Aldosterone in Advancing Age
Don’t Shoot the Messenger
Article, see p 347
The study by Nanba et al, “Age-related autonomous aldosteronism,”1 in this issue of Circulation, addresses 2 areas of current major interest in cardiovascular endocrinology. The first is that of aldosterone-producing cell clusters (APCCs), distinct from a more or less continuous outer layer of aldosterone-producing cells familiar as the zona glomerulosa of the adrenal gland. The second is that of the changes in plasma renin activity (PRA) and serum aldosterone concentration with age in a group of normotensive and stage 1 hypertensive patients from the HyperPATH study (Hypertensive Pathotype), with primary aldosteronism based on current criteria excluded. The study is novel in terms not only of the numbers involved (adrenal glands from kidney donors n=127, HyperPATH cohort n=677), but also in the possibilities of linking the 2 lines of study into a coherent explanation of the changes seen in human aging. The presentation is also straightforward, crisp, and clear, with the discussion, in particular, mercifully free of padding and nit-picking: for this, and the novelty and possible implications of their findings, as well, we owe the authors our thanks.
What the study shows is that there are 2 apparently related changes in aldosterone secretion with age. In adrenal glands from young organ donors, the zona glomerulosa appears as a thin continuous outer layer, as identified by immunohistochemistry with antibody specific for the enzyme aldosterone synthase (CYP11B2). When adrenal glands from older donors are examined, the previously continuous zona glomerulosa appears disrupted (or discontinuous), progressively with increasing age. At the same time, isolated aldosterone-producing cell clusters appear more and more commonly. Whereas the factors controlling aldosterone secretion from zona glomerulosa cells are well established, those from APCCs are not; on the basis that secretion from APCCs may be constitutive, the authors suggest that this may represent age-related autonomous aldosteronism that might provide a potential explanation for age-related cardiovascular risk.
Marc Anthony’s famous opening line, “We come to bury Caesar, not to praise him,” prefaces his panegyric: there is no case for burying Nanba and his colleagues, but rather to point out (as the authors do) the limitations of their study, and to propose possible alternative implications. First, the authors acknowledge the lack of clinical and biochemical data for the adrenal donors. They point out that all their analyses are cross sectional, and therefore, however tempting, cannot determine causality. Small sample size yielded little statistical power for a comparison of findings in whites versus blacks, with no difference found, as similarly for the case of donors with or without hypertension. Additional limitations are part of the design of the HyperPATH study; eg, baseline values for PRA and serum aldosterone are not available. It is unclear whether plasma [K+] values were similarly unavailable; they may have been useful in terms of interpretation of the data.
One unfortunate issue is the repetitive usage of 2 words: normal and dysregulation. These are unwarranted value judgments; eg, if 75% of the adrenal donors aged between 40 and 49 (Figure 1) are abnormal, then we need a redefinition of normality. In this age group, there are normally many top-tier professional golfers, but a 45-year-old top-tier grand slam tennis or Nation Football League player would be very unusual; normal is clearly context dependent. In men, for example, it is normal for testosterone levels to decline markedly with age. Dysregulation is similarly a loaded term, and reflects a very narrow focus on PRA in a number of ways. Renin levels are a sensitive barometer of aldosterone action, but there is scant consideration given to more proximate actors in the renin-angiotensin-aldosterone system (eg, renin substrate, angiotensin II). This inappropriate terminology probably reflects 2 factors. One is that the current study challenges what we were all taught in medical school about the anatomy of the adrenal gland, that is, the adrenal medulla, and 3 cortical zones, glomerulosa, fasciculata, and reticularis. The second is that the rat adrenal gland is very familiar to endocrinologists; in sodium deficiency, the zona glomerulosa thickens, and in sodium repletion, it becomes thin, but in both cases it is continuous. What the present study clearly shows, as has been noted above, is that this does not apply in the human adrenal gland, in which the zona glomerulosa normally becomes progressively discontinuous with age.
In terms of the data, there are a number of points to be made. Their Figure 1G shows continuous versus discontinuous expression of CYP11B2 in the zona glomerulosa; Table 2 shows that a dichotomy with age is perhaps too simple, in that, whereas 70% of adrenal glands without an APCC identified had a continuous zona glomerulosa, in those with at least 1 discernable APCC, more than one-third (35%) still had a continuous zona glomerulosa. Note discernable; unless each adrenal gland is sectioned and examined in toto, an occasional APCC or disrupted zona glomerulosa zone may be missed. This does not negate the findings across the group; it would still appear likely that the overlap between the 2 groups in Table 2 reflects a transition state between no APCCs in youth, and with increasing discontinuity of the zona glomerulosa, and the appearance of APCCs, as well, as age increases.
The authors also showed suppression of PRA in salt loading to increase with age, whereas serum aldosterone levels were essentially unchanged, interpreted as increasing aldosteronism with age. This interpretation needs to be qualified by comparing the effects of sodium loading with those of sodium restriction, as shown in the Table, with mean and median values for serum aldosterone and PRA across quartiles of age. With sodium restriction, the fall in PRA closely parallels that in sodium loading; in contrast with sodium loading, there is a progressive fall in serum aldosterone levels with age.
In terms of aldosterone-to-renin ratio, the current screening test for primary aldosteronism, the values range from 7.1 to 10.2 across age in the sodium-loaded group, and 6.8 to 5.1 in the sodium-restricted group., Screening for primary aldosteronism is commonly on the basis of an elevated aldosterone-to-renin ratio, plus aldosterone levels that are elevated or at least in the upper levels of normal; this is not the case here. Even if screening were positive, saline suppression testing to aldosterone levels found across the age range reported here would exclude such patients from a diagnosis of aldosteronism.
The authors note, first, that aldosterone production from APCCs is suspected to be renin independent in that APCCs are observed even in normal tissue adjacent to an aldosterone-producing adenoma, where circulating renin and angiotensin II levels are suppressed. Second, many APCCs harbor somatic mutations (CACNA1D, ATP1A1; but curiously not the most common KCNJ5). In the first instance, the renin independence may reflect their bearing known (or yet to be described) somatic mutations. The finding of APCCs in normal adrenal glands staining positive for CYP11B2 is not the same as their inevitably producing aldosterone, as evidenced by their being found in the adrenal glands of healthy subjects. The authors describe aldosterone production from APCCs as autonomous, because it may be with a second-hit somatic mutation; it might be more circumspect to say that factors modulating aldosterone secretion from APCCs are yet to be described.
This leads to a consideration of the authors’ clinical perspective. The authors contend that, “This study demonstrates a progressive pattern of abnormal physiology with aging.” A more accurate sentence would be “…demonstrates a pattern of a progressively muted aldosterone response to sodium restriction with age.”
I would also argue that the statement “The study suggests that aging may be associated with a subclinical form of aldosterone excess” is not supported by the data, in that there is no evidence for aldosterone excess; in fact, the ability to secrete aldosterone declines with age (as it does for many adrenal steroids, most notably DHEA [dehydroepiandrosterone]), more noticeably in sodium restriction, and perhaps to some extent in sodium excess.
As noted above, although there does not appear to be evidence for pathophysiologic aldosterone secretion, the suggestion of mineralocorticoid receptor antagonist therapy is indicated for a different reason. In brief, mineralocorticoid receptors (MR) have equivalent high affinity for aldosterone and cortisol.2 In some circumstances, cortisol mimics aldosterone as an MR agonist: one instance is when MR, phosphorylated on serine 843 in tubular intercalated cells, which precludes steroid binding, are dephosphorylated in response to angiotensin and are activated by cortisol or aldosterone.3 Another instance is in the damaged cardiomyocyte, where MR are not protected by the enzyme 11β-hydroxysteroid dehydrogenase type 2, which metaboliizes cortisol to cortisone, which has negligible binding affinity for MR. In such circumstances, cortisol and aldosterone increase infarct size and area at risk, both antagonized by spironolactone and cortisol, with the latter unaffected by the glucocorticoid receptor/progesterone receptor antagonist mifepristone.4 Most renal tubular MR in principal cells are occupied, but normally not activated by glucocorticoid in the presence of 11β-hydroxysteroid dehydrogenase type 2.5 In the context of the progressive decline in renal function with age, and the concomitant tissue damage, it is possible that cortisol might activate a small percentage of the MR, normally occupied but not so activated. This is consistent with the muted aldosterone response to sodium restriction, and underlines the potential utility of MR antagonists in age-related hypertension, as suggested by the authors.
In brief, the changes with age in the sources of aldosterone secretion from the adrenal gland break new ground. In the young, CYP11B2 (aldosterone synthase) expression is from a continuous thin outermost cellular layer of the adrenal cortex. With age, this becomes discontinuous, and cells expressing CYP11B2 are found as scattered aldosterone-secreting cell clusters. When HyperPATH subjects were sodium loaded or restricted, renin levels decreased similarly with age: in sodium-loaded subjects, aldosterone levels did not differ between ages, whereas they were 50% lower in older than younger subjects after sodium restriction. The authors interpret these findings as evidence for age-related aldosteronism; alternative interpretations of the in vivo study, in terms of implications and causality, should be considered.
The opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.
Circulation is available at http://circ.ahajournals.org.
- © 2017 American Heart Association, Inc.
- Nanba K,
- Vaidya A,
- Williams GH,
- Zheng I,
- Else T,
- Rainey WE
- Arriza JL,
- Weinberger C,
- Cerelli G,
- Glaser TM,
- Handelin BL,
- Housman DE,
- Evans RM
- Shibata S,
- Rinehart J,
- Zhang J,
- Moeckel G,
- Castañeda-Bueno M,
- Stiegler AL,
- Boggon TJ,
- Gamba G,
- Lifton RP