Effect of trilostan on steroid excretion in man: Compensated inhibition of 3β-hydroxysteroid dehydrogenase

J. steroid Biochem. Vol. 24, No. 2, pp. 577-580, 1986 Printed in Great Britain. All rights reserved

Copyright 0

EFFECT

0022.4731/86 $3.00 + 0.00 1986 Pergamon Press Ltd

OF TRILOSTAN ON STEROID EXCRETION MAN: COMPENSATED INHIBITION OF 3P-HYDROXYSTEROID DEHYDROGENASE H.

VIERHAPPER,

P. NOWOTNY

and W.

IN

WALDH~~USL

Division of Clinical Endocrinology and Diabetes Mellitus, I. Medizinische Universitltsklinik, Lazarettgasse 14, A-1090 Vienna, Austria (Received 18 April 1985) Summary-Trilostan (240, 360 and 480 mg/day administered p.o. to healthy men) induced an increase (P < 0.05) in the urinary excretion rates of DHEA, androstenediol and pregnenolone but failed to influence the excretion rates of cortisol, aldosterone, and of the four glucocorticoid metabolites, THF, allo-THF, THE and allo-THE. A rise in plasma renin concentrations was seen in the initial phase of the

trial. Administration of dexamethasone in addition to trilostan suppressed plasma and urinary concentrations of cortisol and the excretion rates of all estimated steroid metabolites but did not modify the relative abundance in the excretion rates of DHEA, androstenediol and pregnenolone. These results confinn that trilostan interferes with the activity of 31-hydroxysteroid dehydrogenase in man. However, under physiological conditions production of cortisol and aldosterone is kept at a constant level, most likely by compensatory stimulation of the secretion of ACTH and renin.

INTRODUCTION

Trilostan ((2a,4a&17fl)-4,5-epoxy-17-hydroxy-3oxoandrostane-2a-carbonitrile [l] is an inhibitor of 3/?-hydroxysteroid dehydrogenase (3P-HSD). The activity of the compound has been shown repeatedly both in vitro [2] and in several animal models [3-51. Since inhibition of 38-HSD will impair the production of cortisol by the zona fasciculata and of aidosterone by the zona glomerulosa, Trilostan was employed to treat patients with either Cushing’s syndrome [6, 71 or with primary aldosteronism [8], and a reduction in cortisol secretion was indeed achieved in some [6], though not in all [7] patients. The reports of therapeutic failures have led us to investigate the relationship between the excretion rates of precursors vs products of 38-HSD in 5 healthy male volunteers. EXPERIMENTAL

Experimental protocol

Five healthy, non-obese volunteers aged 20-28 years who had been carefully informed about the aim and the possible risks of the study gave their written consent to participate. The subjects were given Trilostan (Winthrop GmbH, Hamburg, Germany) p.o. in increasing doses of 24Omg/day (60 mg. q.i.d.), 360 mg/day (60 mg q.i.d.) and 480 mg (120 mg q.i.d.) for 3 days each. The trial was completed by a 3-day period of administration of both, Trilostan (480 mg/day) and dexamethasone (Dexamethason Linz, Chemie Linz, Austria, 0.5 mg q.i.d.). Samples of 24 h collections of urine were obtained daily for 3 days prior to the administration of Trilostan and on 571

day 3 of each administered dose of the drug(s) analysis of steroid excretion rates by GC and for determination of cortisol and aldosterone. At same time intervals blood samples were obtained 08:OOh to determine the plasma concentrations cortisol and renin (PRC).

for the the at of

Gas chromatography

Hydrolysis, extraction and derivatization of the urine samples were performed as previously reported [9]. Gas chromatography (GC) was carried out using a Packard Instruments 428 gaschromatograph equipped with a 50 m OV 101 wall coated open tubular column, a solid injection device and a flame ionization detector. Following sample injection (300°C) the temperature of the oven was immediately programmed between 180°C and 265°C at 0.4”C/min followed by an isothermal run for 15 min at the final temperature. Steroids were identified by comparison of retention times expressed as methylene units (MU) as determined by intrapolation between those of 2 n -alkanes (n -Cs4H,, and n-C,* H,, ; Table 1); excreted amounts were calculated using the area under the respective peaks and corrected for injected amounts, response factors of individual steroids, derivatization yield and aliquotization of the original urine samples [9]. Trivial names of steroid metabolites determined by GC and of other steroids mentioned in this report are given in Table 1. Three unidentified peaks at 30.80MU, 31.10MU and 31.86 MU were observed in the urine samples obtained during administration of Trilostan but not under basal conditions. These peaks apparently represent metabolites of Trilostan.

578

H. VIERHAPPER et al. Table 1A. Gas ChromataEraphic characteristics of derivatized standard steroid compounds mentioned in this paper

1 2 3 4 6 7 9

14 15 19

;; 35 36 37 43

Trivial name

Steroid

No.

Sa-Androstane-3a-ol-17-one Sb-Androstane-3a -ol- 17-one 5-Androstene-3f?-ol-]‘I-one 5-Androstene-3b,l7fl-diol Sa-Androstane-3a-ol,ll,17-dione 5fl-Androstane-3a-ol,l I ,17-dione Sa-Androstane-3a-I lb-dial-17.one 58-Androstane-3a-I lj?-dial-17.one 5-Pregnene-3fi-ol-20-one 5~-Pregnane-3cc-17a-,2l-triol-l1,20-dione Sa-Pregnane-3a-17lx-,21-triol-l l,tO-dione 5p-Pregnane-3n-I l~,l7u,2l-tetrol-2O-one 5a-Pregnane-3a,ll~,?7a,21-tetrol-20-one

Methylene units (as MO-TMS derivatives)

n-C,,-Alkane Androsterone Etiocholanolone Dehvdroeoiandrosterone Andiostenediol I I-Ketoandrosterone I I-Ketoetiocholanolone I IS-Hydroxyandrosterone I Ig-Hydroxyetiocholanolone Pregnenolone Tetrahydro-Cortisone Allotetrahydro-Cortisone Tet~hydro-Cortisoi Allotetrahydro-Corti~l n-C,,-Alkane

(TH-E) (al10 TH-E) (THF) (allo TH-F)

24.00 25.09 25.28 25.68 25.92 26.0 I 26.14 26.96 27.13 21.67 29.66 30.25 30.27 30.37 32.00

Table IB. Trivial names of other steroids mentioned in this paper: 4-Androstene-1 lb-ol-3,17-dione 4-Androstene-3,11,17-trione 4-Pregnen-I lfi,17a-diol-3.20-dione 4.Pregnene-l7a,21-diol-3,I 1.20.trione 4-Pregnene-I I~,l7a,21-triol-3,20-dione 4-Prennene-I 18.21-dial-3.IS.20-trione

Other determiantions Plasma concentrations of renin (PRC) were determined by radioimmunoassay as described previously [lo]. Results are expressed as Goldblatt Units (GU). Prepurification by TLC was employed prior to the radioimmunological determination of cortisol and aldosterone [ 11, 121. Statistics Data are presented as mean Ifr SD. Student’s t-test (2-tailed) for matched pairs was used to compare the data obtained on day 3 of each experimental phase with the respective basal values.

RESULTS

The urinary excretion rates of various steroid metaboiites as determined by capillary GC prior to and during administration of Trilostan are summarized in Table 2. The observed rise in the combined excretion rates of pregnenolone, androstenediol and DHEA from basal, 1.16 + 0.75 mg/24 h up to 4.52 f 2.03 mg/24 h (P < 0.05) was largely due to the rise in the excreted amounts of DHEA (Table 2). The excretion rates of the four glucocorticoid metaboiites, tetrahydrocortisone (THE), allotetrahydrocortisone (allo-THE), tetrahy~oco~isol (THF) and alIotetrahydr~ortiso1 (allo-THF} remained unchanged during the administration of Trilostan. This resulted in a rise (P < 0.05) from 0.40 _t 0.24 to 1.33 + 0.51 in the calculated ratio of the combined excretion rates of DHEA, androstenediol and pregnenolone versus those of the THE, allo-THE, THF and allo-THF. This increased ratio was also apparent during the additional administration of dexamethasone, although excretion

I Ifi-Hydroxyandrostenedione I 1-Ketoandrostenedione

(II-OH-A) (II-O-A)

Zl-Desoxycortisol Cortisone Cortisol Aldosterone

rates of all estimated steroid metabolites were suppressed. The excretion of I I-hydroxyandrosterone, 1I-ketoandrosterone, 1 I-hydroxyetiocholanolone, f Iketoetiocholanolone and the relationship of these metabolites to each other was unchanged during the administration of Trilostan, whereas dexamethasone induced the expected suppression of these metabolies. The excretion of androsterone was enhanced by Trilostan at single points of observation but no changes were seen in the excretion rates of etiocholanolone. Thus, a minor rise in the ratio of excreted androsterone versus etiocholanolone was observed when the subjects were given 480 mg Trilostamday. Attention should be given to the non-homogeneity in the behaviour of cortisol excretion rates which resulted in the large standard deviations reported in Table 3. A dose-related suppression of cortisol secretion by Trilostan was apparent neither from urinary nor from plasma concentrations of cortisol, although the latter were below basal levels at single points of observation (Table 3). A slight rise (P < 0.05) in plasma renin cancentrations was observed, but the excretion rates of aldosterone were unchanged by Trilostan. DISCUSSION

In the healthy men investigate in this study Trilostan caused a rise in the calculated ratio of the combined excretion rates of DHEA, pregenolone and androstenediol versus those of the glucorticoid metabolites, THE, allo-THE, THF and allo-THF. Since the former three steroid metabolites are not dependent on the activity of 3/I-HSD these results confirm that Trilostan interferes with the activity of 3/I-HSD in man. However, the excretion rates of

(=2+3) (=2/3)

(=7+9+14+15) (=7+ 14/9+ 1.5) (2) (3)

I::;

(=31 +35+36+37) (Sum l/Sum 2) (7) (9)

16f 19) (31) (35 + 36) (37)

(=4+

(4) (6) (19)

(Peak number)

7.oOrF_1.13 I .43 + 0.60

I .78 _+0.44 4.86 + 3.77 3.98 + 0.89* 2.89 + 0.98

1.58+ 0.46 6.39 f 6.19 2.91 + 1.25 2.52 & 1.59 5.48 + 2.33 I .43 + 0.85

3.98 + 1.32 1.14~0.51* 0.37 * 0.49 0.18t0.11 1.02+0_13 0.24rtO.10

4.55 & 2.02’ 1.81 + 0.58 1.33 2 0.35 0.84 & 0.64

3.60 I 1.86 0.40 & 0.24 0.27 f 0.29 0.36 + 0.35 0.80 + 0.47 0.14 * 0.10

1.16-tO.75 2.09 f 1.30 1.04k 0.34 0.47 _e0.33

T (240) 3.44 * I .83” 0.59 * 0.39 0.55 rt 0.32

BaSaI 0.65 * 0.57 0.13 f 0.07 0.26 + 0.28

T (360)

5.99 i: 2.95 1.83+ 0.66

I .35 I 0.64 6.08 & I .93 3.81 ct 2.09 2.18 & 1.02

3.74 + 2.00 1.14+0.41* 0.33 t 0.37 0.07 + 0.05 0.83 rt 0.34 0.12 f 0.05

4.07 rf: 2.08 1.61 f 1.12 0.64 t 0.47 0.51 t 0.50

3.03 i 1.86’ 0.35 & 0.30 0.69 cf 0.39

Basal

9.2 + 6.6

ir 17.8

L-6.5 7.5 + 1.7 and 2.0 mg Dexamethasonejday + 2.6 4.8 + 3.3

9.6 & 4.1

Aldosterone @/day) 8.5 I4.9

+ 13.9

Urine

(*)P < 0.05 as compared to basal values.

240 mg/Trilostan/day 31.6 360 mg Trilostan/day 23.0 4800 mg Trilostan/day 20.2 480 mg Trilostan/day 4.3

__I__-. Cortisol @g/day) 27.6 + 14.4

Plasma

I .5*

14.2 I: 1.8* 0.5 + o.o*

I.18 kO.73

15.1 & 2.9

14.9 *

Cortisol Wdl) 16.1 +_1.8

. _- -.----

1.26 & 0.53

1.05 + 0.40*

0.96 + 0.25

PRC (GU x 10 E-4/ml) 0.75 * 0.35

,_~_~.

Table 3. Excretion rates of cortisol and aldosterone and plasma concentrations of renin (PRC) and cortisol in healthy men (n = 5, I k SD) before (basal: mean of 3 dete~inatio~s) and during (day 3 of the resoective doses) treatment with Trilostan and with dexamethasone

(*)P < 0.05 as compared to basal excretion rates.

Sum -.- 4 Ratio 3

Sum 3 Ratio 2 Androsterone Etiocholanolone

Sum 2 Ratio I I I-0-Androsterone 1I-O-Etiocholanolone I I-OH-Androsterone 1I-OH-Etiocholanolone

S& THE a-THE + THF a-THF

DHEA Androstenediol Prcgnenolone

Steroid

2.03* 1.00 0.40 0.66

5.19 f 2.28 I .75 + 0.77’

1.61f 0.47 10.50 If- 12.72 3.45 f 1.21* 2.34 + 1.85

3.64 f I .98 1.33 *0.51* 0.35 f 0.33 0.11 *0.08 1.02+ 0.26 0.14 + 0.08

4.52 + 1.91 i I .06 + 0.67 +

2.23 + 2.03 0.77 f 0.60 0.52 + 0.48

T (480)

5 + + +

0.56 0.09 0.08* 0.05’

2.07 + 0.22 1.32f 0.52

0.56 f 0.33* 6.24 i 5.45 1.16+0.30 0.91 +0.10

0.22 F o.os* 2.54 i 1.65 0.25 f 0.35 0.04 + 0.02 O.ZO~O.i4 0.08 + 0.07

0.55 0.09 0.08 0.05

0.16*0.14 0.45 + 0.03 0.35kO.51

T (480) + D (2)

Table 2. Steroid excretion rates in healthy men (n = 5, P f. SD) before (basal: mean of 3 dete~inations) and during (day 3 of the respective doses) treatment (mg/day) with Trilostan (T) and Dexamethasone (D)

580

H. VIERHAPPER et al.

glucocorticoid metabolites per se were unchanged by Trilostan. By analogy with the pathophysiology of congenital adrenal hyperplasia it appears, therefore, that the Trilostan-induced deficiency in 3/I-HSD activity is compensated for by a rise in ACTH secretion which results in an altered relationship of precursors vs products of this enzyme but not necessarily in a decrease in cortisol production. This counter-regulatory rise in ACTH is apparently more or less effective in various individuals, explaining the transient, though not dose-related, fall in plasma and urinary cortisol concentrations in some individuals. The heterogeneity in the effectiveness of Trilostan in suppressing cortisol secretion has been reported by others [6, 71. On a practical level this observation cast doubt on the usefulness of Trilostan in the therapy of pituitary-dependent forms of Cushing’s syndrome, since a compensatory rise in ACTH secretion in these patients is likely to induce an escape from the drug’s effect in regard to inhibition of glucorticoid production. On the other hand the drug may be useful in patients with suppressed endogenous ACTH-secretion, i.e. in patients with glucocorticoidproducing adrenal neoplasms where surgery is not possible on clinical grounds. Treatment with Trilostan may also prove to be an alternative to bilateral adrenalectomy in patients with bilateral macronodular adrenal hyperplasia [ 131. Although our study was performed on an outpatient basis with uncontrolled diet, it seems safe to say that a marked effect of Trilostan on aldosterone excretion was clearly absent in the investigated subjects. Rather, the small rise in plasma renin concentration in the presence of unchanged aldosterone excretion rates seems to indicate a compensatory mechanism via enhanced production of angiotensin II similar to the one postulated above for cortisol [6]. C,,O, precursors such as DHEA contribute to the excretion of androsterone and etiocholanolone in about equal proportions [ 141.Nevertheless, since the 5p-configuration prevails in the glucocorticoidderived CnO, metabolites 1151the relative lack of the SjI-hydrogenated metabolite etiocholanolone and the predominance of its See-hydrogenated homologue androsterone during treatment with Trilostan may result from the increase in DHEA secretion. On the other hand, Trilostan interfered neither with the secretion of C,,O, steroids nor with the relationship of Sa-hydrogenated (1 1-hydroxyandrosterone and 1I -ketoandrosterone) and S/?-hydrogenated (1 Ihydroxyetiocholanolone and 1l-ketoetiocholanolone) C,,O, steroids. Since the former two metabolites are derived mainly from 21-desoxycortisol and 11-ketoandrostenedione whereas the latter are products of cortisol and cortisone the unaltered relationship between these two groups indicates that

Trilostan activity.

does not inferfere

with Zl-hydroxylase

Acknowledgements-The valuable technical assistance of MS A. Fiirst, MS E. Nowotny and MS H. Kieweg is gratefully acknowledged. REFERENCES 1. Neumann H. C., Potts G. O., Ryan W. T. and Stonner

F. T.: Steroidal heterocycles. XIII. 4c@epoxy-5aandrost-2-eno(2,3-d)isoxazoles and related compounds. J. med. Chem. 13 (1970) 948-951. 2. Potts G. O., Creange J. E.. Harding H. R. and Schane P.: Trilostane, an orally active inhibitor of steroid biosynthesis. Steroids 32 (1978) 257-267. 3. Levine J. H., Nicholson W. E., Peytremann A. and Orth D. N.: The mechanism of ACTH stimulation of adrenal ornithine decarboxylase activity. Endocrinology 97 (1975) 136144. 4. Touitou Y., Auzeby A.. Bogdan A.. Luton J.-P. and Galan P.: I ID-hydroxy- 1I -ketosteroids equilibrium, a source of misinterpretation in steroid synthesis: evidence through the effects of trilostane on 1I/?-hydroxysteroid dehydrogenase in sheep and human adrenals in oitro. J. steroid Biochem. 20 (1984) 763-768. 5. Schane H. P., Potts G. 0. and Creange J. E.: Inhibition of ovarian, placental, and adrenal steroidogenesis in the rhesus monkey by trilostane. Fert. Steril. 32 (1979) 464467.

6. Komanicky P., Spark R. F. and Melby J. C.: Treatment of Cushing’s syndrome with trilostane (WIN 24,540), an inhibitor of adrenal steroid biosynthesis. J. clin. Endocr. Metab. 47 (1978) 1042-1051.

I. Dewis P., Anderson D. C., Bu’lock D. E., Eamshaw R. and Kelly W. F.: Experience with trilostane in the treatment of Cushing’s syndrome. Clin. Endocr. 18 (1983) 533-540. 8. Winterberg B. and Vetter H.: Die Behandlung des primaren Aldosteronismus mit Trilostan. Schwei:. med. Wochenschr. 113 (1983) 1735-1738.

9. Vierhapper H., Nowotny P., Waldhausl W. and Frisch H.: Capillary gas chromatography as a tool for characterization of urinary steroid excretion in patients with congenital adrenal hyperplasia. J. steroid Biochem. 22 (1985) 363-369.

10. Waldhausl W. and Lewandovsky J. A.: Measurement of plasma renin concentration (PRC) using exogenous substrate and radioimmunoassay. Eur. J. cfin. Invest. 3 (1973) l-9. 11. Vierhapper H., Waldhiusl W. and Nowotny P.: Suppression of luteinizing hormone induced by adrenocorticotrophin in healthy women. J. Endocr. 91 (1981) 399403.

12. Herkner K., Nowotny P. and Waldhausl W.: Die Bestimmung von Aldosteron in Harn und Plasma. J. Chromat. 146 (1978) 273-281. 13. Smals A. G. H., Pieters G. F. F. M., van Haelst U. J. G. and Kloppenborg P. W. C.: Macronodular adrenocortical hyperplasia in long-standing Cushing’s disease. J. c/in. Endocr. Metab. 58 (1984) 25-31. 14. Fukushima D. K., Bradlow H. L.. Dobriner K. and Gallagher T. F.: The fate of testosterone infused intravenouslv in man. J. biol. Chem. 206 (19541 863-874. 15. Keutmann E. H. and Mason W. B.: Indivihual urinary 17-ketosteroids of healthy persons determined by gas chromatography: biochemical and clinical considerations. J. clin. Endocr. Metab. 27 (1967) 40&420.