原发性醛固酮增多症2015年1月临床综述

临床综述：原发性醛固酮增多症

doi: 10.1210/jc.2014-3663 J Clin Endocrinol Metab, January 2015, 100(1):1–10 jcem.endojournals.org 原发性醛固酮增多症（Primary aldosteronism，PA）是由于肾上腺皮质病变导致醛固酮分泌增多，进而导致水钠潴留、肾素－血管紧张素系统活性受抑制所导致的综合征。随着随着医学技术的发展，各种检测手段越来越精确，在高血压人群中原发性醛固酮增多症的检出率也随之增多。最近，来自澳大利亚昆士兰大学医学院的 Stowasser 教授在 Journal of Clinical Endocrinology & Metabolism 杂志发表综述，详细介绍了近年原发性醛固酮增多症的病因、诊断以及治疗的相关研究进展。

遗传学病因

2011 年，耶鲁大学医学院 Choi 博士发现 22 例来自瑞典的肾上腺醛固酮腺瘤患者的切除组织中 8 例伴有编码内向整流钾通道的 KCNJ5 基因突变。

也有报道醛固酮分泌腺瘤患者体细胞 ATP1A1、ATP2B3 、CACNA1D 基因突变，但发病率较低。

临床表现

由于醛固酮受体广泛分布于人体心脏、肾脏、神经系统、脂肪组织中，原发性醛固酮增多症患者可表现为独立于血压以外的心脑血管损害作用，如冠状动脉疾病、心肌梗死、心房颤动、脑卒中等。近来研究也发现，醛固酮的过度分泌与阻塞性睡眠呼吸障碍、胰岛素抵抗、代谢综合征、肾脏疾病及精神性疾病、骨质疏松均有一定关系。但这些影响可在对抗醛固酮治疗后在一定程度上得到逆转。

实验室筛查及确诊

血浆醛固酮 / 血浆肾素活性比值（ARR）广泛应用于筛查 PA，使得原发性醛固酮增多症的检出率明显提高，但该数值容易受到多种因素影响。

现在通常采用的确诊实验共有 4 种：生理盐水试验、口服高钠负荷试验、氟氢可的松抑制试验、卡托普利试验。

其中，氢化可的松抑制实验（FST）是最可靠的诊断实验，但该实验过程比较复杂、费用相对较高并且需要病人入院进行。Gouli 教授等更改了原来实验方法，在 FST 最后一天使用 1mg 地塞米松来消除任何内源性因素造成的促肾上腺皮质激素增多诱发的醛固酮升高，并且实验结果已经得到认可。

亚型分型及其意义

原发性醛固酮增多症常见的亚型为醛固酮瘤和双侧特发性醛固酮增多症，少见亚型见于主要由一侧肾上腺球状带增生所致单侧增生。

分型后进行诊断治疗对治疗效果以及患者今后的生活质量具有重要影响。如，单侧原发性醛固酮增多症患者行单侧肾上腺切除对于心血管以及生活质量的改善均优于使用药物治疗的双侧肾上腺醛固酮分泌增多。

治疗

1、安体舒通（螺内酯）是目前常用的醛固酮拮抗剂，是内科治疗的主要手段。但它是非选择性醛固酮受体拮抗剂，也可作用于性激素受体并导致男性乳腺发育，女性月经紊乱等不良反应。

2、依普利酮是螺内酯的衍生物，可选择性作用于醛固酮受体而避免了上述不良反应。

3、对于肾上腺醛固酮腺瘤患者，一侧肾上腺切除术是最优的手术方式。

腹腔镜下肾上腺切除术是一种理想的手术方式，其具有住院时间短和长期死亡率低的优点，但对手术者的要求也较高。这些方法均可使血压明显下降。

4、无论是动物还是人体内，内源性的醛固酮生成以及钠盐饮食摄入对于器官的损害是相互影响的。增加饮食中钠盐摄入会增加呼吸睡眠暂停综合征的发生，因此限制钠盐摄入对于较少原发性醛固酮增多症患者的死亡率和致残率具有重要意义。

总之，应在高血压人群中采用血浆醛固酮 / 血浆肾素活性比值来更加广泛地筛查, 确定为原发性醛固酮增多症者需行体位试验或影像学检查，必要时作选择性肾上腺静脉激素检测以明确其亚型分型，以指导治疗。

Update in Primary Aldosteronism

Michael Stowasser

Endocrine Hypertension Research Center, University of Queensland School of Medicine, Greenslopes and

Princess Alexandra Hospitals, Brisbane 4102, Australia Primary aldosteronism (PA) is a condition well worth detecting because it is a common cause of

hypertension and is associated with excessive morbidity for the degree of hypertension and reduced

quality of life, all of which can be abrogated with specific surgical or medical treatment.

Recent years have seen an explosion in knowledge concerning the genetic bases of this disorder,

andparticularly of somatic mutations associated with aldosterone-producingadenomasandgermline

mutations causing rare familial forms, both involving genes encoding ion channels. Inroads

have also been made into understanding molecular pathways that may be involved in the development

of PA. With evidence continuing to mount for non-blood pressure-dependent adverse

effects of aldosterone excess and for superior effects of specific over non-specific treatment, the

need for accurate yet readily applicable and available diagnostic approaches and methodologies

has become a matter of urgency. Advances in approaches to confirmatory testing, subtype differentiation,

and assay methodology are helping to improve feasibility and reliability of the diagnostic

workup for PA, and new treatment approaches are emerging. (J Clin Endocrinol Metab

100: 1–10, 2015)

Primary aldosteronism (PA) remains in a “golden era” in terms of clinical and research interest that has evolved over the past two to three decades with the realization

that it is highly prevalent among hypertensive populations and, through non-blood pressure-dependent

means, is associated with cardiovascular, renal, and psychological

morbidity that appears to be excessive for the

degree of hypertension. The last few years have seen ongoing, intense research activity into its causations, diagnostic workup, and management that appears unlikely to

wane anytime soon. Perhaps the most exciting advances have been in understanding the pathophysiology of PA, and especially its genetic bases, opening doors to newer and better methods of diagnosis and treatment. Evidence that the adverse effects of aldosterone excess go way beyond

merely causing hypertension has been further

strengthened. Challenges to physicians keen to pursue the diagnosis of this condition, correctly determine the subtype, and offer optimal specific treatment have received

considerable attention and, in some cases, have been met by the development of better diagnostic methodologies, more accurate assays, and new therapies. This report will review some of these advances.

The New Genetics of PA

It would be impossible to complete an update on PA without including the major recent developments that have

been made in understanding the genetic bases of this disorder. In 2011, Choi et al (1) reported somatic mutations (G151R and L168R) in KCNJ5, encoding an inwardly

rectifying potassium channel (Kir 3.4), in eight of 22 aldosterone-

producing adenomas (APAs) removed from

Swedish patients. In the same report, a third germline mutation

(T158A) in KCNJ5 was identified in an American

family with familial, florid PA with severe childhood-onset hypertension, hypokalemia, and markedly elevated al- ISSN Print 0021-972X ISSN Online 1945-7197

Printed in U.S.A.

Received September 27, 2014. Accepted October 27, 2014. First Published Online November 3, 2014

Abbreviations: APA, aldosterone-producing adenoma; ARR, aldosterone/renin ratio;

AT1RAA, agonistic autoantibodies against angiotensin-II type-1 receptor; AVS, adrenal

venous sampling; BAH, bilateral adrenal hyperplasia; CT, computed tomography; FST,

fludrocortisone suppression testing; miRNA, microRNA; mTOR, mammalian target of rapamycin;

NT-proBNP, N-terminal probrain natriuretic peptide; PA, primary aldosteronism;

RSST, recumbent SST; SSST, seated SST; SST, saline suppression testing; TASK-2, TWIKrelated

acid-sensitive K channel 2.

S P E C I A L F E A T U R E

U p d a t e

doi: 10.1210/jc.2014-3663 J Clin Endocrinol Metab, January 2015, 100(1):1–10 jcem.endojournals.org 1

dosterone and hybrid steroid (18-hydroxy- and 18- oxo-cortisol) levels, which, unlike in

glucocorticoid-remediable

aldosteronism (familial hyperaldosteronism type I), was not glucocorticoid-suppressible. Resected adrenals (the hypertension could not be controlled with medications, including spironolactone) were grossly enlarged and

showed marked, diffuse hyperplasia of the zona fasciculata

(2). Subsequently, several groups confirmed the presence of somatic KCNJ5 mutations with a prevalence of

approximately 40% of APAs among Caucasian patients

(3– 6) and even higher in Japanese patients (7). Patients with APAs bearing these mutations have shown a female

predominance and youngermeanage than those with nonmutated tumors (1, 3–8), with a tendency to more severe

PA (1, 5–7), unresponsiveness of plasma aldosterone to

upright posture (4), and larger tumors (3, 4) composed predominantly of cells resembling zona fasciculata (9). In vitro, reported KCNJ5 mutations are almost always

located within the selectivity filter and associated with loss

of selectivity of the mutated Kir 3.4 channels to potassium, with enhanced sodium conductance predisposing to cell

membrane depolarization, influx of calcium, and ultimately stimulation of aldosterone synthesis (1, 3, 6,

9–11). How they lead to adrenal cell proliferation and tumor development remains less well understood and a fascinating area of ongoing research (12).

Among reported families with germline KCNJ5 mutations, severity of phenotype has varied considerably from

mild PA readily treated with spironolactone with adrenals of apparently normal morphology on computed tomography (CT) to very florid PA requiring bilateral adrenalectomy yielding massively enlarged adrenals, and appears

to be at least partly related to the KCNJ5 mutation genotype (1, 2, 10, 13–15). Surprisingly, G151E-mutated

channels, which tend to be associated with a milder clinical phenotype, showed markedly higher (more abnormal) sodium conductance and conferred greater lethality to transfected HEK 293T cells than G151R-mutated channels associated with a much more florid phenotype. This increased lethality may explain the lack of adrenal hyperplasia

occurring clinically in patients bearing germline G151E (vs G151R) mutations (14).

In a study of 251 Caucasian subjects with PA, we identified three heterozygous missense mutations (R52H,

E246K, andG247R)and found that 12 subjects(5%of the cohort) were carriers for the rare nonsynonymous single nucleotide polymorphism rs7102584 causingE282Qsubstitution

(16). Unlike previously described mutations associated with PA, these three mutations and rare polymorphism were all remote from the potassium channel

selectivity filter. In vitro studies showed R52H, E246K, and E282Q (but not G247) substitutions to be functional,

affecting the inward rectification, the ability of the Kir

3.4

channels to conduct Na currents, and angiotensin II-induced aldosterone release from the H295R cell line (16). Hence, germline variation in the KCNJ5 gene may have a role to play in the common sporadic form as well as the much rarer syndromic forms of PA.

Somatic mutations within APAs have since also been

described in ATP1A1 (encoding the -subunit of Na/K ATPase), ATP2B3 (encodes aCa2 ATPase calcium channel), and CACNA1D (encodes a voltage-gated calcium

channel), but with much lower frequency (8, 17–20). As with KCNJ5 mutations, in vitro studies suggest that these mutations also lead to zona glomerulosa cell membrane depolarization, resulting in increased aldosterone production

(17–20). Unlike patients with KCNJ5-mutated APAs, however, those with APAs carrying the newer mutations have tended to be more commonly males (17, 19), with

smaller APAs (8, 19) containing predominantly zona glomerulosa-

like cells (19).

Pathophysiology

Several groups of investigators have recently begun to shed some light on the molecular pathways that may be involved in the development of PA. Lenzini et al (21) reported blunted expression of the TWIK-related acid-sensitiveK channel 2 (TASK-2) gene inAPAscompared with

normal adrenal cortex. H295R adrenocortical cells transfected

with a TASK-2 dominant-negative mutant construct

showed increased aldosterone production and expression of the genes encoding aldosterone synthase

(CYP11B2) and the steroidogenic acute regulatory protein. Two microRNAs (miRNAs), hsa-miR-23 and hsamiR-

34, the expression of which was shown to negatively correlate with that of TASK-2 in APAs, were found to

decrease TASK-2 expression by binding to the 3 untranslated region of the TASK-2 gene. These data suggest that

lower expression of the TASK-2 channel in APA cells, perhaps induced by miRNAs, could contribute to excessive aldosterone secretion in human PA.

Robertson et al (22) investigated further the role of miRNA in APA development. Knockdown of Dicer1, a key enzyme inmiRNAmaturation, significantly enhanced CYP11B1 and CYP11B2 expression in H295R cells, suggesting a regulatory role for Dicer-dependent miRNAs. In

vitro manipulation of one of these (miR-24, which was down-regulated in APA vs normal human adrenal cortex) confirmed its ability to modulate CYP11B1 and CYP11B2

expression, as well as cortisol and aldosterone production. If specificity to CYP11B2 could be conferred, adrenal

2 Stowasser Update in Primary Aldosteronism J Clin Endocrinol Metab, January 2015, 100(1):1–10

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 January 2015. at 21:39 For personal use only. No other uses without permission. . All rights reserved.

miRNAmay therefore be a therapeutic target for the treatment

of PA.

Berthon et al (23) reported constitutive activation of WNT/-catenin signaling, possibly due to decreased expression

of the WNT inhibitor SFRP2, to be present in

70% of APAs and postulated that this may be an important step in APA development. In support of this, mice

lacking Sfrp2 demonstrated increased aldosterone production and ectopic differentiation of zona glomerulosa

cells. Furthermore, in vitro studies inH295Rcells revealed a role for-catenin to play in the control of both basal and angiotensin II-induced aldosterone secretion by activating AT1R, CYP21, and CYP11B2 transcription. The authors concluded that aberrant WNT/

associated

with APA development and that the WNT pathway

may be worth considering as another therapeutic target in PA.

The role of the phosphoinositide 3-kinase/protein kinaseB( or AKT)/mammalian target of rapamycin(mTOR) -catenin activation is

signaling pathway (involved in tumor development and

metastasis in adrenocortical carcinoma, pheochromocytoma, and many other human tumors) in the pathogenesis

ofPAwas assessed by Su et al (24). Compared with normal zona glomerulosa, the levels of phospho-AKT, phosphomTOR, phospho-S6, and vascular endothelial growth factor

were significantly up-regulated in APA and hyperplastic adrenal cortex from patients with PA. No significant

differences in the expression of these proteins were found betweenAPAand hyperplasia. ThemTORinhibitor rapamycin inhibited both cell proliferation and aldosterone production by H295R cells. The authors concluded that the phosphoinositide 3-kinase/AKT/mTOR signaling pathway may mediate aldosterone hypersecretion and contribute to the development of PA.

Two groups have recently studied the potential role for agonistic autoantibodies against the angiotensin-II type-1 receptor (AT1RAA) in PA. Rossitto et al (25) found titers of AT1RAA to be higher in both PA patients (n 46; 26 with APA, 20 with bilateral PA) and essential hypertensive

patients (n 62) than in normotensive subjects (n 45). The titer was higher in patients with APA than in bilateral PA and appeared to allow discrimination of these two PA subtypes by receiver operator characteristics curve analysis.

Plasma aldosterone concentrations fell more in

AT1RAA-positive than AT1RAA-negative PA patients in

response to captopril, in keeping with an agonistic role for these autoantibodies. In a smaller but more detailed analysis by Kem et al (26), sera from each of 13 patients with PA(three APA, two bilateral, and eight with indeterminate subtype) significantly increased angiotensin-II type-1 receptor

(AT1R) activation in AT1R-transfected HAC15

adrenocortical cells compared with 20 control subjects, and this activity was inhibited by the selective AT1R blocker losartan. Both sera and IgG purified from

AT1RAA-positive sera demonstrated vasoconstrictive effects in isolated rat cremaster arterioles (again blocked by

losartan), and the IgG stimulated basal and

angiotensininduced

aldosterone production by HAC15 cells (blocked

by candesartan). These suggest the possibility that, at least in some individuals, the acquisition of antibodies directed against and capable of activating AT1R may contribute to the development of PA.

Non-Blood Pressure-Dependent Adverse

Effects of Aldosterone Excess: Evidence

Continues to Mount

Numerous published studies have demonstrated aldosterone excess to have adverse cardiovascular and renal effects that are at least partly independent of its effects on blood pressure and that appear to be reversible with specific treatment directed against aldosterone excess (27–

32). Savard et al (33) compared the prevalence of cardiovascular

events in 459 patients with PA and 1290 essential

hypertensives matched for sex, age, and office systolic

blood pressure. Patients withPAhad a 2-fold greater prevalence

of left ventricular hypertrophy (even after adjustment for hypertension duration) and significantly higher prevalence rates for coronary artery disease, nonfatal myocardial infarction, heart failure, and atrial fibrillation.

Mulatero et al (34) reported a higher rate of cardiovascular events (and especially strokes and arrhythmias) among

patients with PA (n 270) than among essential hypertensives (n810) matched 1:3 in a case control fashion for

blood pressure levels and several other risk factors. This was the case both for patients with APA and those with bilateral adrenal hyperplasia (BAH). Patients with PA

were also more likely to develop type II diabetes during a 12-year period of follow-up after diagnosis. In this context, Fischer et al (35) reported mean glucose-induced first phase insulin secretion to be reduced in a group of 22 patients with PA, compared to that in 11 patients with essential hypertension, and to improve in nine with APA

after unilateral adrenalectomy, suggesting a negative effect of aldosterone excess on pancreatic -cell function.

Su et al (36) reported magnetic resonance imaging evidence of increased myocardial fibrosis and stiffening

among 25 patients with PA vs 12 age-matched normotensive controls. Using magnetic resonance imaging to measure aortic distensibility, Mark et al (37) reported that PA patients (n 14) displayed increased arterial stiffness

33). compared with matched essential hypertensives (n

Lin et al (38) found plasma procollagen levels to be higher doi: 10.1210/jc.2014-3663 jcem.endojournals.org 3

and cyclical variation of integrated backscatter (a marker of left ventricular systolic function) on echocardiography to be lower among 20 patients with APA compared to 20 with essential hypertension. Parameters improved in the APA group after adrenalectomy. These investigators also

described higher carotid intima-medial thicknesses on duplex ultrasound scanning and higher arterial pulse wave

velocities in the patients with APA, again improving after surgery (39). In a multivariate regression analysis of 54 patients with PA followed for a mean of 6.4 years after unilateral adrenalectomy or commencement of spironolactone, Catena et al (40) found that the degree of reduction in left ventricular mass was independently predicted

not only by change in systolic blood pressure but also by pretreatment plasma aldosterone levels, further supporting a role for aldosterone in promoting left ventricular

hypertrophy that is independent of the hypertension-related hemodynamic load.

Detection and Diagnosis of PA: What’s

New?

Although still regarded as the most reliable available screening test for PA, the plasma aldosterone/renin ratio (ARR) can be affected, and results thereby can be confounded by many factors (41), indicating a need for improved approaches. Pizzolo et al (42) reported on the use

of N-terminal probrain natriuretic peptide (NT-proBNP) measurement as an adjunct to the ARR. Among 132 hypertensive subjects who underwent ARR testing, 81 with

highARRunderwent iv saline infusion suppression testing (SST). NT-proBNP levels correlated positively with ARR and negatively with renin level, were higher in patients with high ARR and in those with positive SST, and were an independent predictor of positive SST. The proportion of patients with a positive SST ranged from only 23% in females with a low NT-proBNP to 93% in males with a high NT-proBNP, leading the authors to conclude that NT-proBNP may prove useful to identify which patients

with a high ARR should receive formal confirmatory testing

(42).

In patients who screen positive for PA, confirmatory testing is necessary to definitely confirm the diagnosis before

proceeding to adrenal venous sampling (AVS) to differentiate unilateral, surgically correctable forms of PA

from bilateral forms usually treated medically. Four confirmatory

tests are commonly used, namely, oral sodium

loading, saline infusion, fludrocortisone administration with oral sodium loading, and captopril challenge testing. Of these, fludrocortisone suppression testing (FST) has been considered the most reliable (43), but it is cumbersome, difficult to perform, and relatively expensive, requiring hospital admission for several days. Gouli et al

(44) recently reported on a modified form of the test in which 1 mg of dexamethasone was administered on the

last day of the FST to eliminate any stimulatory effect of endogenous stress-induced adrenocorticotropin on aldosterone

secretion, a recognized confounder of test results

(43). Westerdahl et al (45) recently reported captopril challenge testing to have poor discriminatory power for PA in hypertensive patients with high basal ARR. We previously found SST, performed in the recumbent position,

to lack sensitivity for PA, missing most patients (including

those with APA) who tested positive by FST (46). Willenberg et al (47) similarly found SST to lead to greater suppression of plasma aldosterone than FST. Because aldosterone levels can be higher upright (eg, seated) than

recumbent in patients with PA and upright levels are used during FST, we hypothesized that seated SST (SSST) is more sensitive than recumbent SST (RSST), especially for posture-responsive PA. In a pilot study of 31 patients who underwent FST (upright plasma aldosterone levels measured at 10 AM basally and after 4 d of fludrocortisone 0.1 mg every 6 h and oral salt loading) and SST (aldosterone levels measured basally at 8 AM and after infusion of 2 L normal saline over 4 h), both recumbent and seated in

randomized order, FST confirmed PA in 23 patients, excluded PA in three, and was originally “inconclusive” in

five (48). However, one patient with inconclusive FST had PA confirmed by lateralizing AVS and was reclassified “unilateral PA.” Of the 24 with confirmed PA (eight unilateral,

11 bilateral, and five undetermined subtype), 23

(96%) tested positive by SSST compared to only eight (33%) by RSST (P .001) (Table 1). RSST missed one

unilateral, all 11 bilateral, and four with as yet undeter- Table 1. Comparison of Results of SSST vs RSST in 24 Patients With Confirmed PA

SSST

Positive

RSST

Positive

Total patients with PA 24 23 (96)** 8 (33)

Unilateral PA 8 8 (100) 7 (88)

Bilateral PA 11 10 (91)** 0 (0)

PA subtype yet

to be determined

5 5 (100)* 1 (20)

Posture-responsive 14 13 (93)** 1 (7)

Posture-unresponsive 10 10 (100) 7 (70)

Data are expressed as number or number (percentage). Postureresponsive

denotes plasma aldosterone responsive to upright posture (levels measured at 1000 AM after 3 hours of upright posture at least

50% higher than those measured basally at 7 AM after overnight recumbency). Posture-unresponsive denotes plasma aldosterone

unresponsive to upright posture (upright levels less than 50% higher

or lower than recumbent).

* P .01; and ** P .001 vs RSST.

4 Stowasser Update in Primary Aldosteronism J Clin Endocrinol Metab, January 2015, 100(1):1–10

mined subtype. RSST was positive in seven of 10 (70%)

posture-unresponsive vs one of 14 (7.1%) posture-responsive patients (P .005). Clearly, larger studies are required, but these preliminary results suggest that SSST

may be superior to RSST in terms of sensitivity for detecting PA, especially posture-responsive forms, and may represent a reliable, yet much simpler, cheaper, and quicker alternative to FST.

Subtype Differentiation

Currently, the only reliable way to differentiate unilateral (mainly APA) from bilateral PA and to lateralize APAs preoperatively is by adrenal venous sampling (AVS).

However, achieving high success rates for cannulation of the adrenal veins (and especially the right, which is more difficult to locate) represents a challenge to institutions offering AVS. Recent studies have demonstrated that limiting the number of proceduralists at each institution to one or two and providing them with a higher throughput of procedures probably have the most profound effects on improving cannulation success rates (judged by adrenal/ peripheral venous cortisol gradients) (49, 50). Rapid, point-of-care cortisol assays can alert the proceduralist, at

the time of AVS, whether adrenal venous blood has been

successfully obtained, and have also resulted in higher success

rates in centers using this technique (51–53). We previously reported on the use of CT to localize the right

adrenal vein before AVS (46, 54). The use of CT during angiography to determine correctness of catheter placement was recently reported by two groups to lead to catheter repositioning in approximately 10% of patients and

to be associated with achievement of high (90%) rates of successful cannulation (55, 56). Improved success rates have also been reported with the use ofACTHstimulation during AVS (57, 58). Presumably, at least in some of these instances, cannulation had already been achieved but was not confirmed by cortisol gradients because of a transient lack of cortisol production at the time of sampling. A concern about ACTH stimulation has been that, in a patient with APA, ACTH could lead to stimulation of the contralateral gland, causing results to suggest bilateral,

rather than unilateral, disease. In the study by

Monticone et al (58), the use of more stringent (for example, adrenal/peripheral venous cortisol gradients of 3.0 basally and 4.0 post-ACTH) criteria for

cannulation proved critical for optimizing pre- vs post- ACTH diagnostic reproducibility.

Traditional AVS provides information about the relative degree of aldosterone production by an entire adrenal gland vs the other gland. In a recent advance, Japanese workers have reported on the use of superselective AVS that has the ability to sample different branches of the adrenal vein and thereby compare aldosterone production by different regions within a gland (59, 60). In this way, it

has been possible to localize small aldosterone-producing lesions with sufficient resolution to permit their removal by partial, rather than total unilateral, adrenalectomy (59,

60). Although this may only seem clinically relevant for a minority of patients withPA(for example, those suspected as having bilateral APAs, or who have only one functioning adrenal that may contain an APA, who otherwise

would require steroid replacement therapy postoperatively), it nevertheless represents a promising new approach with potentially wider future applications.

Once considered a rarity, autonomous tumorous cosecretion of cortisol has been recently reported by two

groups to be relatively common (10%) among patients

with APA, even in the absence of obvious clinical features of Cushing syndrome (“subclinical Cushing syndrome”) (61, 62). This has implications not only for perioperative management with the need, at least in some patients, to consider glucocorticoid replacement, but also for interpretation

of AVS results. Cosecretion of cortisol would be

expected to raise cortisol levels in the adrenal vein draining theAPA(thus lowering the ipsilateral aldosterone/cortisol ratio), while causing chronic suppression of ACTH and thus lowering of cortisol secretion by the contralateral gland (thus raising the contralateral aldosterone/cortisol ratio and lowering the adrenal/peripheral venous cortisol

ratio). This could result in loss of lateralization (and a lost

opportunity to offer potentially curative surgery) or give the impression that cannulation had failed on the contralateral

side (and thereby lead to inappropriate consideration for repeat AVS). One potential solution to this

problem is the measurement of an alternative, non- ACTH-dependent adrenal hormone to cortisol. Dekkers et al (63), for example, found measurement of plasma metanephrine during AVS to be superior to measuring

cortisol in assessing success of cannulation, with adrenal/ peripheral ratios 6-fold higher among procedures deemed to be successful according to cortisol criteria. Baba et al

(64) reported catecholamine levels to be useful for judging not only cannulation success but also lateralization. The search for reliable alternative approaches that are less costly, invasive, and technically demanding than AVS has been ongoing in recent years but has met with limited success.Aclinical prediction score developed by Küpers et

al (65), who reported the presence of a unilateral mass lesion on CT “typical” for APA plus a serum potassium of3.5 mmol/L or an estimated glomerular filtration rate of at least 100 mL/min/1.73 m2 to have 100% specificity (but only 53% sensitivity) for unilateral PA, was found to doi: 10.1210/jc.2014-3663 jcem.endojournals.org 5

be much less reliable when applied to patients within the German Conn’s Registry (66), except in those under 40 years of age. Sze et al (67) also found the clinical prediction score to have a lower specificity (88%) among aUKcohort of patients with PA and were unable to show superiority over an imaging-based strategy. Other more promising new approaches for differentiatingAPAfrom BAH, which require validation in different centers and with larger numbers, have included the use of semiquantification of 131I-6-iodomethyl-norcholesterol (NP-59) single photon

emission CT/CT (68) and 11C-metomidate positron emission tomography/CT (69).

Studies attempting to assess (including by receiver operator characteristics analysis) and validate approaches to distinguishing unilateral from bilateral PA have been challenged

by the lack of a robust “gold standard” for diagnosing APA. The blood pressure response to surgery, for

example, may be incomplete if the patient is studied too soon after adrenalectomy or has other causes of residual hypertension. The response to surgery of plasma aldosterone plus or minus renin levels and the ARR is better, but

it is still imperfect because cutoff criteria are arbitrary to

some extent and the ARR is associated with false positives and negatives. Our own approach is to repeat the FST

postoperatively (70), but, as mentioned above, this is time consuming, cumbersome, and expensive to perform.

Demonstration of an adenoma in surgically removed adrenals does not prove that it was actually producing aldosterone,

and reliance on this would have the potential to

misclassify unilateral PA due to micronodular hyperplasia as bilateral PA. Recent studies have reported on the use of immunohistochemical staining of resected adrenals in an attempt to confirm APA removal (71–75). Using

isoformspecific

monoclonal antibodies against the 3-hydroxysteroid dehydrogenase/isomerase family, Doi et al (71) reported hyperplasia of zona glomerulosa seen in bilateral PA to be accompanied by a robust expression of the HSD3B1 isoform. In contrast, tumor cells in APA were immunonegative to HSD3B1 but showed expression of HSD3B2. Volpe et al (72) found APAs to show strong immunoreactivity for aldosterone synthase. In adrenals from a significant minority of patients, they found either “adenomas” (presumably defined as adrenocortical tumors of at least 1 cmin diameter) that were not CYP11B2

positive or no adenoma at all, but they found smaller nodules with strong CYB11B2 immunoreactivity, indicating

aldosterone-producing nodular hyperplasia. Nanba et al

(73), in a study of 32 patients with PA who underwent unilateral adrenalectomy, found 22 to show positive CYP11B2 immunostaining in their tumors (designated APAs) and eight with either CYP11B2-negative adenomas

(n 4) or without tumors on CT to show aldosteroneproducing cell clusters with CYP11B2 immunostaining in

the zona glomerulosa. In APAs, CYP11B2 score adjusted for tumor volume was positively correlated with plasma aldosterone and negatively with serum potassium. APAs with either similar or greater amounts of CYP11B1 (11hydroxylase, which catalyzes the last step of cortisol biosynthesis)

to CYP11B2 staining showed significantly

higher serum cortisol after 1 mg dexamethasone and larger tumor size than the CYP11B2-dominant APA

group. High-quality antibodies produced in the laboratory of Celso Gomez-Sanchez (University of Mississippi)

have also been used to demonstrate correlations between the degree of expression of CYP11B2 in APAs and preoperative plasma (and urinary) aldosterone levels (74). In -

addition, this group has shown expression levels to be similar in zona glomerulosa within normal adrenals and in those removed from patients with PA due to BAH (and cited increased expression of upstream enzymes to be a possible explanation for the greater aldosterone production in the latter group), but lower in that in zona glomerulosa adjacent to an APA (75), consistent with suppression of aldosterone production.

Importance of Subtype Differentiation

For both cardiovascular outcomes (76–78) and quality of life (79, 80), beneficial effects of surgery in unilateral PA

have been reported to be superior (more marked and/or

more rapid) to those of medical therapy in bilateral PA. In a prospective study of 54 patients with PA who were treated by adrenalectomy (n24) or with spironolactone (n 30), Catena et al (76) reported left ventricular mass to have decreased significantly by 1 year of follow-up only in adrenalectomized patients. Although overall change

from baseline to the end of a mean of 6.4 years of follow-

up was comparable in the two groups, there remains

the possibility that the apparent delay in benefit afforded by medical (over surgical) treatment could translate to less impressive reductions in cardiovascular morbidity and mortality. A nationwide epidemiological study conducted in Japan, which analyzed data from 1706 patients with PA, found that among those with APA, surgical, but not medical, treatment was significantly associated with amelioration

of hypertension (81). By contrast, among patients with hyperplastic forms of PA (mostly BAH), there

was no relationship between either surgical or medical treatment and the prognosis of hypertension. In an observational

study reported by Reincke et al (82) on 300 patients with PA, adrenalectomy (presumably mostly for

unilateral PA) was associated with reduced all-cause mortality

when compared with medical treatment.

6 Stowasser Update in Primary Aldosteronism J Clin Endocrinol Metab, January 2015, 100(1):1–10

Toward High-Quality Assay Development

Screening, confirmation, and subtype differentiation of PA are all dependent on accurate measurement of plasma aldosterone, and because theARRis predominantly

renindependent,

it is critical that the methods used to assay

aldosterone and renin in clinical samples are of high quality. The recent push for using automated platforms and immunometric methodology has been primarily on the

basis of faster throughput and lower cost and potentially at the expense of lower reliability. In 2009, we reported on the development of a high-throughput, highly accurate, and reproducible method of measuring plasma aldosterone using HPLC and tandem mass spectrometry (LCMS/

MS) (83). Since then, several other investigators have established similar assays for aldosterone (84, 85), and at least two groups have developed assays that use LCMS/ MS to measure angiotensin I and thereby determine

plasma renin activity (86, 87). The hope for incorporation of tandem mass spectrometry (LC-MS/MS) technology

into routine clinical laboratory practice for measuring aldosterone

and renin (or alternatively angiotensin II, with

implications for the development of aldosterone/angiotensin II ratio testing as a new and possibly superior

screening approach) now appears to be well within reach. New Approaches to Treatment

The mainstays of treatment for PA continue to be unilateral adrenalectomy for most patients with unilateral forms and medications that antagonize aldosterone action (for example, spironolactone 12.50–50 mg daily, eplerenone 25–100 mg daily, or amiloride 2.5–20 mg daily) for most with bilateral PA. Both treatment approaches can result in marked blood pressure lowering in patients with hypertension,

including those with resistant forms.

Recent surgical developments have included reports of single incision laparoscopic adrenalectomy (88, 89),

which offers superior cosmetic results to multiple incision approaches but requires further experience to establish its value from a risk vs benefit perspective.

On the medical front, a randomized, open-label,

blinded end-point study involving 34 patients with BAH reported in 2009 by Karagiannis et al (90) found eplerenone and spironolactone to be of similar efficacy in reducing blood pressure. Conversely, a more recent randomized, double-blinded trial comparing the efficacy, safety, and tolerability of eplerenone to that of spironolactone (given in relatively large daily doses of 100–300 and 75–

225 mg, respectively) in patients with PA found

spironolactone

to be superior in terms of blood pressure lowering,

but to be associated, as expected, with higher rates of male gynecomastia (21 vs 5% for eplerenone) and female mastodynia

(21 vs 0%) (91). Aldosterone synthase inhibitors

remain a promising treatment approach, although a recent study showed that in patients with PA, the effects of eplerenone

treatment (50–100 mg twice daily) on blood pressure, plasma potassium, and renin were more marked than

those of 4 weeks of treatment with the aldosterone synthase inhibitor LCI699 (0.5–1 mg twice daily) (92). An

ongoing concern for these new drugs is the potential for inhibiting cortisol production in addition to that of aldosterone.

Nonsteroidal dihydropyridine-based mineralocorticoid receptor antagonists (93, 94) are a new drug class that has displayed similar in vitro potency to spironolactone, without apparent effects on androgen and progesterone receptors, and therefore presumably without the same risk of sex steroid-related side effects.BAY94–8862, the firstin-

class of this new generation of drugs, is currently in development for treatment of heart failure (93), with early

results showing considerable promise (95), and studies in PA are eagerly awaited.

In both animals and humans, endogenous aldosterone and dietary salt intake appear to interact to induce targetorgan

deterioration. Pimenta et al demonstrated that both

left ventricular mass (96) and degree of proteinuria (97) in patients with PA were dependent on 24-hour sodium excretion rate (as a marker of dietary salt intake). Increased dietary sodium and aldosterone were also shown by these workers to be related to severity of obstructive sleep apnea in patients with resistant hypertension (98). Incorporation of dietary salt restriction in the management of patients withPAmay therefore help to limit and even reverse target organ damage and morbidity.

Conclusions

In an exciting time for research into PA, major inroads have been made into understanding its pathophysiology, including its underlying genetic bases, with the potential for development ofnewtreatment modalities. Advances in

approaches to confirmatory testing, subtype differentiation, assay technology, and treatment are helping to enable treating physicians to optimize diagnosis and management for patients with this important condition, with

the promises of reduced morbidity and mortality and improved quality of life.

Acknowledgments

Address all correspondence and requests for reprints to: Michael

Stowasser, Hypertension Unit, University of Queensdoi: 10.1210/jc.2014-3663 jcem.endojournals.org 7

land School of Medicine, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, Brisbane 4102, Australia.

E-mail: [email protected].

Work performed within the author’s laboratory described herein was funded by The National Health and Medical Research

council of Austrialia, The PA Research Support Scheme, and the Irene Patricia Hunt Memorial Trust for Hypertension Research.

Disclosure Summary: The author has nothing to disclose.