| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Research Center Studies |
Department of Medicine (W.R.L., C.C., G.H.W., R.G.D.), Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts 02115; Department of Medicine (P.S.), New England Deaconess Hospital; Howard Hughes Medical Institute (R.P.L.), Boyer Center for Molecular Genetics, Yale University School of Medicine, New Haven, Connecticut 06510; Division of Endocrinology (F.F.), University of Padova, 35122 Padova, Italy
Address all correspondence and requests for reprints to: W. Reid Litchfield, Endocrine/Hypertension Division, Brigham and Women’s Hospital, 221 Longwood Avenue, Boston, Massachusetts 02115-5817.
Abstract
Unlike other forms of primary aldosteronism, recent prospective studies have paradoxically revealed that glucocorticoid-remediable aldosteronism (GRA) is usually characterized by normal potassium (K+) levels. To evaluate this paradox we studied 10 GRA subjects and 14 healthy controls in two protocols: 1) the renal K+ excretory response to acute oral administration of 50 mmol K+ chloride and to fludrocortisone, 0.2 mg po q12 h x 4 doses; and 2) the aldosterone response to administration of 50 mmol K+ chloride.
The K+ excretion rate (KER) in GRA subjects (n = 6) at baseline (45.6 ± 8.3 µEq/min), after K+ (134 ± 34.2 µEq/min), and after fludrocortisone (100 ± 35.0 µEq/min) was not significantly different than that seen in the control (n = 8) subjects (54.9 ± 19.0, 154 ± 35.5, 112 ± 45.8 µEq/min, respectively). Thus the renal kaliuretic response to K+ ingestion and exogenous mineralocorticoid is normal in GRA. Serum aldosterone increased from 5.0 ± 3.8 at baseline to a maximum of 13.1 ± 6.6 ng/dL 60 min after K+ ingestion in control subjects (n = 7), but failed to increase in GRA subjects (n = 14), going from 8.7 ± 3.8 (baseline) to 8.8 ± 5.4 ng/dL at 60 min (P = 0.004 vs. control). The blunted aldosterone response to K+ in GRA in association with the sharp diurnal decline in aldosterone in this ACTH-regulated syndrome probably results in a milder degree of hyperaldosteronism compared with other forms of primary aldosteronism, thereby producing volume expansion with minimal renal K+ wasting.
GLUCOCORTICOID-REMEDIABLE aldosteronism (GRA) represents a rare form of primary aldosteronism (1, 2, 3, 4). Normokalemia may occur in some patients with primary aldosteronism (5, 6), but spontaneous hypokalemia is present in a majority of these patients. In contrast, in GRA, normokalemia is the rule (7, 8, 9). This reduced incidence of hypokalemia is seen in spite of the fact that GRA is a mineralocorticoid-excess state characterized by moderate to severe hypertension and suppressed plasma renin activity (PRA) (10).
The failure to consistently find low potassium (K+) levels in GRA could be explained by several possibilities including: self-selected low sodium (Na+) diets in GRA subjects, decreased renal K+ excretory response to mineralocorticoids, or the unique steroidogenesis and regulation of aldosterone secretion seen in GRA. This study was designed to evaluate K+ homeostasis in GRA by assessing: 1) the renal response to kaliuretic stimuli, including oral K+ loading and exogenous mineralocorticoid administration; and 2) the regulation of aldosterone secretion by K+ in this disorder.
Patients and Methods
Ten patients who tested positive for the chimeric gene that causes GRA (1) and 14 healthy age-matched control subjects were studied in the General Clinical Research Center of the Brigham and Women’s Hospital. Control subjects were normotensive at the time of screening and were not taking medications. The four GRA patients receiving treatment for hypertension (spironolactone in 1 subject, spironolactone and atenolol in 1 subject, tenormin/hydrochlorothiazide in 1 subject, and amiloride and propranolol in 1 subject) were taken off all medications 1 week before the study, but the severity of hypertension necessitated retreatment with sustained-release nifedipine in 3 of the 4 subjects. The other GRA patients were not taking antihypertensive agents. No patients took K+-wasting diuretics or K+ supplements at the time of study. All subjects were studied on a salt replete diet, and all were known to be normokalemic before the commencement of the study. Studies were performed with the patients in the fasting state and in a semi-recumbent position. Blood samples were drawn through indwelling venous catheters that had been placed at least 60 min before the study. The study was approved by the Human Research Committee of the Brigham and Women’s Hospital, and all patients gave informed and written consent before impanelment in the study.
Protocol 1: Kaliuretic response to oral K+ loading and fludrocortisone
Protocol 1A: Oral K+ loading.Six normokalemic GRA patients and eight healthy control subjects were studied supine after an overnight fast on a 150 mmol Na+, 100 mmol K+ diet. The subjects were given 50 mmol oral K+ chloride (KCl) at 0800, followed by 0.9% NaCl at 85 mL/h for 6 h, to ensure that renal K+ excretion was not limited by inadequate luminal Na+ concentration. Blood was sampled for K+ and aldosterone at baseline, 60, and 360 min; urine was collected during the 6-h infusion. The potassium excretion rate (KER, µEq urinary K+/min) was calculated at baseline (24-h urine collection obtained on the day of admission) and compared with the response following the administration of KCl.
Protocol 1B: Exogenous mineralocorticoid loading. Six normokalemic GRA patients and eight healthy control subjects began this study on the day after the K+ loading study; the diet was kept constant. Fludrocortisone (Florinef, Apothecon, Princeton, NJ) 0.2 mg po q12h x 4 doses was given at 0600 and 1800 h, with the last dose being given at 0600 on the morning of study. Approximately 2 h after the last dose of fludrocortisone was given, 0.9% NaCl at 85 mL/h was administered over 6 h as described above. Blood and urine were collected and KER calculated as described in protocol 1A.
Protocol 2: Aldosterone response to oral K+ loading
This study was conducted in ten additional patients (4 GRA and 6 controls) and was designed to determine the peak serum K+ and aldosterone response to oral KCl loading. Baseline (-30, 0 min) blood samples for cortisol, aldosterone, and PRA were drawn at 0800 through an indwelling venous catheter; 50 mmol KCl elixir was then administered orally followed by blood sampling for cortisol, aldosterone, and K+ (30, 60, and 120 min). Cortisol was assessed to ensure that the aldosterone responses were not related to a rise in ACTH levels. The KCl was mixed in 120 ml sugar-free cranberry juice to mask the bitter taste.
Laboratory methods
K+ was measured using an ion-selective electrode (NOVA Biomedical, Waltham, MA). Commercial radioimmunoassay (RIA) kits were used to measure aldosterone (Diagnostic Products Corporation, Los Angeles, CA) and cortisol (INCSTAR Corporation, Stillwater, MN). The intra- and interassay coefficients of variation (CV) for aldosterone were 6.0 and 10% (normal range 1–10 ng/dL high salt diet and upright posture). Intra- and interassay CV for cortisol were 4.5 and 6.7% (normal range 9–24 ug/dL at 0800). A commercial RIA kit (INCSTAR Corporation, Stillwater, MN) was used to measure PRA in supine subjects on a high Na+ diet (6). The intra- and interassay CV were 7.0 and 12.3% (normal range 0.5–3.0 ng/mL/h).
Statistical analysis
Results are expressed as mean ± SD. Groups
were compared using a nonpaired t-test with an 
value of
0.05. The multiple comparisons in protocol 2 were performed using
repeated measures ANOVA, with one grouping factor (GRA vs.
normal) and one repeated factor (time of sampling). The interaction of
group and time was assessed to determine if group membership predicted
the response pattern after provocation.
Results
Protocol 1: Kaliuretic response to oral K+ loading and fludrocortisone
At the time of admission to the General Clinical Research Center, all GRA and control subjects were similar with regard to serum K+ (4.12 ± 0.29 mmol/L vs. 4.28 ± 0.54 mmol/L, P = 0.53), age (34.5 ± 8.0 vs. 33.1 ± 10.6, P = 0.79), body mass index (28.2 ± 1.17 vs. 26.4 ± 3.1, P = 0.21), urinary Na+ excretion (195 ± 96.1 mmol/d vs. 270 ± 84.3 mmol/d, P = 0.15), and urinary K+ excretion (65.7 ± 11.9 mmol/d vs. 79 ± 27.3 mmol/d, P = 0.29). The proportion of female subjects in each study group (83% for GRA, 50% for controls) was not significantly different by Fisher’s exact test. As expected, the GRA patients were hypertensive compared with controls (systolic blood pressure 156 ± 26.2 vs. 128 ± 17.6 mm Hg, P = 0.03, and diastolic blood pressure 103 ± 11.9 vs. 71 ± 10.3 mmHg. P = 0.0002).
Serum K+ increased after oral KCl (50 mmol) in both GRA
patients and controls; the mean increment in serum K+ at 60
min was 0.39 ± 0.34 mmol/L and 0.69 ± 0.47 mmol/L in GRA
and controls, respectively (P = 0.29). The mean KER was
not significantly different at baseline (admission) between GRA and
control subjects (45.6 ± 8.3 and 54.9 ± 18.9 µEq/min
respectively, P = 0.29; Fig. 1
). There
were significant increases in KER after oral KCl in both GRA
(P < 0.01) and control subjects (P <
0.001) compared with baseline. After KCl ingestion, neither the mean
KER (134 ± 34.2 vs. 154 ± 35.5 µEq/min) nor
the increment in KER above baseline (88 ± 37.5 vs.
99 ± 40.1 µEq/min) were significantly different between GRA and
control subjects.
|
).
However, the mean KER after fludrocortisone loading was not
significantly different between GRA (100 ± 35.0 µEq/min) and
control subjects (112 ± 45.8 µEq/min; P =
0.60). Protocol 2
PRA was lower in the GRA group (0.23 ± 0.13 ng/mL/hr) than the control group (0.78 ± 0.36 ng/mL/hr, P = 0.022). Serum K+ peaked 60 min after administration of 50 mmol KCl in both controls and GRA patients (mean serum K+ 4.9 ± 0.26 and 4.9 ± 0.58 mmol/L, respectively); there was no difference in the increment in serum K+ in either group of patients. The control subjects showed a peak aldosterone response 60 min after KCl loading, which coincided with the peak in serum K+ (mean serum aldosterone at -30, 0, 30, 60, and 120 min was 5.7 ± 4.1, 4.7 ± 2.6, 7.2 ± 4.8, 14.5 ± 7.0, and 10.4 ± 7.6 ng/dL, respectively). By contrast, the GRA subjects had insignificant aldosterone responses to KCl loading at all time points (corresponding mean aldosterone levels of 11.4 ± 6.2, 7.5 ± 3.1, 9.7 ± 6.1, 8.7 ± 5.7, and 8.4 ± 3.4 ng/dL). Repeated measures ANOVA showed that the aldosterone response to KCl loading was significantly different in the two groups (P = 0.017). In both GRA and control groups, the serum cortisol declined over the course of the study; the patterns were not significantly different by ANOVA.
In protocol 2, the peak serum aldosterone and K+ responses to KCl loading in normal subjects occurred after 60 min. We therefore added the aldosterone, cortisol, and K+ data from time 60 for 3 GRA and 8 control subjects studied in protocol 1A to the data at time 60 for the patients studied in protocol 2. These 7 GRA patients and 14 control subjects were similar with respect to age (30.6 ± 10.6 vs. 31.1 ± 8.7) and baseline K+ (4.10 ± 0.28 vs. 4.2 ± 0.42) and cortisol (8.7 ± 4.7 vs. 12.6 ± 7.7) levels and gender breakdown. The GRA subjects had significantly higher BMI (30.0 ± 3.5 vs. 25.2 ± 2.8; P = 0.003), systolic blood pressures (134 ± 16 vs. 117 ± 18 mm Hg; P = 0.04) and diastolic blood pressures (91 ± 13 vs. 71 ± 8.4 mm Hg; P = 0.004).
The baseline and increment in serum K+ at 60 min in the GRA
patients (4.1 ± 0.28 and 0.61 ± 0.40 mmol/L, respectively)
were not significantly different from control subjects (4.2 ±
0.42 and 0.74 ± 0.39 mmol/L). Basal aldosterone levels (Fig. 2) were significantly higher in GRA patients compared
with controls (8.7 ± 3.8 vs. 5.0 ± 3.8 ng/dL
respectively, P = 0.045). Sixty minutes after oral
K+ administration, aldosterone increased from baseline in
controls (mean increase 8.1 ± 5.7 ng/dL) but was blunted in GRA
patients (0.13 ± 4.0 ng/dL; P = 0.004
vs. control).
|
In earlier studies, GRA was described primarily in individuals with overt hypokalemia, although several normokalemic kindreds were reported (8). However, recent prospective screening of large kindreds has revealed that the majority of GRA subjects are normokalemic, indicating that earlier observations reflected ascertainment bias (7, 9).
Normokalemia in the setting of a mineralocorticoid-excess state is not unique to GRA. Although the prevalence of normal K+ levels in primary aldosteronism varies from 7–38% (5, 7, 11, 12), Bravo et al. (6) had reported that 90% of such individuals become hypokalemic with salt loading. In the present study, the mean urinary Na+ in the normokalemic GRA patients was 195 ± 96 mmol/day, and they were further salt-loaded with 75 mmol Na+ on each of their study days. Thus, in GRA, normokalemia is seen despite a dietary Na+ intake similar to what Bravo et al. used to induce hypokalemia in normokalemic primary aldosteronism patients (6).
Another explanation for normokalemia in GRA subjects is the possibility that they have a decreased renal K+ excretory response to kaliuretic stimuli. K+-stimulated kaliuresis and mineralocorticoid-induced kaliuresis represent the most important stimuli for renal K+ excretion. However, GRA patients in our study responded to KCl loading and fludrocortisone with kaliuretic responses that were the same as in control subjects. Other determinants of renal K+ excretion (13), such as negative lumen potential, adequate luminal flow rates, and Na+ concentration were considered in this study. These potential limitations of renal K+ excretion were avoided by invoking a high Na+ diet and saline infusion during the study. Consequently, the response to each of the kaliuretic stimuli could be assessed independently of these confounding factors.
The hybrid compounds 18-hydroxy-cortisol and 18-oxo-cortisol, which are greatly over produced in GRA, have been shown to have weak mineralocorticoid activities (14, 15). However, the normal kaliuretic response to fludrocortisone argues against competitive antagonism of the renal type I mineralocorticoid receptor by these compounds.
The observed normokalemia in GRA could reflect the altered regulation of aldosterone secretion in GRA. There are normally three secretagogues for aldosterone production: angiotensin II (Ang II), adrenocorticotropin (ACTH), and K+ (16). The chimeric gene that causes GRA (1, 3) results in aldosterone production primarily under the regulation of ACTH as a result of ectopic expression of aldosterone synthase activity in the zona fasciculata (17, 18). The resulting hyperaldosteronism suppresses the renin-angiotensin system and the biosynthetic pathways in the zona glomerulosa (19, 20, 21, 22). Unlike the zona glomerulosa, steroidogenesis in the zona fasciculata is not responsive to K+ (23). In this study, control subjects had an expected increase in aldosterone levels after KCl administration (24), while the GRA patients failed to respond. This lack of response was seen in the setting of similar increments in serum K+ levels in controls and GRA subjects after oral KCl loading.
It is impossible to prove that the diminished K+-stimulated aldosterone response seen in GRA is not caused by the effects of nifedipine, which some of the GRA patients took during the study. However, it is important to concede that the in vivo experience with dihydropyridine calcium channel blockers in the treatment of primary aldosteronism has been disappointing at best (25, 26, 27). The fact that the patients treated with nifedipine had a response that was identical to the GRA patients studied off all medication makes us think that nifedipine therapy was not the cause of the blunted aldosterone response to potassium loading.
The consequences of absent K+-induced aldosterone production may be clinically relevant. In fact, it may result in a milder degree of hyperaldosteronism in GRA compared with other forms of primary aldosteronism. This possibility is supported by the wide overlap in the aldosterone excretion rates of GRA patients in comparison to normal subjects or unaffected family members (7). A direct comparison of aldosterone levels in GRA vs. aldosterone-producing adenoma (APA) or bilateral idiopathic hyperplasia (IHA) has not been formally done. However, a comparison of the serum aldosterone in the 12 GRA patients of Rich et al.(7) with 70 APA patients studied by Bravo et al. (6), revealed significantly lower levels in GRA (21.6 ± 12 ng/dL) vs. APA (44.8 ± 36.6 ng/dL; P < 0.001). Aldosterone levels in 10 IHA patients (41.0 ± 14.4 ng/dL) were intermediate between APA and GRA (P < 0.01 vs. GRA). It is also noteworthy that Fallo et al.(28) have shown that aldosterone secretion in both APA and IHA patients is responsive to K+ loading. In spite of being hypokalemic, IHA and APA patients, respectively, had 67% and 61% increases in serum aldosterone levels after a 40 mmol iv K+ load.
Beyond the absence of K+-induced aldosterone secretion associated with meals, aldosterone secretion in GRA follows ACTH, declining diurnally, with low levels during the late evening hours. Thus, the GRA patient may be afforded an escape from mineralocorticoid-induced kaliuresis that is not seen in other mineralocorticoid-excess states, such as APA and IHA, where there is autonomous aldosterone production.
In summary, an impairment in the renal responsiveness to K+ or mineralocorticoids cannot explain the normal serum K+ seen in the majority of GRA subjects. However, the absence of a K+-induced increase in aldosterone and the sharp diurnal decline in aldosterone in GRA could result in a milder form of hyperaldosteronism that is sufficient to suppress PRA in most patients, but not severe enough to result in renal K+ wasting.
Footnotes
1 This work was supported by the following grants: GCRC Grant 5 MO1
RR002635, NIH Grant 5 RO1 HL 53693, NIH Grant T32HL07609, and an
Endocrine Fellows Foundation Grant. 
Received October 18, 1996.
Accepted January 15, 1997.
References
1–24 ACTH, and
potassium: a comparative study. Am J Physiol. 233:E402–E406.
This article has been cited by other articles:
 |
|
 |
|
  W. F. YOUNG Jr. Primary Aldosteronism: Management Issues Ann. N.Y. Acad. Sci., September 1, 2002; 970(1): 61 - 76. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Mulatero, S. M. di Cella, T. A. Williams, A. Milan, G. Mengozzi, L. Chiandussi, C. E. Gomez-Sanchez, and F. Veglio Glucocorticoid Remediable Aldosteronism: Low Morbidity and Mortality in a Four-Generation Italian Pedigree J. Clin. Endocrinol. Metab., July 1, 2002; 87(7): 3187 - 3191. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. V Milford Investigation of hypertension and the recognition of monogenic hypertension Arch. Dis. Child., November 1, 1999; 81(5): 452 - 455. [Full Text] [PDF]
|
|
|
|
|
|
P. Mulatero, F. Veglio, C. Pilon, F. Rabbia, C. Zocchi, P. Limone, M. Boscaro, N. Sonino, and F. Fallo Diagnosis of Glucocorticoid-Remediable Aldosteronism in Primary Aldosteronism: Aldosterone Response to Dexamethasone and Long Polymerase Chain Reaction for Chimeric Gene J. Clin. Endocrinol. Metab., July 1, 1998; 83(7): 2573 - 2575. [Abstract] [Full Text]
|
|
|
|
 |
|
 W. R. Litchfield, B. F. Anderson, R. J. Weiss, R. P. Lifton, and R. G. Dluhy Intracranial Aneurysm and Hemorrhagic Stroke in Glucocorticoid-remediable Aldosteronism Hypertension, January 1, 1998; 31(1): 445 - 450. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ganguly Potassium and Aldosterone Secretion in Glucocorticoid-Remediable Aldosteronisme J. Clin. Endocrinol. Metab., December 1, 1997; 82(12): 4276a - 4277. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
