0021-972X/90/7006-01637$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 70, No. 6 Printed in U.S.A.
The Influence of Glycyrrhetinic Acid on Plasma Cortisol and Cortisone in Healthy Young Volunteers MARIUS A. MACKENZIE, WILLIBRORD H. L. HOEFNAGELS, RENfi W. M. M. JANSEN, THEO J. BENRAAD, AND PETER W. C. KLOPPENBORG Department of Medicine, Division of Endocrinology (M.A.M., P.W.C.K.), and the Departments of Geriatric Medicine (W.H.L.H., R. W.M.M.J.) and Experimental and Chemical Endocrinology (T.J.B.), University Hospital, Nijmegen, The Netherlands
ABSTRACT. Based on studies in laboratory animals and on measurements of the urinary metabolites (allo)tetrahydrocortisol and tetrahydrocortisone in human volunteers it has been claimed that liquorice-induced mineralocorticoid excess is caused by a unique defect in the conversion of cortisol to cortisone. To further evaluate this hypothesis we have investigated the influence of glycyrrhetinic acid (GA), the mineralocorticoid-active constituent of liquorice, on plasma cortisol and cortisone in 10 healthy young normotensive volunteers. Pure GA (500 mg/day), administered orally from days 3-10
L
of the study, exerted pronounced mineralocorticoid activity. Ingestion of GA resulted in an elevated urinary excretion of free cortisol and virtually unchanged plasma cortisol levels in the presence of markedly decreased levels of both plasma cortisone and urinary free cortisone. These results provide direct clinical support for the hypothesis that GA induces an inhibition of the activity of 11/3-dehydrogenase, resulting in a blockade in the conversion of cortisol to cortisone. (J Clin Endocrinol Metab 70: 1637-1643,1990)
IQUORICE extracts, derived from the roots of the plant Glycyrrhiza glabra, have been used for many decades as flavoring agents for food and luxury products, such as liquorice, chocolates, and beer, as well as for medical use. In 1946 Revers (1) reported on the usefulness of succus liquiritiae in the treatment of peptic ulcers. Thereafter, he observed that edema and hypertension or cardiac asthma occurred in about 20% of the patients treated with this medication (2). In 1950 Molhuysen and coworkers (3) demonstrated that liquorice extracts administered to normal subjects exerted deoxycortone-like action, as was established by retention of sodium and chloride as well as by an increased urinary excretion of potassium. They demonstrated that glycyrrhizic acid was the mineralocorticoid principle of succus liquiritiae (3). Later, it was established that glycyrrhetinic acid (GA), a hydrolytic product of glycyrrhizic acid, is mainly responsible for the mineralocorticoid activity of liquorice extracts (4, 5). Although liquorice abuse is a well known cause of mineralocorticoid hypertension in The Netherlands, this diagnosis may be easily overlooked because these paReceived November 21,1989. Address all correspondence and requests for reprints to: M. A. MacKenzie, M.D.~, Department of Medicine, Division of Endocrinology, University Hospital, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands.
tients tend to deny addiction to liquorice. Excessive intake of liquorice has been shown to induce a syndrome of pseudohyperaldosteronism, including hypertension, retention of sodium and water, hypokalemia, and suppression of the renin-angiotensin-aldosterone axis (6-8). A valuable contribution toward the understanding of the mechanism of mineralocorticoid action of glycyrrhetinic acid has been the recognition of a remarkable similarity between the clinical picture of excessive liquorice ingestion and the syndrome of apparent mineralocorticoid excess that has been described in patients with congenital deficiency of Ilj8-hydroxysteroid dehydrogenase (11/3-OHSD); this enzyme complex catalyzes the (reversible) conversion of cortisol to cortisone (9-15). In these patients an impaired conversion of cortisol to cortisone is found, resulting in an elevated ratio of the metabolites of cortisol to cortisone, a prolonged plasma cortisol half-life, and a decreased cortisol secretion rate. In addition, increased urinary excretion of unconjugated cortisol has been demonstrated (9,10,12-15). It has been claimed that cortisol acts as a potent mineralocorticoid in this disorder (11,16). Furthermore, it has been established that both patients with congenital deficiency of 11/3-OHSD and subjects with excessive liquorice ingestion respond to spironolactone (17). Based on measurements of the urinary metabolites
1637
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 07:03 For personal use only. No other uses without permission. . All rights reserved.
MACKENZIE ET AL.
1638
(allo)tetrahydrocortisol and tetrahydrocortisone in human volunteers, it has been claimed that liquorice-induced mineralocorticoid excess is caused by an inhibition of 110-OHSD (18). To further evaluate and provide direct clinical evidence for an inhibition of the conversion of cortisol to cortisone we have studied the influence of pure glycyrrhetinic acid on plasma cortisol and cortisone in healthy volunteers.
Subjects and Methods Ten healthy normotensive volunteers [five male and five female; mean age, 24.3 ± 2.6 (±SD) yr] were studied in the outpatient department for 10 consecutive days. For several months preceding the study the subjects had not used medication (including oral contraceptives) and had only occasionally taken very small amounts of liquorice; no liquorice had been used in the 4 weeks preceding the investigation. All volunteers consumed their regular diet during the study and underwent a routine physical examination before entering the study. From days 3-10 of the study pure GA (500 mg/day) was administered orally in two divided doses; this dose is equivalent to the consumption of 200 g confectionery liquorice/day. Daily at 0900 h blood samples for measurements of plasma cortisol and cortisone were collected after at least 10 min of rest in the supine position. Body weight was recorded, and blood pressure was measured with a mercury sphygmomanometer in the supine position and after 2 min in the standing position. On days 1 and 10 of the study at 0900 h blood samples were taken for measurements of sodium, potassium, bicarbonate, PRA, aldosterone, and atrial natriuretic peptide (ANP). Each day of the study 24-h urine was collected for determination of sodium, potassium, free cortisol, and free cortisone excretion. Before and after the first oral administration of GA (250 mg), blood samples were collected, using an indwelling iv catheter, at regular intervals during 2 h for measurement of plasma cortisol and cortisone. On days 2 and 10 of the study adrenocortical response was evaluated by the administration of 0.25 mg tetracosactrin [Synacthen®, synthetic ACTH-(l-24), Ciba-Geigy, Basel, Switzerland] in 100 mL 0.9% NaCl, iv, in 10 min. Blood samples were taken 30 min before and 0, 30, 60, and 90 min after administration for measurement of plasma cortisol and cortisone.
JCE & M • 1990 Vol70-No6
RIA. The sensitivity of the RIA for aldosterone was 6.2 fmol/ tube, i.e. 0.04 nmol/L plasma, when procedural losses are taken into account. The intraassay coefficient of variation (CV) was 4.8% at a level of 0.35 nmol/L (n = 12) and 3.8% at a level of 1.17 nmol/L (n = 12). Interassay CVs were 12.3% at a level of 0.32 nmol/L (n = 19) and 11.5% at a level of 1.27 nmol/L. The sensitivity of the RIA for cortisol was 9 fmol/tube, i.e. 0.06 nmol/L. The intraassay CV, expressed as the relative duplovariation, was 5.3% in samples with a concentration of 0.36 ± 0.19 Mmol/L (n = 50). The interassay CV was 5.0% at a level of 0.38 jumol/L. The sensitivity of the RIA for cortisone was 13.4 fmol/tube, i.e. 0.09 nmol/L. The intra- and interassay CVs were 3.8% in samples with a mean concentration of 44.9 nmol/ L (n = 20) and 4.8% at a level of 40.3 nmol/L (n = 16), respectively. PRA was measured with the Phadebas angiotensin-I test (Pharmacia Diagnostics, Uppsala, Sweden). ANP was measured by RIA (21). The blood as well as urine samples of each subject were analyzed in the same assay run. Statistical methods The effect of GA on plasma cortisol and cortisone, body weight, and blood pressure was established by comparing these parameters after GA administration (days 4-10) with the means of the values measured on days 1-3 of the study. The urinary parameters after GA intake were compared with the basal values measured on day 1 of the study. Statistical analysis was performed using repeated measures analysis of variance with multiple comparison procedures by means of the contrast option of the GLM procedure (SAS Institute, Inc., Cary, NC). For the comparison of blood parameters between days 1 and 10 the paired t test was applied. For comparison of the maximum and overall responses of plasma cortisol and cortisone to synthetic ACTH before and after GA administration the Wilcoxon signed rank test was performed. For all procedures P values of 0.05 were required for statistical significance. All results are expressed as the mean ± SEM unless stated otherwise. Approval for the study was obtained from the Ethical Committee of the University Hospital Nijmegen; informed consent was obtained from all subjects.
Results Mineralocorticoid effect
Laboratory methods Plasma cortisol, cortisone, and aldosterone as well as urinary free (i.e. extractable by organic solvents) cortisol and cortisone were measured after extraction and paper chromatography, as described previously (19, 20). In short, 0.2 mL urine or plasma (2.0 mL plasma in the case of aldosterone) were incubated with 10,000 dpm recovery tracer. The samples were extracted with 15.0 mL dichloromethane. The aqueous layer was removed, and the organic phase dried under a stream of air. Separation of the different steroids was obtained by descending paper chromatography and a modified Bush B5 system (toluene: methanol: water = 2:1:1) as eluent. The chromatogram was scanned for radioactivity, and the appropriate area was cut and eluted in ethylene glycol in water (0.2%). The eluates were subjected to
The administration of GA induced a mineralocorticoid effect, as established by a significant increase in plasma sodium and urinary potassium (day 3-7) as well as a significant fall of plasma potassium, PRA, and aldosterone; plasma ANP was significantly enhanced after GA administration (Table 1). The ratio of urinary sodium to potassium was significantly lower on days 3-5 and 8 of the study. Blood pressure was 112/73 ± 10/6 (mean ± SD) mm Hg in the supine position and 117/80 ± 9/5 mm Hg after 2 min in the standing position. The intake of GA did not influence either supine or standing blood pressure in the present study. During the administration of GA four volunteers (two men and two women) com-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 07:03 For personal use only. No other uses without permission. . All rights reserved.
EFFECT OF GA ON CORTISOL AND CORTISONE TABLE 1. Mineralocorticoid effect of GA administration
PRA(ng/Ls) Aldosterone (pmol/L) ANP (pmol/L) Sodium (mmol/L) Potassium (mmol/L) Bicarbonate (mmol/L) Creatinine (/xmol/L)
Day 1
Day 10
0.37 (0.06) 364 (75) 9.0 (1.1) 141 (0.5) 3.7 (0.9) 27.3 (4.8) 79 (4.0)
0.13 (0.03)° 166 (96)" 28.3 (4.2)° 144 (0.6)° 3.3 (0.4)° 29.0 (6.5)° 79 (4.5)c
Plasma parameters of mineralocorticoid action of GA before (day 1) and after (day 10) GA administration. Values are means, with the SEM in parentheses. °P
Vol. 70, No. 6 Printed in U.S.A.
The Influence of Glycyrrhetinic Acid on Plasma Cortisol and Cortisone in Healthy Young Volunteers MARIUS A. MACKENZIE, WILLIBRORD H. L. HOEFNAGELS, RENfi W. M. M. JANSEN, THEO J. BENRAAD, AND PETER W. C. KLOPPENBORG Department of Medicine, Division of Endocrinology (M.A.M., P.W.C.K.), and the Departments of Geriatric Medicine (W.H.L.H., R. W.M.M.J.) and Experimental and Chemical Endocrinology (T.J.B.), University Hospital, Nijmegen, The Netherlands
ABSTRACT. Based on studies in laboratory animals and on measurements of the urinary metabolites (allo)tetrahydrocortisol and tetrahydrocortisone in human volunteers it has been claimed that liquorice-induced mineralocorticoid excess is caused by a unique defect in the conversion of cortisol to cortisone. To further evaluate this hypothesis we have investigated the influence of glycyrrhetinic acid (GA), the mineralocorticoid-active constituent of liquorice, on plasma cortisol and cortisone in 10 healthy young normotensive volunteers. Pure GA (500 mg/day), administered orally from days 3-10
L
of the study, exerted pronounced mineralocorticoid activity. Ingestion of GA resulted in an elevated urinary excretion of free cortisol and virtually unchanged plasma cortisol levels in the presence of markedly decreased levels of both plasma cortisone and urinary free cortisone. These results provide direct clinical support for the hypothesis that GA induces an inhibition of the activity of 11/3-dehydrogenase, resulting in a blockade in the conversion of cortisol to cortisone. (J Clin Endocrinol Metab 70: 1637-1643,1990)
IQUORICE extracts, derived from the roots of the plant Glycyrrhiza glabra, have been used for many decades as flavoring agents for food and luxury products, such as liquorice, chocolates, and beer, as well as for medical use. In 1946 Revers (1) reported on the usefulness of succus liquiritiae in the treatment of peptic ulcers. Thereafter, he observed that edema and hypertension or cardiac asthma occurred in about 20% of the patients treated with this medication (2). In 1950 Molhuysen and coworkers (3) demonstrated that liquorice extracts administered to normal subjects exerted deoxycortone-like action, as was established by retention of sodium and chloride as well as by an increased urinary excretion of potassium. They demonstrated that glycyrrhizic acid was the mineralocorticoid principle of succus liquiritiae (3). Later, it was established that glycyrrhetinic acid (GA), a hydrolytic product of glycyrrhizic acid, is mainly responsible for the mineralocorticoid activity of liquorice extracts (4, 5). Although liquorice abuse is a well known cause of mineralocorticoid hypertension in The Netherlands, this diagnosis may be easily overlooked because these paReceived November 21,1989. Address all correspondence and requests for reprints to: M. A. MacKenzie, M.D.~, Department of Medicine, Division of Endocrinology, University Hospital, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands.
tients tend to deny addiction to liquorice. Excessive intake of liquorice has been shown to induce a syndrome of pseudohyperaldosteronism, including hypertension, retention of sodium and water, hypokalemia, and suppression of the renin-angiotensin-aldosterone axis (6-8). A valuable contribution toward the understanding of the mechanism of mineralocorticoid action of glycyrrhetinic acid has been the recognition of a remarkable similarity between the clinical picture of excessive liquorice ingestion and the syndrome of apparent mineralocorticoid excess that has been described in patients with congenital deficiency of Ilj8-hydroxysteroid dehydrogenase (11/3-OHSD); this enzyme complex catalyzes the (reversible) conversion of cortisol to cortisone (9-15). In these patients an impaired conversion of cortisol to cortisone is found, resulting in an elevated ratio of the metabolites of cortisol to cortisone, a prolonged plasma cortisol half-life, and a decreased cortisol secretion rate. In addition, increased urinary excretion of unconjugated cortisol has been demonstrated (9,10,12-15). It has been claimed that cortisol acts as a potent mineralocorticoid in this disorder (11,16). Furthermore, it has been established that both patients with congenital deficiency of 11/3-OHSD and subjects with excessive liquorice ingestion respond to spironolactone (17). Based on measurements of the urinary metabolites
1637
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 07:03 For personal use only. No other uses without permission. . All rights reserved.
MACKENZIE ET AL.
1638
(allo)tetrahydrocortisol and tetrahydrocortisone in human volunteers, it has been claimed that liquorice-induced mineralocorticoid excess is caused by an inhibition of 110-OHSD (18). To further evaluate and provide direct clinical evidence for an inhibition of the conversion of cortisol to cortisone we have studied the influence of pure glycyrrhetinic acid on plasma cortisol and cortisone in healthy volunteers.
Subjects and Methods Ten healthy normotensive volunteers [five male and five female; mean age, 24.3 ± 2.6 (±SD) yr] were studied in the outpatient department for 10 consecutive days. For several months preceding the study the subjects had not used medication (including oral contraceptives) and had only occasionally taken very small amounts of liquorice; no liquorice had been used in the 4 weeks preceding the investigation. All volunteers consumed their regular diet during the study and underwent a routine physical examination before entering the study. From days 3-10 of the study pure GA (500 mg/day) was administered orally in two divided doses; this dose is equivalent to the consumption of 200 g confectionery liquorice/day. Daily at 0900 h blood samples for measurements of plasma cortisol and cortisone were collected after at least 10 min of rest in the supine position. Body weight was recorded, and blood pressure was measured with a mercury sphygmomanometer in the supine position and after 2 min in the standing position. On days 1 and 10 of the study at 0900 h blood samples were taken for measurements of sodium, potassium, bicarbonate, PRA, aldosterone, and atrial natriuretic peptide (ANP). Each day of the study 24-h urine was collected for determination of sodium, potassium, free cortisol, and free cortisone excretion. Before and after the first oral administration of GA (250 mg), blood samples were collected, using an indwelling iv catheter, at regular intervals during 2 h for measurement of plasma cortisol and cortisone. On days 2 and 10 of the study adrenocortical response was evaluated by the administration of 0.25 mg tetracosactrin [Synacthen®, synthetic ACTH-(l-24), Ciba-Geigy, Basel, Switzerland] in 100 mL 0.9% NaCl, iv, in 10 min. Blood samples were taken 30 min before and 0, 30, 60, and 90 min after administration for measurement of plasma cortisol and cortisone.
JCE & M • 1990 Vol70-No6
RIA. The sensitivity of the RIA for aldosterone was 6.2 fmol/ tube, i.e. 0.04 nmol/L plasma, when procedural losses are taken into account. The intraassay coefficient of variation (CV) was 4.8% at a level of 0.35 nmol/L (n = 12) and 3.8% at a level of 1.17 nmol/L (n = 12). Interassay CVs were 12.3% at a level of 0.32 nmol/L (n = 19) and 11.5% at a level of 1.27 nmol/L. The sensitivity of the RIA for cortisol was 9 fmol/tube, i.e. 0.06 nmol/L. The intraassay CV, expressed as the relative duplovariation, was 5.3% in samples with a concentration of 0.36 ± 0.19 Mmol/L (n = 50). The interassay CV was 5.0% at a level of 0.38 jumol/L. The sensitivity of the RIA for cortisone was 13.4 fmol/tube, i.e. 0.09 nmol/L. The intra- and interassay CVs were 3.8% in samples with a mean concentration of 44.9 nmol/ L (n = 20) and 4.8% at a level of 40.3 nmol/L (n = 16), respectively. PRA was measured with the Phadebas angiotensin-I test (Pharmacia Diagnostics, Uppsala, Sweden). ANP was measured by RIA (21). The blood as well as urine samples of each subject were analyzed in the same assay run. Statistical methods The effect of GA on plasma cortisol and cortisone, body weight, and blood pressure was established by comparing these parameters after GA administration (days 4-10) with the means of the values measured on days 1-3 of the study. The urinary parameters after GA intake were compared with the basal values measured on day 1 of the study. Statistical analysis was performed using repeated measures analysis of variance with multiple comparison procedures by means of the contrast option of the GLM procedure (SAS Institute, Inc., Cary, NC). For the comparison of blood parameters between days 1 and 10 the paired t test was applied. For comparison of the maximum and overall responses of plasma cortisol and cortisone to synthetic ACTH before and after GA administration the Wilcoxon signed rank test was performed. For all procedures P values of 0.05 were required for statistical significance. All results are expressed as the mean ± SEM unless stated otherwise. Approval for the study was obtained from the Ethical Committee of the University Hospital Nijmegen; informed consent was obtained from all subjects.
Results Mineralocorticoid effect
Laboratory methods Plasma cortisol, cortisone, and aldosterone as well as urinary free (i.e. extractable by organic solvents) cortisol and cortisone were measured after extraction and paper chromatography, as described previously (19, 20). In short, 0.2 mL urine or plasma (2.0 mL plasma in the case of aldosterone) were incubated with 10,000 dpm recovery tracer. The samples were extracted with 15.0 mL dichloromethane. The aqueous layer was removed, and the organic phase dried under a stream of air. Separation of the different steroids was obtained by descending paper chromatography and a modified Bush B5 system (toluene: methanol: water = 2:1:1) as eluent. The chromatogram was scanned for radioactivity, and the appropriate area was cut and eluted in ethylene glycol in water (0.2%). The eluates were subjected to
The administration of GA induced a mineralocorticoid effect, as established by a significant increase in plasma sodium and urinary potassium (day 3-7) as well as a significant fall of plasma potassium, PRA, and aldosterone; plasma ANP was significantly enhanced after GA administration (Table 1). The ratio of urinary sodium to potassium was significantly lower on days 3-5 and 8 of the study. Blood pressure was 112/73 ± 10/6 (mean ± SD) mm Hg in the supine position and 117/80 ± 9/5 mm Hg after 2 min in the standing position. The intake of GA did not influence either supine or standing blood pressure in the present study. During the administration of GA four volunteers (two men and two women) com-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 November 2015. at 07:03 For personal use only. No other uses without permission. . All rights reserved.
EFFECT OF GA ON CORTISOL AND CORTISONE TABLE 1. Mineralocorticoid effect of GA administration
PRA(ng/Ls) Aldosterone (pmol/L) ANP (pmol/L) Sodium (mmol/L) Potassium (mmol/L) Bicarbonate (mmol/L) Creatinine (/xmol/L)
Day 1
Day 10
0.37 (0.06) 364 (75) 9.0 (1.1) 141 (0.5) 3.7 (0.9) 27.3 (4.8) 79 (4.0)
0.13 (0.03)° 166 (96)" 28.3 (4.2)° 144 (0.6)° 3.3 (0.4)° 29.0 (6.5)° 79 (4.5)c
Plasma parameters of mineralocorticoid action of GA before (day 1) and after (day 10) GA administration. Values are means, with the SEM in parentheses. °P
