Journal of Human (2002) 16, 159–162  2002 Nature Publishing Group All rights reserved 0950-9240/02 $25.00 www.nature.com/jhh REVIEW ARTICLE , a common entity? the myth persists

PL Padfield Department of Medical Sciences, Western General Hospital, University of Edinburgh, Scotland, UK

Primary excess or is aldosterone is in fact inappropriately elevated if the an important cause of hypertension which, when asso- level is low. A single measurement of the ratio of ciated with an aldosterone secreting , is amen- aldosterone to renin levels is claimed to be highly pre- able to surgical cure. The biochemical hallmarks of the dictive of patients who will have primary aldosterone condition are a relative excess of aldosterone pro- excess. This paper examines the logic behind such duction with suppression of plasma levels of renin (a claims and presents evidence from the literature that an proxy for II, the major trophic substance abnormal ratio is simply a different description of the regulating aldosterone secretion). This combination of low renin state and that such patients do not necessarily a high aldosterone and a low renin is however more have mineralocorticoid hypertension. Most patients ‘dis- commonly associated with ‘nodular hyperplasia’ of the covered’ by this test will have what many call low-renin adrenal glands, a condition not improved by surgery hypertension, a condition not amenable to specific ther- and variably responsive to the effects of the mineral- apy. Claims that they are peculiarly sensitive to the ocorticoid antagonist, . Until recently the hypotensive effects of spironolactone have not been prevalence of either form of tested in controlled trials. The test would however be has been thought to be low such that few clinicians expected to pick up those individuals with true Conn’s ‘hunted’ for it in the absence of hypokalaemia (the tra- syndrome but such patients remain too few in number ditional clue for the syndrome). This view has been chal- to justify widespread use of an expensive screening lenged, firstly by the realisation that no more than 50% test. of such patients will have a low plasma and Journal of Human Hypertension (2002) 16, 159–162. DOI: secondly by the assumption that a ‘normal’ plasma 10.1038/sj/jhh/1001321

Keywords: primary aldosterone excess; Conn’s syndrome; low-renin; hyperaldosteronism; aldosterone/renin ratio

Introduction single causative factor, the removal of which will cure hypertension. I have moved from a time when I have lived and worked throughout a quarter of a we investigated everybody1 to one where I investi- century where clinical hypertension research has gate few and I am not persuaded that this is the been both productive and exciting. This has been no wrong approach. more so that in the area of mineralocorticoid hyper- tension. Fashions change however and I well remember Background starting out in hypertension research in a unit where On 29 October 1954 Jerome Conn gave his presiden- every patient had a full assessment of the renin- tial address at the 27th Annual Meeting of the Cen- angiotensin-aldosterone system. This was only tral Society for Clinical Research in the United partly for research purposes as there was a genuine States of America.2 He described a patient with belief that this system was intimately concerned in hypertension who was cured by the removal of an the pathogenesis of high blood pressure. As the adrenal tumour. I was privileged to work in the years have passed my own philosophical approach institution where he investigated the index patient, to the hypertensive patient has changed. I no longer a middle-aged female, for many weeks before finally believe that most hypertensive patients harbour a persuading a surgeon that she must have an adrenal tumour and that he (the surgeon) had but to find it. Correspondence: Dr PL Padfield, Department of Medical It is worth remembering that Conn had none of the Sciences, Western General Hospital, University of Edinburgh, benefits of the modern clinician and indeed, at that Edinburgh, Scotland, UK. time, aldosterone had not even been discovered. It Based upon a paper presented at the Annual Scientific Meeting of the British Hypertension Society, held in Glasgow, Sep- was not long before electrocortin, as it was called by tember 1999. the Taits in London, became ‘aldosterone’ and it was Received 20 August 2001; accepted 11 October 2001 clear that what Conn had described was a benign Primary aldosteronism, a common entity? PL Padfield 160 aldosterone-producing adenoma of the adrenal identifying this or that adrenal steroid that might be gland. This condition has since been eponymously the cause of hypertension in such patients. In retro- named, Conn’s syndrome. Renin had been known spect it is clear of course that a relative aldosterone about for over 50 years and as assays became avail- excess (for a given level of renin) was present in able for the measurement of both renin activity in those patients with low renin hypertension. The plasma and aldosterone in , and subsequently absence of a suppressed aldosterone level was plasma, it became clear that Conn’s syndrome of against a role for an alternative mineralocorticoid aldosterone excess resulted in retention and (see Padfield et al9). suppression of renin. The hallmark of Conn’s syn- Through the 1970s it became clear to us in Glas- drome was high aldosterone production with a sup- gow that the differences between so-called idio- pressed renin. Initially all patients investigated had pathic hyperplasia and low renin hypertension were hypokalaemia but it became apparent that not all quantitative rather than qualitative. Seminal studies patients with low renin aldosterone excess had an performed around that time showed quite clearly adrenal adenoma.3 In early series it seemed clear that: that about a quarter of such patients had what came • Patients with true Conn’s syndrome had plasma to be called idiopathic hyperplasia or pseudo hyper- aldosterone levels that were insensitive to changes aldosteronism. These patients did not have a dis- 4 in circulating angiotensin II (induced by assump- creet tumour but had nodular adrenal glands. Some tion of the erect posture or by infusion of the pep- nodules could be large being indistinguishable from 10 5 tide itself). true tumours in a variety of imaging techniques. • Patients with the idiopathic hyperplastic variety While the removal of an aldosterone producing of primary aldosteronism had an enhanced aldos- adrenal adenoma resulted in cure of hypertension in terone response to such manoeuvres.11,12 at least 70% of cases, surgery for nodular hyper- plasia did not.6 There were even examples in the Idiopathic hyperplasia of the adrenal glands asso- literature of patients who had both adrenal glands ciated with aldosterone excess was not primary removed but who remained hypertensive following aldosteronism at all but some form of secondary replacement therapy. This is not surprising if it is aldosteronism whereby the was appreciated that the nodules from hyperplastic exquisitely sensitive to small changes in circulating glands do not make aldosterone; whereas of course angiotension II. The coup de grace for those who tumours do.7 This observation threw into question believed in qualitative differences was the obser- the whole notion of what so-called primary aldos- vation that when patients with low renin hyperten- teronism meant. The optimism of the 1960s where sion (a low circulating renin with a normal it was genuinely believed by some that up to 20% aldosterone) were infused with angiotensin II they of patients with hypertension might have primary behaved exactly as did the hyperplastic patients aldosteronism gradually subsided throughout the with an exaggerated aldosterone response. 1970s and most of the 1980s to a belief that primary We wrote in several papers around the middle aldosterone excess was a rare cause of hypertension. 1970s that idiopathic hyperaldosteronism and low Most textbooks and recent reviews now give it a renin hypertension were simply a continuum in the prevalence rate of probably less than 1% of the spectrum of essential hypertension (see Padfield et hypertensive population.8 al4, Davies et al10). There was much other circumstantial evidence to support this view. Pathophysiology • In patients with tumerous Conn’s syndrome a plot The measurement of plasma renin was always more of circulating renin or angiotensin II against readily available to clinicians than was the measure- aldosterone produced a negative correlation ment of aldosterone and in the search for aldos- (aldosterone induced sodium retention sup- terone excess it was found that a significant pro- pressing renin) whereas in those with the hyper- portion of patients (who otherwise had essential plastic form of the disease the relationship was hypertension) had a low or suppressed plasma renin positive (see Davies et al10). This is exactly what activity. While working at the MRC BP unit in Glas- would be expected if the aldosterone was high as gow 25 years ago, I showed that plasma renin levels, a result of angiotensin II. just like blood pressure itself, were normally distrib- • Patients with Conn’s syndrome had other evi- uted in hypertensive patients and therefore any sub- dence of mineralocorticoid excess in that division of hypertension based on the renin level exchangeable sodium was always elevated; it was was as arbitrary as the definition of hypertension not in the hyperplastic variety being on a par with 9 itself. Nonetheless, there was a flurry of literature those patients with low renin hypertension.10 in the 1970s describing the condition of low renin hypertension. These patients did not have aldos- terone excess but because low renin was a hallmark Clinical consequences of mineralocorticoid effect a search ensued for other Arguing as a purist for a particular disease entity, or mineralocorticoids. Isolated case reports appeared lack of it, is less important than appropriate treat-

Journal of Human Hypertension Primary aldosteronism, a common entity? PL Padfield 161 ment for patients. Although it had become clear that hypertension and over the years he has developed surgical therapy was not an option for patients with complicated protocols for the investigation of hyperplastic adrenal glands there was still the hope patients with hypertension. Initially he only investi- that they might respond to the aldosterone antagon- gated patients with hypokalaemia for the presence ist spironolactone. The response to spironolactone of aldosterone excess but as it became clear that 20– had been shown to be a good predictor of the 50% of patients with aldosterone excess did not response to adrenal surgery in those patients with present with hypokalaemia he investigated more an adrenal tumour.13 Patients with the hyperplastic widely, including patients with resistant hyperten- form of the disease did have good blood pressure sion. His current claim is that by doing this (using responses to spironolactone (300 mg or 400 mg per the aldosterone: renin ratio as the screening test) he day).14 Spironolactone was also used for low renin believes that primary aldosteronism has a preva- hypertension and although equally successful, was lence of around 12% amongst patients with hyper- no better than conventional thiazide diuretics.15 I tension.18 Most patients identified do not have surgi- know of no studies to address the question as to cally curable disease and it is difficult to be clear whether thiazides would be equally efficacious in from his published papers how often he discovers patients with hyperplastic aldosterone excess. an aldosterone producing adenoma? The Brisbane group have recently changed their stance on how the The diagnosis of ‘primary aldosteronism’ test should be performed and they now suggest that drug therapy is withdrawn before testing.19 This is Diagnosing aldosterone excess was always difficult no longer a simple process and it is my contention because most of the drugs that were used in the that most of the patients being identified by this treatment of hypertension had an effect on the renin screening test have what I would have called low angiotension aldosterone axis. Drugs either stimu- renin hypertension or idiopathic hyperplasia lated (diuretics, ACE inhibitors) or suppressed renin depending where the cut-off point is drawn. The (beta-blockers) and in the early days we always took same argument applies to recent studies within patients off their drugs, changing treatment to Tayside where a similar prevalence of primary aldo- bethanidine, a drug known not to impact on the sterone excess is claimed20 (again evidence from sur- renin angiotension system. This was time consum- gical cure is lacking). ing (it took several weeks for the effect of drugs on the renin angiotension system to wear off), uncomfortable for the patients (bethanidine was not Pragmatic perspectives a good drug) and the pick-up rate after investigation It would truly be foolish to argue against using the was of course low. The fact that there are no drugs aldosterone renin ratio as a screening test if it indeed that simultaneously suppress renin while stimulat- led to better targeted therapy for a significant pro- ing aldosterone led to arguments that the ratio of portion of hypertensive patients and I am prepared aldosterone to renin might be a good predictor of the to eat my words if this is shown to be the case. A state of primary aldosterone excess. It was postu- recent publication from Tayside (from the group lated as early as 1981 that use of the ratio was a good who suggested a prevalence of aldosterone excess of way to diagnose patients with primary aldosteron- 13%) has suggested that spironolactone is successful ism with an adrenal tumour.16 Since that original in treating blood pressure in these patients where paper others have used the ratio to diagnose all other drugs have failed.21 A claim, which if true, forms of low-renin aldosterone excess.17 Such a pos- would refute all my arguments above. The study ition ignores the fact that the ratio of aldosterone to however was totally uncontrolled and it is distinctly renin is increased in hypertensive patients com- possible that within the careful observation of a trial pared with normal subjects.10 In addition, the ratio situation patients were more compliant than they is affected predominantly by the level of PRA rather might otherwise have been. The answer to this ques- than aldosterone such that, almost by definition, low tion would only be resolved by a randomised con- renin hypertension is reclassified as hyperaldos- trol trial of spironolactone or other aldosterone teronism. As already argued the differentiation of antagonists against some other form of therapy. low renin hypertension from idiopathic hyperplasia It is not a simple matter to confirm or refute a diag- as a cause of aldosterone excess is probably quanti- nosis of primary hyperaldosteronism and until we tative rather than qualitative. The only important have better data on the outcomes of an expensive therapeutic decision therefore would be whether process we should leave well alone!22 such a test would identify patients with true Conn’s syndrome? Conclusions Hyperaldosteronism—common enough When we first wrote in 1981 in the Lancet that the to hunt? term idiopathic hyperaldosteronism was a myth (see Padfield et al4), we suggested that there is something Richard Gordon in Brisbane has spent much of his within the human psyche that needs to classify. It is academic life researching into mineralocorticoid inherent in our whole handling of the topic of blood

Journal of Human Hypertension Primary aldosteronism, a common entity? PL Padfield 162 pressure in that we need cut-off points to define of primary aldosteronism. Clin Endocrinol 1996; 45: hypertension from normotension; even though we 47–52. know that blood pressure is a continuum in the 9 Padfield PL et al. Is low-renin hypertension a stage in population. The same is true regarding aldosterone the development of essential hypertension or a diag- levels in plasma. Even if aldosterone is related to the nostic entity? Lancet 1975; 1: 548–550. 10 Davies DL et al. Aldosterone and its stimuli in normal simultaneous renin level there will not be daylight and hypertensive man: are essential hypertension and between normal and abnormal and unless or until primary hyperaldosteronism without tumour the same there is clear evidence that a significant proportion condition? J Endocr 1979; 81:79–91. of patients are identified who will be managed dif- 11 Wisgerhof M, Carpenter PC, Brown RD. Increased ferently it is hard to justify advocating the wide- adrenal sensitivity to angiotensin II in idiopathic hy- spread use of this test even though it is relatively peraldosteronism. J Clin Endocrinol Metab 1978; 47: simple. I still hold to the quotation we used in the 938–943. Lancet in 1981 as follows: 12 Kisch ES, Dluhy RG, Williams GH. Enhanced aldos- terone response to angiotensin II in human hyperten- sion. Circulation Res 1976; 38: 502–505. ‘The tendency has always been strong to believe that 13 Brown JJ et al. Comparison of surgery and prolonged whatever receives a name must be an entity or being, spironolactone therapy in patients with hypertension, having an independent existence of its own; and if aldosterone excess, and low plasma renin. BMJ 1972; no real entity answering to the name could be found 2: 729–734. man did not for that reason suppose that none 14 Ganguly A, Luetscher JA. Spironolactone therapy in existed but imagined that it was something peculi- primary aldosteronism: diagnostic and therapeutic arly abstruse and mysterious, too high to be an implications. In: Sambhi MP (ed). Systemic Effects of object of sense.’ Antihypertensive Agents. Stratton Intercontinental Medical Book: New York, 1976, pp 383–392. J Stewart Mill (1869)23 15 Carey RM et al. The syndrome of essential hyperten- sion and suppressed plasma renin activity. Normaliz- ation of blood pressure with spironolactone. Arch References Intern Med 1972; 130: 849–854. 1 Ferriss JB et al. Low—renin (‘primary’) hyperaldos- 16 Hiramatsu K et al. A screening test to identify aldos- teronism differential diagnosis and distinction of sub- terone-producing adenoma by measuring plasma renin groups within the syndrome. Am Heart J 1978; 95: activity. Arch Intern Med 1981; 141: 1589–1594. 641–658. 17 Gordon RD. Mineralocorticoid hypertension. Lancet 2 Conn JW. Primary aldosteronism, a new clinical syn- 1994; 344: 240–243. drome. J Labo Clini Med 1955; 45:3–17. 18 Gordon RD et al. Primary aldosteronism and other 3 Young WF et al. Primary aldosteronism: diagnosis and forms of mineralocorticoid hypertension. In: Swales J treatment. Mayo Clinic Proc 1990; 65:96–110. (ed.). Textbook of Hypertension. Blackwell Scientific: 4 Padfield PL et al. The myth of idiopathic hyperaldos- Oxford, 1994, pp 865–892. teronism. Lancet 1981; 2:83–84. 19 Stowasser M. Primary aldosteronism: revival of a syn- 5 Radin DR, Manoogian C, Nadicr JL. Diagnosis of pri- drome. J Hypertens 2001; 3: 363–366. mary hyperal-dosteronism: importance of correlating 20 Lim PO et al. High prevalence of primary aldosteron- CT findings with endocrinologic studies. Am J Roent- ism in the Tayside hypertension clinic population. J genol 1992; 158: 553–557. Hum Hypertens 2000; 5: 311–315. 6 Spark RF, Melby JC. Aldosteronism in hypertension: 21 Lim PO, Jung RT, MacDonald TM. Raised aldosterone the spironolactone response test. Ann Intern Med to renin ratio predicts antihypertensive efficacy of spi- 1968; 169: 685–695. ronolactone: a prospective cohort follow-up study. Br 7 Neville AM, O’Hare MJ. The human adrenal gland: J Clin Pharm 1999; 48: 756–760. aspects of structure, function and pathology. In: James 22 Kaplan NM. Cautions over the current epidemic of pri- VHT (ed). The Human Adrenal Gland, Academic mary aldosteronism. Lancet 2001; 357: 953–954. Press: London, 1979. 23 Mill JS. Analysis of the Phenomenon of the Human 8 Vallotton MB, Primary aldosteronism. Part 1 Diagnosis Mind, by James Mill, vol 2, London, 1869.

Journal of Human Hypertension