or reliably differentiated from low-renin primary (essential) hypertension.3 Therefore, we conducted the Primary Aldosteronism Prevalence in Hypertensives (PAPY) study, a large-scale, multicenter, prospective study to determine the prevalence of aldosterone-producing adenoma in newly diagnosed hypertensive subjects referred to specialized centers for hypertension.4
Subjects and methods
Consecutive newly diagnosed hypertensive subjects who had been referred to specialized hypertension centers throughout Italy were enrolled in the study. Exclusion criteria included a previous diagnosis of any secondary forms of hypertension and refusal to participate in the study.
Mineralocorticoid receptor antagonist therapy was discontinued for at least 6 weeks before the screening test; diuretics, beta blockers, angiotensin-converting enzyme inhibitors, and angiotensin II type 1 receptor antagonists were discontinued for at least 2 weeks before screening tests were conducted. Subjects were permitted to take doxazosin (Cardura) or a long-acting calcium channel blocker, or both, to minimize the risks of uncontrolled hypertension.
After fasting overnight and sitting quietly for 1 hour, baseline measurements of Na+, K+, plasma renin activity, aldosterone, and cortisol were taken between 7 and 9 AM. The same measurements were taken 60 minutes after captopril (Capoten; 50 mg by mouth) administration. Na+ and K+ excretion were measured by 24-hour urine collection. We calculated the ratio of aldosterone (in ng/dL) to plasma renin activity (in ng/mL/hr, expressed as ARR) at baseline and after captopril administration.3 For all subjects with an ARR ≥ 40 at baseline or an ARR ≥ 30 after captopril administration, measurements of aldosterone, cortisol, and plasma renin activity were also taken after the infusion of 2 liters of saline over 4 hours.
All subjects with an ARR ≥ 40 at baseline and ≥ 30 after captopril administration underwent high-resolution computed tomography (CT) or magnetic resonance imaging (MRI), or both. Adrenal vein sampling5 was performed to identify the cause of primary aldosteronism without corticotropin (ACTH) stimulation4,6 and was interpreted as described previously.5 At centers where adrenal vein sampling was not available, dexamethasone-suppressed adrenocortical scintigraphy was used to demonstrate a lateralized aldosterone excess.
Any condition that might artificially increase serum K+, including fist clenching and the use of a tourniquet, were avoided when drawing blood. Subjects with a serum K+ of ≤ 3.5 mEq/L were considered to have hypokalemia. Details regarding plasma renin activity, aldosterone, and cortisol assays, and normal ranges have been previously reported.7
Subjects with evidence of autonomous excess aldosterone production, including an ARR ≥ 40 at baseline and ≥ 30 after captopril administration, were determined to have primary aldosteronism.8 Subjects with: (1) lateralization of aldosterone secretion at adrenal vein sampling or adrenocortical scintigraphy; (2) surgery; (3) pathology; and (4) outcome of adrenalectomy at follow-up were determined to have aldosterone-producing adenoma.7 Subjects with primary aldosteronism but without evidence of a lateralized aldosterone excess were considered to have idiopathic hyperaldosteronism (Figure 1).
Figure 1. The protocol for the recruitment and investigation of the Primary
Aldosteronism Prevalence in Hypertensives (PAPY) study. ARR indicates
aldosterone/renin ratio; LDF, logistic discriminant analysis; CT, computed
tomography; MR, magnetic resonance; AVS, adrenal vein sampling; IHA,
idiopathic hyperaldosteronism; APA, aldosterone-producing adenoma; AHA,
American Heart Association. (Reprinted with permission from Rossi GP, Bernini
G, Caliumi C, et al. A prospective study of the prevalence of primary
aldosteronism in 1125 hypertensive patients. J Am Coll Cardiol. 2006;48:
One-way analysis of variance with Bonferroni test and chi-square analysis was performed. A P value of < .05 was accepted as statistical significance. The sensitivity, specificity, and accuracy (the area under the receiver operating characteristic [ROC] curve) of each test for prediction of the diagnosis were calculated. Receiver operating characteristic curve analysis was also used to determine the optimal cutoff values for each test.
Table 1 shows the baseline data for the subjects. Besides the expected differences of plasma renin activity, aldosterone, and serum K+, the comparison of subjects with and without primary aldosteronism also showed an older age and a higher blood pressure in the primary aldosteronism group.
Untreated subjects and those treated with captopril experienced no problems or adverse effects. There was no effect of treatments on the screening test results in either the entire primary aldosteronism group or the primary hypertension group. Of the 1180 subjects, 55 subjects did not complete the entire diagnostic workup, making a conclusive diagnosis possible in 1125 (95.3%). Of these, 126 subjects were diagnosed with primary aldosteronism, which resulted in an overall prevalence of 11.2%, with no difference between the sexes. Primary aldosteronism was caused by an aldosterone-producing adenoma in 54 subjects (42.8%) and by idiopathic hyperaldosteronism in the remaining 72 (57.2%).
Table 1. Demographic characteristic of the subjects (n = 1125) enrolled in
the PAPY study who had a conclusive diagnosis.
s-Aldosterone indicates sitting plasma aldosterone; s-PRA, sitting plasma
renin activity; BP, blood pressure; HR, heart rate; K+, potassium level; Na+,
sodium level. Data are mean ± standard error of the mean, except for
serum and urinary K+, serum and urinary Na+, s-aldosterone, s-PRA, and s-
cortisol, where median (range) are shown. Normal values for s-aldosterone:
20-150 pg/mL; s-PRA: 0.51-2.64 ng/mL/h; s-cortisol: 50-250 ng/mL.
(Reprinted with permission from Rossi GP, Bernini G, Caliumi C, et al. A
prospective study of the prevalence of primary aldosteronism in 1125
hypertensive patients. J Am Coll Cardiol. 2006;48:2293-2300.)
About 20% of the subjects tested positive for primary aldosteronism by either the ARR at baseline or ARR after captopril administration. Table 2 shows the sensitivity, specificity, and accuracy of these 2 tests at the prespecified cutoff levels. These tests, therefore, allowed selection of subgroups of subjects with a higher primary aldosteronism prevalence. The ARR after captopril administration had a lower sensitivity but a higher specificity than the ARR at baseline.
Table 2. Sensitivity, specificity, and overall accuracy of each test at the pre-
ARR indicates aldosterone/renin ratio. (Reprinted with permission from Rossi
GP, Bernini G, Caliumi C, et al. A prospective study of the prevalence of
primary aldosteronism in 1125 hypertensive patients. J Am Coll Cardiol. 2006;
On CT and MRI scans, the tumor diameter was < 10 mm in 17% of the subjects with aldosterone-producing adenoma, between 10 and 20 mm in 28%, and > 20 mm in 55%. The prevalence of spontaneous hypokalemia was 9.6%; however, it was observed in about half of the subjects with aldosterone-producing adenoma and in only 16.9% of those with idiopathic hyperaldosteronism (Figure 2). The prevalence of aldosterone-producing adenoma and idiopathic hyperaldosteronism increased significantly ( < .001) with increasing severity of hypertension at the screening test, from high-normal blood pressure (because of ongoing treatment) to grades 1, 2, and 3 hypertension, in both subjects with aldosterone-producing adenoma and idiopathic hyperaldosteronism (Figure 3).
The accuracy of serum K+, the ARR at baseline, and the ARR after captopril administration as shown by ROC curve analysis was > 0.5, whereas that for the ratio of 24-hour urinary K+ excretion to serum K+ did not differ (Table 3). The ARR after captopril administration and the ARR at baseline performed better ( < .001) than serum K+. Table 3 shows that the optimal cutoff values for each test differed from those prespecified based on retrospective studies.9
Table 3. Results of the ROC curve analysis for the screening tests used in
ROC indicates receiver operating characteristic; AUC, area under the curve;
K+, potassium level, ARR, aldosterone/renin ratio. P value denotes significant
difference from the AUC of the identity line. The optimal cutoff was the value
that provided the highest accuracy, that is, the best trade-off between
sensitivity and specificity. CI indicates confidence interval. (Reprinted with
permission from Rossi GP, Bernini G, Caliumi C, et al. A prospective study of
the prevalence of primary aldosteronism in 1125 hypertensive patients. J Am Coll Cardiol. 2006;48:2293-2300.)
Five of the 14 centers participating in the study had adrenal vein sampling available. Adrenal vein sampling was bilaterally selective in 93% of subjects. By categorizing the centers according to whether or not they performed adrenal vein sampling, we found that significantly more aldosterone-producing adenoma and less idiopathic hyperaldosteronism were diagnosed at the centers that performed adrenal vein sampling; the opposite occurred at the centers where adrenal vein sampling was not available. Therefore, the availability of adrenal vein sampling affects the prevalence of adrenocortical pathologies underlying primary aldosteronism.
Figure 2. This graph shows that a substantial proportion of the subjects with
aldosterone-producing adenoma (APA) and idiopathic hyperaldosteronism (IHA)
did not have hypokalemia (black bar) at the time of presentation. PH indicates
primary hypertension. (Reprinted with permission from Rossi GP, Bernini G,
Caliumi C, et al. A prospective study of the prevalence of primary aldosteronism
in 1125 hypertensive patients. J Am Coll Cardiol. 2006;48:2293-2300.)
Figure 3. The proportion of subjects without primary aldosteronism (PA; white
bars), with idiopathic hyperaldosteronism (IHA; red bars), and with aldosterone-
producing adenoma (APA; black bars) in the subjects at the screening test. The
proportion of the patients with PA caused by both APA and IHA rose significantly
(from 7.2% to 19%) with increasing severity of hypertension. HT indicates
hypertension; BP, blood pressure. (Reprinted with permission from Rossi GP,
Bernini G, Caliumi C, et al. A prospective study of the prevalence of primary
aldosteronism in 1125 hypertensive patients. J Am Coll Cardiol. 2006;48:
Hyperaldosteronism carries serious cardiovascular consequences10; however, removal of an aldosterone-producing adenoma or unilateral autonomous hyperplasia corrects the hyperaldosteronism and cures or significantly decreases hypertension and the cardiovascular alterations.11 Therefore, early efforts to identify the causes of primary aldosteronism are warranted.
The PAPY study provided strong data on the prevalence of primary aldosteronism using a state-of-the-art diagnostic protocol prospectively in newly diagnosed hypertensive subjects. This new information may affect future strategies for the screening and diagnosis of primary aldosteronism and its causes.
The overall prevalence of primary aldosteronism in our study was 11.2%, which concurs with an earlier study.12 However, more importantly, an aldosterone-producing adenoma could be unequivocally diagnosed with the aforementioned strict criteria in 42% of the primary aldosteronism cases, that is, in 4.8% of the 1125 consecutive hypertensive subjects. This high rate is even more striking because the proportion of aldosterone-producing adenoma might have been underestimated for several reasons, as discussed elsewhere.7
Moreover, even though these results were obtained in subjects referred to specialized hypertension centers, they can be generalized because of the exclusion of subjects with any possibility of a secondary form of hypertension and because of the enrollment of consecutive subjects without any selection criteria (such as resistance to antihypertensive therapy, blood pressure levels, and hypokalemia). Furthermore, the anthropometric and blood pressure data (Table 1) were comparable with those of the general population of individuals with hypertension and the prevalence of a high ARR was even lower than that found in a general practitioners study.13 Thus, these results suggest the need to screen all newly diagnosed patients with hypertension for primary aldosteronism, particularly those with spontaneous or diuretic-induced hypokalemia or severe or resistant hypertension, as this diagnosis can permit a cure of arterial hypertension in several cases that otherwise would go undetected.
The ARR at baseline and after captopril administration outperformed hypokalemia and the ratio of 24-hour urinary K+ excretion to serum K+ because many subjects without primary aldosteronism had hypokalemia, and, conversely, most subjects with confirmed primary aldosteronism had normokalemia. Therefore, we could confirm that the normokalemic cases of primary aldosteronism are currently the rule rather than the exception.3 This change is probably explained by the prevalent use of low-sodium and high-potassium diets, and the widespread prescription of potassium-sparing anti-hypertensive drugs. This implies that if the condition of hypokalemia were used as an insensitive screening test, most of the primary aldosteronism would be overlooked.3
Our results showed that 45% of the aldosterone-producing adenomas were < 20 mm and 17% were < 10 mm, as previously described.14 These small tumors can be conclusively identified on CT or MRI in only half of the cases.14 Considering the high prevalence of both incidentally discovered adrenal masses, of which only 1.6% are aldosterone-producing adenoma,15 and hypertension, the chance concurrence of primary hypertension and nonfunctioning adrenal tumors is about 3.6%. Moreover, a nonfunctioning adrenal mass can be associated with a low ARR in patients with low-renin hypertension. Therefore, adrenal vein sampling is necessary to demonstrate lateralization of aldosterone secretion from the affected adrenal gland in patients with a high ARR, particularly in those with the common small-sized aldosterone-producing adenoma.5,14,16 In fact, in the PAPY study, the lack of availability of adrenal vein sampling affected the number of cases of primary aldosteronism attributed to aldosterone-producing adenoma or idiopathic hyperaldosteronism, as significantly more aldosterone-producing adenomas were diagnosed at centers where adrenal vein sampling was available. Therefore, identification of the pathology underlying primary aldosteronism requires adrenal vein sampling which, albeit technically demanding and not universally available, can be performed at a very low (0.5%-1%) risk of major complications in experienced hands.14
Results of our study showed that a surgically curable aldosterone-producing adenoma can be identified in almost 5% of newly diagnosed hypertensive patients referred to specialized hypertension centers, and this high rate might even underestimate the true prevalence of the tumors because of the lack of systematic use of adrenal vein sampling in this study. An additional 6.3% of subjects had idiopathic hyperaldosteronism, resulting in an overall prevalence of primary aldosteronism of 11.2%. Once primary aldosteronism has been identified, adrenal vein sampling should be performed to demonstrate lateralized aldosterone excess and provide an indication for adrenalectomy.
This study was supported by grants from FORICA (The Foundation for Advanced Research in Hypertension and Cardiovascular Diseases) to Gian Paolo Rossi and from the Società Italiana dell'Ipertensione Arteriosa. The secretarial assistance of Mrs. Carla Franceschin in preparing this manuscript is gratefully acknowledged.
A complete list of participating centers and PAPY investigators, is in the Appendix.
Appendix. Participating centers and Primary Aldosteronism Prevalence in
Hypertensives (PAPY) study investigators.